ANGIOMOTIN-LIKE PROTEIN 1-DERIVED MOLECULES FOR THE MODULATION OF ANGIOGENESIS AND THE TREATMENT OF TUMOURS

The present invention relates to the modulation of angiogenesis and/or tumour formation

The development of the embryonic vascular system into a highly ordered network requires precise control over the migration and branching of endothelial cells (ECs).

The vascular structure of the developing embryo is formed through two consecutive processes (i) vasculogenesis, which is the formation of a primitive vascular plexus and (ii) angiogenesis, which is the remodeling and maturation of the initial vascular structure (Risau 1997). EC migration, proliferation and polarization play fundamental roles in establishing a functional blood vessel network.

The development of this network is tightly controlled by growth factors and their receptors, as well as integrins and their extra cellular matrix (ECM) ligands (YancopouJos et al. 1998; Hynes 2002). VEGF is a major regulator of embryonic vascular development (Carmeliet et al. 1996; Ferrara et al. 1996) and exerts its function by binding to two high affinity receptor tyrosine kinases, VEGF receptor 1 (VEGFR1 ) and VEGF receptor 2 (VEGFR2) (Olsson et a/.2006), and the co- receptors neuropilin (NP) 1 and 2 (Soker et al. 1998; Gluzman-Poltorak et al. 2000).

The development of the embryonic vascular system into a highly ordered network requires strict control over the sprouting ECs that are involved in the migration and branching of the early vasculature (Jones et al. 2006). The cells at the tip of the sprout sense the environment, whereas the stalk cells proliferate and form a lumen. Whether cells migrate or proliferate is controlled by VEGF, where a gradient of VEGF induces migration of the tip cells and the concentration of VEGF regulates proliferation of the stalk cells (Gerhardt et al. 2003).

It has been shown that the membrane associated protein Angiomotin (Amot) is involved in the control of cell migration (Troyanovsky et al. 2001 ; Levchenko et al.

2003). Furthermore, it binds to and mediates the anti-migratory properties of

angiostatin through an angiostatin-binding domain (Troyanovsky et al. 2001 ; Bratt et al. 2005).

Angiomotin belongs to a protein family that also comprises two additional members, Angiomotin like- proteins 1 and 2 (AMOTL-1 and AMOTL-2). These proteins are all characterized by a glutamine rich domain, a conserved coil-coil domain, and a C4 terminal PDZ-binding domain (Bratt et al. 2002). Angiomotin itself is expressed in two different isoforms with distinct functions, p80-Amot enhances cell migration and stabilizes tubes in vitro, whereas p130-Amot associates to actin and affects cell shape (Troyanovsky et al. 2001 ; Levchenko et al. 2004; Ernkvist et al. 2006).

The functional importance of the PDZ-binding domain is indicated by the migratory defect displayed by cells expressing a C-terminal mutant form of Amot (Levchenko et al. 2003). Transgenic mice expressing the mutant Amot under the EC specific Tie promoter lose their response to growth factors, which leads to insufficient vascularization and death around E 9.5 (Levchenko er al. 2003). In addition to the important function during cell migration, Wells and coworkers have recently shown that the PDZ binding domain of Amot functions as a link between Cdc42-activity and the PaIsI , Patj, and Par-3 polarity protein complex in epithelial cells (Wells et al. 2006).

The importance of Amot and AMOTL proteins during angiogenesis has now been characterised by examining Amot-deficient mice and Amot-knockdown zebrafish embryos.

Amot-deficiency has now been shown to cause vascular defects, including dilated brain vessels and a defective intersomitic vasculature. The work demonstrates an important role for Amot during tube formation, with defects in EC polarization and growth factor-induced migration.

Cancer is a major health problem causing great deal of suffering a high number of mortalities. Early diagnosis followed by surgery is so far the only way to cure the disease. Treatment of patients with advanced metastatic disease is still unsatisfactory with high treatment failure. Pre-clinical evidence show that angiogenesis, the process where blood vessels are formed from pre-existing

ones, play a pivotal in the growth of local and distant tumors. In patients, micro vessel density and production of vascular endothelial growth factor (VEGF) correlates to the incidence of metastasis and survival. There is now also evidence that anti-angiogenic therapy is beneficial for patients with metastatic cancer. Firstly, anti-angiogenic schedules of chemotherapy, "metronomic chemotherapy" has been used to target tumor vessels. Patients treated with low dose methotrexate and cyclophosphamide showed an overall response rate of 19% (stable disease in 13% of the patients). Secondly, phase III studies with Avastin (antibodies against VEGF) in combination with chemotherapy for the treatment of patients with spread colorectal, lung and breast cancer was presented at ASCO 2003 and 2005. This study found that Avastin improved time to progression by 2- 6 months and increased survival by approximately 50 %.

Pre-clinical and clinical evidence has shown that targeting the VEGF receptor pathway inhibits angiogenesis and tumor growth. However, tumors are continually demonstrating their possession of an increasing number of angiogenesis stimulators during tumor progression, there is a need for additional strategies to target vessel formation. An attractive alternative are the endogenous angiogenesis inhibitors that have been identified. One of the first inhibitors to be reported, angiostatin, was shown to be specific for endothelial cells and could maintain dormancy of established metastases in vivo. However, the pharmacodynamics of angiostatin and other inhibitors constitute a major obstacle for the use of these agents in cancer patients.

Although these are significant and encouraging results they also indicate that targeting a single pathway that triggers angiogenesis is just not enough. This may be explained by the findings that malignant cells produce a plethora of factors that trigger vessel formation. Indeed, multiple angiogenic factors are commonly expressed by invasive breast cancers. It is therefore highly likely that over time tumors will elaborate alternative signaling pathways to trigger vessel growth.

In a strategy to identify the angiostatin receptor by yeast-two hybrid system we have previously found a novel protein named angiomotin (AMOT, Troyanovsky et a/. JCB 2001 ). This is a membrane associated protein that mediates the inhibitory effects of angiostatin in vitro. In vivo AMOT is expressed in angiogenic vessels.

We have further shown that όverexpression of the shorter isoform, p80, confers a hyper-migratory and invasive phenotype in transfected cells (Levchenko et al. Oncogene 2004). The publication of a DNA vaccine that efficiently inhibits angiogenesis and breast cancer tumor growth was an important milestone in this project (Holmgren et al. PNAS 2006) and was highlighted in Nature Reviews in Cancer. (http://www.nature.com/nrc/journal/v6/n7/full/nrc1938.html). The inventors have now worked on the targeting of angiomotin-like protein 1 (AMOTL1) in order to reduce tumour growth via angiogenesis inhibition.

In a first aspect of the invention there is provided an siRNA molecule for modulating angiogenesis and/or tumour formation wherein the siRNA is complementary to the nucleotide sequence encoding angiomotin-like protein 1 or fragments or variants thereof.

By "polynucleotide" we include single-stranded and/or double-stranded molecules of DNA (deoxyribonucleic acid) and/or RNA (ribonucleic acid) and derivatives thereof. By "encoding polynucleotide" we include a polynucleotide the sequence of which that may be translated to form a desired polypeptide.

It has now also been found that angiogenesis inhibition can be induced using siRNA (small interfering RNA molecules).

RNA interference (RNAi) is a natural mechanism for silencing specific genes. Genes provide cells with the instructions for making proteins, and when a gene is silenced, the cell stops making the protein specified by that gene. RNA interference was first observed in plants, but the first crucial breakthrough in understanding the RNAi mechanism came from studies of worms. This came in 1998 with the recognition that double-stranded RNA (dsRNA) played a pivotal role in RNAi. The first evidence for in vivo silencing of genes using siRNA was published 2002 (McCaffrey, A. P., Meuse, L., Pham, T. T., Conklin, D. S., Hannon, G. J. and Kay, M. A. (2002) RNA interference in adult mice. Nature (London) 418, 38-39) followed by the publication of Song et al. that showed that siRNA may be used for therapeutic intervention. This study showed that RNA interference targeting Fas protected mice from fulminant hepatitis. (Song, E., Lee, S. K., Wang, J. et al. (2003) Nat. Med. 9, 347-351).

siRNA molecules may be single-stranded (ss) or double-stranded (ds). The siRNA molecules may be delivered using a construct, which is capable of expressing the siRNA molecule upon delivery to the target cell.

A "small interfering RNA" or "short interfering RNA" or "siRNA" or "short hairpin RNA" or "shRNA" is a double-stranded RNA molecule that is complementary to a target nucleic acid sequence. A double-stranded RNA molecule is formed by the complementary pairing between a first RNA portion and a second RNA portion. The length of each portion generally is less than 30 nucleotides in length (e.g., 29, 28, 27, 26, 25, 24, 23, 22, 21 , 20, 19, 18, 17, 16, 15, 14, 13, 12, 1 1 , or 10 nucleotides). In some embodiments, the length of each portion is 19 to 25 nucleotides in length. In some siRNA molecules, the complementary first and second portions of the RNA molecule are the "stem" of a hairpin structure. The two portions can be joined by a linking sequence, which can form the "loop" in the hairpin structure. The linking sequence can vary in length. In some embodiments, the linking sequence can be 5, 6, 7, 8, 9, 10, 1 1 , 12 or 13 nucleotides in length. A representative linking sequence is 5'-TTC AGA AGG-3', but any of a number of sequences can be used to join the first and second portions. The first and second portions are complementary but may not be completely symmetrical, as the hairpin structure may contain 3' or 5' overhang nucleotides (e.g., a 1 , 2, 3, 4, or 5 nucleotide overhang).

RNA molecules have been shown by many researchers to be effective in suppressing mRNA accumulation. siRNA-mediated suppression of nucleic acid expression is specific as even a single base pair mismatch between siRNA and the targeted nucleic acid can abolish the action of RNA interference. siRNAs generally do not elicit anti-viral responses.

Preferably the siRNA is an antisense polynucleotide which is capable of hybridising to the nucleotide sequence encoding angiomotin-like protein 1 or fragments or variants thereof under stringency conditions of 2xSSC at 65 ⁰ C.

In a second aspect of the invention there is provided a nucleotide sequence encoding a siRNA as described in the first aspect of the invention.

In a third aspect of the invention there is provided an expression vector containing a nucleotide sequence encoding the siRNA as described in the first aspect of the invention.

In a fourth aspect of the invention there is provided a host cell producing a siRNA as described in the first aspect of the invention resulting from the expression of a nucleotide sequence encoding the siRNA.

In a fifth aspect of the invention there is provided a method of treating angiogenic disease comprising administering an effective amount of at least one siRNA molecule as as described in the first aspect of the invention.

Preferably the method also comprises administering an effective amount of an siRNA molecule complementary to the nucleotide sequence encoding angiomotin or fragments or variants thereof.

In a sixth aspect of the invention there is provided an siRNA molecule as defined in the first aspect of the invention for use in medicine.

In a further aspect of the invention there is provided a use of at least one siRNA molecule as defined in the first aspect, in the manufacture of a medicament for modulating angiogenesis and/or tumour formation.

Preferably the medicament prevents and/or reduces angiogenesis and/or tumour formation.

In a further aspect of the invention there is provided a use of at least one siRNA molecule as described in the first aspect of the invention in the manufacture of a medicament for treating a subject with an angiogenesis-related disease or disorder.

In a further aspect of the invention there is provided a use of at least one siRNA molecule as described in the first aspect of the invention in the manufacture of a

vaccine for vaccinating a subject with, or at risk of, angiogenesis and/or tumour formation and/or an angiogenesis-related disease or disorder.

Preferably the medicament further comprises at least one siRNA molecule complementary to the nucleotide sequence encoding angiomotin or fragments or variants thereof.

Conveniently the angiogenesis-related disease or disorder is cancer, a solid tumour, haemangioma, ocular neovascularisation, diabetic retinothapy, macular degeneration, proliferative ischemic retinapathy, corneal neovascularisation, rheumatoid arthritis, inflammatory conditions, psoriasis, chronic inflammation of the intestines, asthma or endometriosis.

In a further aspect of the invention there is provided a use of at least one siRNA molecule as described in the first aspect of the invention in the detection and/or measurement of angiogenesis and/or tumour formation in a test sample.

Preferably the siRNA molecule is accompanied by at least one further siRNA molecule that is complementary to the nucleotide sequence encoding angiomotin or fragments or variants thereof.

In a further aspect of the invention there is provided a pharmaceutical composition for modulating angiogenesis and/or tumour formation, comprising an effective amount of at least one siRNA molecule as described in the first aspect of the invention and a pharmaceutical excipient or diluent.

Preferably the pharmaceutical composition further comprises at least one siRNA molecule complementary to the nucleotide sequence encoding angiomotin or fragments or variants thereof.

Conveniently the pharmaceutical composition prevents and/or reduces angiogenesis and/or tumour formation.

In a further aspect of the invention there is provided a vaccine for modulating angiogenesis and/or tumour formation, comprising an effective amount of an

siRNA molecule as described in the first aspect of the invention and an exipient or diluent.

Preferably the vaccine or pharmaceutical composition further comprises at least one additive for assisting or augmenting the action of the siRNA molecule(s) therein.

Conveniently the at least one additive is an immunostimulatory molecule.

Advantageously the immunostimulatory molecule is a cytokine or polynucleotide (and/or antisense polynucleotide) encoding a cytokine. For example, the antisense polynucleotide is at least one siRNA molecule complementary to the nucleotide sequence encoding angiomotin or fragments or variants thereof.

Preferably the vaccine comprises a cell or cell extract, wherein the cell may be an antigen presenting cell which is loaded with the siRNA molecule as described in the first aspect of the invention or transfected with the encoding polynucleotide.

Conveniently the cell is a tumour cell expressing angiomotin or an endothelial cell expressing angiomotin.

Advantageously the siRNA molecule(s) of the invention is administered to the eye. Preferably the siRNA molecule(s) is administered intravitreally or in the form of an eye bath or eye drop.

In a further aspect of the invention there is provided a use of an antibody or an antigen-binding fragment thereof, which binds specifically to either (a) angiomotin-like protein 1 or fragments or variants thereof or (b) a polynucleotide encoding angiomotin-like protein 1 or fragments or variants thereof, in the manufacture of a medicament for modulating angiogenesis and/or tumour formation.

Antibodies comprise two identical polypeptides of M _r 50,000-70,000 (termed

"heavy chains") that are linked together by a disulphide bond, each of which is linked to one of an identical pair of polypeptides of M _r 25,000 (termed "light chains"). There is considerable sequence variability between individual N-termini

of heavy chains of different antibody molecules and between individual light chains of different antibody molecules and these regions have hence been termed "variable domains". Conversely, there is considerable sequence similarity between individual C-termini of heavy chains of different antibody molecules and between individual light chains of different antibody molecules and these regions have hence been termed "constant domains".

The antigen-binding site is formed from hyper-variable regions in the variable domains of a pair of heavy and light chains. The hyper-variable regions are also known as complementarity-determining regions (CDRs) and determine the specificity of the antibody for a ligand. The variable domains of the heavy chain (V _H ) and light chain (V _L ) typically comprise three CDRs, each of which is flanked by sequence with less variation, which are known as framework regions (FRs).

The variable heavy (V _H ) and variable light (V _L ) domains of the antibody are involved in antigen recognition, a fact first recognised by early protease digestion experiments. Further confirmation was found by "humanisation" of rodent antibodies. Variable domains of rodent origin may be fused to constant domains of human origin such that the resultant antibody retains the antigenic specificity of the rodent parented antibody (Morrison et al., 1984, Proc. Natl. Acad. Sci. USA, 81 , 6851-6855).

That antigenic specificity is conferred by variable domains and is independent of the constant domains is known from experiments involving the bacterial expression of antibody fragments, all containing one or more variable domains. These molecules include Fab-like molecules (Better et al., 1988, Science, 240:1041 ). Fv molecules (Skerra et al., 1988, Science, 240, 1038); single-chain Fv (ScFv) molecules where the V _H and V _L partner domains are linked via a flexible oligopeptide (Bird et al., 1988, Science 242:423; Huston et al., 1988, Proc. Natl. Acad. ScL USA, 85:5879) and single domain antibodies (dAbs) comprising isolated V domains (Ward et al., 1989 Nature 341 , 544). A general review of the techniques involved in the synthesis of antibody fragments which retain their specific binding sites is to be found in Winter et al., 1991 , Nature, 349, 293-299.

Preferably, the antibody is a human or humanised antibody or fragment thereof.

Conveniently the antibody fragment of the invention is an scFv molecule or Fab.

By "ScFv molecules" we mean molecules wherein the V _H and V _L partner domains are linked via a flexible oligopeptide.

The advantages of using antibody fragments which have antigen-binding activity, rather than whole antibodies, are severai-foid. The smaller size of the fragments may lead to improved pharmacological properties, such as better penetration of solid tissue. Effector functions of whole antibodies, such as complement binding, are removed. Fab, Fv, ScFv and dAb antibody fragments can all be expressed in and secreted from Escherichia coli (E. cols), thus allowing the facile production of large amounts of the said fragments.

Whole antibodies, and F(ab') ₂ fragments are "bivalent" . By "bivalent" we mean that the said antibodies and F(ab') ₂ fragments have two antigen combining sites. In contrast, Fab, Fv, ScFv and dAb fragments are monovalent, having only one antigen combining site.

Methods for generating, isolating and using antibodies for a desired antigen or epitope are well known to those skilled in the relevant art. For example, an antibody may be raised in a suitable host animal (such as, for example, a mouse, rabbit or goat) using standard methods known in the art and either used as crude antisera or purified, for example by affinity purification. An antibody of desired specificity may alternatively be generated using well-known molecular biology methods, including selection from a molecular library of recombinant antibodies, or grafting or shuffling of complementarity-determining regions (CDRs) onto appropriate framework regions. Human antibodies may be selected from recombinant libraries and/or generated by grafting CDRs from non-human antibodies onto human framework regions using well-known molecular biology techniques.

Methods for formulating polypeptides, polynucleotides and antibodies into medicaments, pharmaceutical compositions and vaccines are well known to

those in the relevant art. Preferred formulations of medicaments, pharmaceutical compositions and vaccines comprising the polypeptides, polynucleotides and antibodies of the invention are described in the Examples.

Antibodies may be used in therapy - for example, a medicament comprising therapeutic antibodies may be introduced into a subject to modulate the immune response of that subject. For example, a therapeutic antibody specific for an antigen in the subject will stimulate an immune response to that antigen, thereby inducing and/or promoting an immune response and aiding recovery. Methods for administering therapeutic antibodies to a patient in need thereof are well known in the art.

In one embodiment the antibody or antigen-binding fragment thereof also binds specifically to (a) angiomotin or fragments or variants thereof or (b) a polynucleotide encoding angiomotin or fragments or variants thereof.

In a alternative embodiment there is provided a use as claimed in wherein the medicament also comprises an further antibody or an antigen-binding fragment thereof, which binds specifically to either (a) angiomotin or fragments or variants thereof or (b) a polynucleotide encoding angiomotin or fragments or variants thereof.

In a further aspect of the invention there is provided a method for treating a subject for angiogenesis and/or tumour formation; the method comprising the step of administering to the subject an antibody or an antigen-binding fragment thereof, which binds specifically to either (a) angiomotin-like protein 1 or fragments or variants thereof or (b) a polynucleotide encoding angiomotin-like protein 1 or fragments or variants thereof.

In a first embodiment the antibody or antigen-binding fragment thereof also binds specifically to (a) angiomotin or fragments or variants thereof or (b) a polynucleotide encoding angiomotin or fragments or variants thereof.

In an alternative embodiment the method also incorporates the step of administering to the subject a further antibody or antigen-binding fragment that

binds specifically to (a) angiomotin or fragments or variants thereof or (b) a polynucleotide encoding angiomotin or fragments or variants thereof.

In a further aspect of the invention there is provided a pharmaceutical composition for modulating angiogenesis and/or tumour formation; comprising an antibody or an antigen-binding fragment thereof which binds specifically to either (a) angiomotin-like protein 1 or fragments or variants thereof or (b) a polynucleotide encoding angiomotin-like protein 1 or fragments or variants thereof and a pharmaceutical carrier, excipient or diluent.

In one embodiment antibody or antigen-binding fragment thereof also binds specifically to (a) angiomotin or fragments or variants thereof or (b) a polynucleotide encoding angiomotin or fragments or variants thereof.

In an alternative embodiment the composition also comprises a further antibody or an antigen-binding fragment thereof, which binds specifically to either (a) angiomotin or fragments or variants thereof or (b) a polynucleotide encoding angiomotin or fragments or variants thereof.

Preferably the angiogenesis-related disease or disorder is cancer, a solid tumour, haemangioma, ocular neovascularisation, diabetic retinothapy, macular degeneration, proliferative ischemic retinapathy, corneal neovascularisation, rheumatoid arthritis, inflammatory conditions, psoriasis, chronic inflammation of the intestines, asthma or endometriosis.

Conveniently the antibody or antigen-binding fragment thereof is human or humanised.

Advantageously the antigen-binding antibody fragment is an scFv or Fab.

Preferably the antibody or antigen-binding fragment thereof binds to an epitope of full length human angiomotin-like protein 1.

Conveniently the antibody or antigen-binding fragment thereof binds specifically to an epitope of a fragment of human angiomotin-like protein 1 , wherein the

fragment has substantially the same angiogenic activity as full length angiomotin- like protein 1.

In a further aspect of the invention there is provided a use of an antibody or antigen binding fragment thereof in the detection and/or measurement of angiogenesis and/or tumour formation in a test sample.

In a further aspect of the invention there is provided a use of angiomotin-like protein 1 or fragments or variants thereof, in the manufacture of a medicament for modulating angiogenesis and/or tumour formation.

Preferably, the medicament also comprises angiomotin or fragments or variants thereof.

In a further aspect of the invention there is provided a method of treating angiogenic disease comprising administering an effective amount of angiomotin- like protein 1 or fragments or variants thereof.

Preferably, the method also comprises administering an effective amount of angiomotin or fragments or variants thereof.

In a further aspect of the invention there is provided a use of angiomotin-like protein 1 or fragments or variants thereof as in the manufacture of a vaccine for vaccinating a subject with, or at risk of, angiogenesis and/or tumour formation and/or an angiogenesis-related disease or disorder.

Preferably the medicament further comprises angiomotin or fragments or variants thereof.

Conveniently the angiogenesis-related disease or disorder is cancer, a solid tumour, haemangioma, ocular neovascularisation, diabetic retinothapy, macular degeneration, proliferative ischemic retinapathy, corneal neovascularisation, rheumatoid arthritis, inflammatory conditions, psoriasis, chronic inflammation of the intestines, asthma or endometriosis.

In a further aspect of the invention there is provided a use of angiomotin-like protein 1 or fragments or variants thereof in the detection and/or measurement of angiogenesis and/or tumour formation in a test sample.

In a further aspect of the invention there is provided a pharmaceutical composition for modulating angiogenesis and/or tumour formation, comprising an effective amount of angiomotin-like protein 1 or fragments or variants thereof and a pharmaceutical exipient or diluent.

Preferably the pharmaceutical composition comprises angiomotin or fragments or variants thereof.

Conveniently the pharmaceutical composition prevents and/or reduces angiogenesis and/or tumour formation.

In a further aspect of the invention there is provided a vaccine for modulating angiogenesis and/or tumour formation, comprising an effective amount of angiomotin-like protein 1 and an excipient or diluent.

Preferably the vaccine further comprises at least one additive for assisting or augmenting the action of the angiomotin-like protein 1 therein.

Preferably the at least one additive is an immunostimulatory molecule.

Conveniently the immunostimulatory molecule is a cytokine or polynucleotide (and/or antisense polynucleotide) encoding a cytokine.

Advantageously the angiomotin-like protein 1 is administered to the eye.

This may be for example be administered intravitreally or in the form of an eye bath or eye drop.

In a further aspect of the invention there is provided a nucleic acid molecule encoding angiomotin-like protein 1.

In a further aspect of the invention there is provided a expression vector containing a nucleotide sequence encoding the angiomotin-like protein 1.

In a further aspect of the invention there is provided a host cell producing angiomotin-like protein 1 resulting from the expression of a nucleotide sequence encoding the angiomotin-like protein 1.

In a further aspect of the invention there is provided a use of a nucleic acid molecule encoding angiomotin-like protein 1 or fragments or variants thereof, in the manufacture of a medicament for modulating angiogenesis and/or tumour formation.

Preferably the medicament also comprises a nucleic acid molecule encoding angiomotin or fragments or variants thereof.

In a further aspect of the invention there is provided a method of treating angiogenic disease comprising administering an effective amount of a nucleic acid molecule encoding angiomotin-like protein 1 or fragments or variants thereof.

Preferably the method also comprises administering an effective amount of a nucleic acid molecule encoding angiomotin or fragments or variants thereof.

In a further aspect of the invention there is provided a use of a nucleic acid molecule encoding angiomotin-like protein 1 or fragments or variants thereof in the manufacture of a vaccine for vaccinating a subject with, or at risk of, angiogenesis and/or tumour formation and/or an angiogenesis-related disease or disorder.

Preferably the medicament further comprises a nucleic acid molecule encoding angiomotin or fragments or variants thereof.

Conveniently the angiogenesis-related disease or disorder is cancer, a solid tumour, haemangioma, ocular neovascularisation, diabetic retinothapy, macular degeneration, proliferative ischemic retinapathy, corneal neovascularisation, rheumatoid arthritis, inflammatory conditions, psoriasis, chronic inflammation of the intestines, asthma or endometriosis.

In a further aspect of the invention there is provided a use of a nucleic acid molecule encoding angiomotin-like protein 1 or fragments or variants thereof as described in the first aspect of the invention in the detection and/or measurement of angiogenesis and/or tumour formation in a test sample.

In a further aspect of the invention there is provided a pharmaceutical composition for modulating angiogenesis and/or tumour formation, comprising an effective amount of a nucleic acid molecule encoding angiomotin-like protein 1 or fragments or variants thereof and a pharmaceutical exipient or diluent.

Preferably the pharmaceutical composition further comprises a nucleic acid molecule encoding angiomotin or fragments or variants thereof.

Conveniently the pharmaceutical composition prevents and/or reduces angiogenesis and/or tumour formation.

In a further aspect of the invention there is provided a vaccine for modulating angiogenesis and/or tumour formation, comprising an effective amount of a nucleic acid molecule encoding angiomotin-like protein 1 and an excipient or diluent.

Preferably the vaccine or pharmaceutical composition further comprises at least one additive for assisting or augmenting the action of the nucleic acid molecule encoding angiomotin-like protein 1 therein.

Preferably the at least one additive is an immunostimulatory molecule.

Conveniently the immunostimulatory molecule is a cytokine or polynucleotide (and/or antisense polynucleotide) encoding a cytokine.

Advantageously, the nucleic acid molecule encoding angiomotin-like protein 1 is administered to the eye, this may be intravitreally or in the form of an eye bath or eye drop.

In a further aspect of the invention there is provided a kit of parts comprising:

a pharmaceutical composition or vaccine as described in any of the aspects of the invention; apparatus for administering the pharmaceutical composition to the eye; and instructions for use.

Preferably the apparatus of (ii) is an intra-ocular needle, an eye bath or an eye dropper.

Preferred, non-iimiting examples which embody certain aspects of the invention will now be described, with reference to the following figures:

Figure 1. Angiomotin is essential for intersegmental vessel formation.

Depletion of Amot expression in developing zebrafish embryos using amot antisense morpholinos leads to vascular defects in the head and trunk regions. A- D, lateral views, dorsal is up, anterior to the left. The cranial arteries are normal in embryos injected with the mismatch control morpholino {A), whereas the pMBC and pHBC are dilated in the AmotKD embryos (B, arrowheads) at 36 hpf. The sprouting of the ISVs and the formation of the DLAV is complete at 36 hpf in the control (C), but the migration of the ECs in the ISVs are halted midway in the AmotKD (D, arrows) and no continuous DLAV is formed. Percent of defective embryos at 36 and 60 hpf using different morpholinos (E). Injection of amot antisense morpholinos leads to 72% defective embryos at 36 and 60 hpf, whereas the control mismatch embryos does not display any defects. Co-injection of human amot mRNA leads to rescue of the phenotypes.

Co-injection of murine amotl-1 mRNA leads to a delayed rescue at 60 hpf. The number of filopodia per cell was reduced fivefold in the AmotKD (F).

In panel E: ¹ AmotKD embryos display dilated pMBC, pHBC and arrested ISV migration at 36 hpf.

² At 60 hpf, Amot KD embryos display only the ISV defects. Abbreviations: DLAV, dorsal lateral anastomosing vessel; ISV, intersegmental vessel; pHBC, primordial hindbrain channel; pMBC, primordial midbrain channel. Scale bar 100 μm.

Figure 2. Vascular defects in the Angiomotin-deficient yolk sacs and embryos at E10.5.

lmmunofluorescent staining at E10.5 using Amot-specific antibodies in combination with the endothelial marker isolectin-B4. Amot is expressed by the capillaries of the brain (A) where the staining overlaps with isolectin, a marker for EC [A-C). More expression data is found in supplemental figures 1 and 2. Wt yolk sacs form a branched vascular network with different 36 sized vessels (D), whereas the Amot-deficient yolk sacs display a less organized vessel network (E). At a higher magnification, the difference is shown more clearly, with narrow blood vessels in the wt (F) and dilated blood vessels that have anastomosed into lagoon-like structures (asterisk) in the Amot- yolk sacs (G). Whole mount staining at E10.5 of wt (H) and Amot- (/) embryos using the PECAM antibody. The blood vessels in the brain of wt embryos have expanded into a vascular plexus with large vessels that branch into smaller capillaries in the developing brain and somites (J, arrows). The brains of the Amot- embryos show similar structures as in the yolk sacs with dilated vessels (arrows) and lagoon-like structures (K, asterisk). There is an accumulation of erythrocytes in the Amot- vessels (M), compared to the wt vessels (L). There is a lack of paraneural capillaries in the somitic region in the Amotembryos (O, arrowhead) that are present in the wt embryos (N, arrowheads). Scale bar in AC 14μm, in D, E, H, I 1 mm, in F, G 21 μm and in J-O 10 μm.

Figure 3. VEGF-induced sprouting is impaired in Angiomotin-deficient embryoid bodies.

Embryonic stem cells were allowed to differentiate and grow in 3D collagen I gels. ECs were visualized with the PECAM antibody (red) and supporting perivascular cells with αSMA (green). In the presence of VEGF the wt cells migrate and form sprouting tubes that invade the collagen gel, whereas the Amot- cells only a few sprouts are formed (A). Quantification of the tubes that invade the collagen I gel (β). Invading tubes were coated with perivascular cells in both the wt and Amot- bodies (C). αSMA-positive cells were not dependent on either VEGF or Amot for their migration (D). Scale bar in A, 500 μm, B, 50 μm and C, 250 μm.

Figure 4. siRNA knockdown of Angiomotin in the CNV-model impairs angiogenic sprouting in vivo.

Neovascularization was induced by laser-induced rupture of the inner most layer of the choroid. The lesions were treated with intra ocular injections of siRNA at 0, 3 and 6 days.

37 The sprouting ECs within the lesion were visualized using the PECAM antibody. Mice treated with the control siRNA (A) display extensive sprouting in the lesion, whereas the sprouting area of the lesion in mice treated with the Amot siRNA (B) is half the size (C). ^*** , PO.001.

Figure 5. Angiomotin is not required for tight junction formation, but for the organization of actin and focal adhesions.

Wt PmT-ECs express both isoforms of Amot, whereas no expression is detected in the Amot- PmT-ECs (A). RT-PCR analysis shows that both cell lines express equal amount of EC-specific markers (S). Amot expression overlaps with the tight junction marker ZO- 1 in cell-cell contacts in the wt PmT-EC (C upper panel). ZO- 1 still localizes to cell-cell junctions in the absence of Amot (C lower panel). Visualization of focal adhesions (arrowheads) and actin fibers using a paxillin antibody and phalloidin (D). The Amot- PmT-EC displays an irregular pattern of actin fibers and shorter and smaller focal adhesions (D lower panel). The highlighted areas show focal adhesions at a higher magnification. Box diagram showing the area (E) and length (F) of focal adhesions. Scale bar, 14 μm. * ^** , PO.001.

Figure 6. Angiomotin is critical for growth factor mediated migration.

Wt and Amot- PmTECs were allowed to migrate towards VEGF, bFGF or serum in the Boyden chamber assay (A). Amot- PmT-ECs fail to respond to growth factors, but show an equal migratory capacity as wt PmT-ECs in the presence of serum. In the in vitro wound healing assay wt PmT-ECs respond to VEGF with an increase in migration rate (B), whereas the Amot- PmT-ECs fail to respond (C). There is no difference in migratory response in the presence of serum (D). Both wt and Amot- PmT-ECs display a similar response in proliferation to VEGF (E,

FJ.siRNA knockdown of Amot in BCE cells was confirmed using Western blot (G). Amot siRNA38 transfected BCE cells show a decrease in basal migration in the Boyden chamber assay and do not respond to bFGF (H).

Figure 7. Angiomotin-deficiency results in an increase in lamellipodia formation, an increase in Rac activity and loss of polarization.

Photographs from time lapse studies at indicated time points show that the wt PmT-ECs form one large iameliipodia (asterisk) in the direction of migration (A, upper panel). In contrast the Amot- PmT-ECs fomn several cell protrusions, and lamellipodia (arrows, A lower panel). Subconfluent cells are stained with a Golgi apparatusspecific antibody and scored for degree of polarization (S). A majority of the wt cells are polarized whereas only about half of the Amot- cells display the same pattern (C). Rac1 activation is increased in the Amot- PmT-ECs (D). Knockdown of both isoforms of Amot in 293T cells also leads to an increase in the activation of Rac1 (E). Equal expression of Rich-1 in wt and Amot- PmT-ECs (6). In wt cells Rich-1 localizes to lamellipodia (G, arrow, upper panel), but its localization is altered in the Amot- PmT-ECs (G, lower panel). Lamellipodia are highlighted with white dotted lines in A. Scale bar 14 μm. ^*** , PO.001 ; ^** , P<0.01 ; ^* , P<0.05 39.

Figure 8. Whole mount in situ hybridization using Amot-specific probes (A, C, D).

Expression is detected in the intersomitic vessels (arrowheads) at E8.5 (D) and at E10.5 (A).

Amot expression is also observed in the branchial arches, region of the midbrain (C, arrow) and limb buds (A). Amot sense probe was used as a negative control (S). lmmunofluorescent staining using Amot-specific antibodies together with the endothelial marker isolectin-B4 (EV).

Amot is not expressed in the aorta expressed at E12.5 (E), but in the capillaries in the somitic region (H). AMOTL-1 , but not AMOTL-2 is also expressed in the capillaries (K, N).

Amot is present in the blood vessels of the placenta (Q, Sj in wt, but not in the Amot-deficient placentas (T, V). Abbreviations: ba, branchial arch; br, brain; h, heart; Ib, limb bud. Positive staining of erythrocytes is due to non-specific cross reactivity of the secondary antibody (asterisk). Scale bar in A-D, 1 mm and in E-V 14μm.

Figure 9. - Immunofluorescence stainings using an Amot-specific antibody

Immunofluorescence stainings using an Amot-specific antibody reveals that Amot is expressed in the epithelial cells in Rathke's pouch of the developing pituitary at

E10.5 (A) and at E12.5 (B). Other epithelial cells that express Amot are the cells that line the tongue (C) and the branchial arches (D) at E12.5. AMOTL-2 shows a different expression pattern, with positive staining in the neuroepithelium of the midbrain at E12.5 (£, F). Amot is expressed by the cytotrophoblasts in the placenta (G) AMOTL-1 (/, K), but not AMOTL-2 (J, L) are expressed by the cytotrophoblasts both in wt and Amot- placentas. Scale bar, 21 μm.

Figure 10.

Head and rump length of embryos from the same litter (n=6) at E10.5 were compared and the length of the wt embryos set to one. Amot- embryos were 75% the length of their wt littermates (A, B). Hematoxylin and eosin stainings of placental sections from wt and 40 Amot- at E10.5 show that the Amot- placentas appear to have a normal layered structure (C).

The weight of Amot- placentas is 14% less than wt placentas (D). Scale bar, 1mm. ^*** , PO.001 ; ^** , P<0.01.

Figure 11. Comparison of the vascular structure in wt at E9.5 (A) and Amot- at E10.5 (B).

PECAM whole mount staining of wt (C) and Amot- (D) embryos at E1 1. Higher magnification of the boxed areas in C and D shows the difference in vascular structure in wt (E) and Amot- (F) brain at E11. The somitic region (G-/) in wt at E9.5 (G), Amot- at E10.5 (H), and at E11 (/). Quantification of the vessel diameter

in the brain at E10 5 Wt embryo (K) and Amot- embryos (L-N) at day 11 Scale bars A, B, E-H, 10 μm, C, D, K-N 1 mm ^*** , P<0 001

Figure 12. Western blot analysis of Amot ^' embryoid bodies

Western blot analysis show that wt, but not Amot- embryoid bodies express both isoforms of Amot (A) RT-PCR analysis shows that there is no statistical difference in the amount of differentiated ECs between wt and Amot- (S) Proliferation of ECs in response to VEGF in the embryoid body assay was not affected in the Amot- bodies (C) Tubular structures in the wt embryoid body express both Amot and PECAM (D-F) Epithelial-like cells also express Amot in their cell-cell contacts (G) Western blot analysis of the knock down of both isoforms of Amot in mouse aortic endothelial (MAE) cells using siRNA (H)

Scale bar, 14 mm, ^* , P<0 05

Figure 13. Expression of EC markers in the PmT-ECs

FACS analysis of wt and Amot- PmT-ECs show that both cell lines express the EC markers PECAM and VEGFR2 (A)

Western blot of cell lysates confirm that the cells also express vWF, confirming that the cells were blood vessel ECs (S) Visualization of the focal adhesions (arrowheads) using the FAK 41 and vinculin antibodies (C) Amot- PmT-ECs display four times as many protrusions compared to wt PmT-ECs (D) Wt PmT- ECs migrate five times as fast compared to Amot- PmT-ECs Quantification of the distance (E) and cell tracks (F) ^** , P<0 01 42

Figure 14. Angiomotin is required for embryonic blood vessel formation and for endothelial polarization during migration

E wt embryo and F AMOT- embryo Note the smaller size of the embryo in F, G and H Whole mount PECAM staining visualizing the endothelial lining of blood vessels of the embryonic brain (E day 10 5) Note the lack of vascular structure of the blood vessels of the AMOT deficient embryo (H) I and J shows by

fluorescence the flow of erythrocytes of brain capillaries in wt and AMOT- embryos.

Figure 15. Angiomotin is required for embryonic blood vessel formation and for endothelial polarization during migration.

Time lapse analysis of wt and AMOT- endothelial migration. Not the polarized morphology of the wt cell (top panel) with the lamellipodia and front(star) and retracting tail in the end. In contrast, AMOT deficient cells extend lamellae in several different orientations simultaneously and therefore are defective in directional migration.

Figure 16. AMOT therapeutic antibodies inhibit tumor and retinal angiogenesis.

Top panel shows treatment of the mouse TUBO breast cancer cell-lines with either control or anti-AMOT antibodies. The TUBO tumors grow as ductal breast cancers and the ducts are indicated at higher maginification with an asterisk. Green color shows positive staining for the endothelial marker CD31. Note the absence of positive staining in the anti-Amot treated tumors. Bottom Panel. Visualization of the vascular front during retinal vessel formation in neonatal mice. The leading endothelial cells, "tip cells", extend filopodia which guide the establishment of the vascular network. Not the disorganization of the filopodia of vessels from retina treated with anti-AMOT antibodies.

Figure 17. AMOTL1 expression

AMOTL1 is expressed in angiogenic endothelial cells, colocalizes with AMOT and AM0TL1 and AMOT forms a protein complex. A Expression of AM0TL1 in endothelial cells of a human testis tumor. B. Co-localization of AMOT and AMOTL1 to tight junctions of endothelial cells in vitro. C. AMOT and AM0TL1 form a protein complex.

Figure 18. Synergistic effect of combining AMOT and AMOTL1 DNA vaccination.

Balbc mice were vaccinated twice with pcDNA3 plasmid constructs encoding AMOT, AM0TL1 or both in combination. Plasmids were introduced by

intramuscular injections followed in vivo electroporation (mice were under anaestesia). Mice were vaccinated at day-21 and -7 . At day 0 mice were challenged with the invasive mouse breast cancer cell-line TSA (which is AMOT and AMOTL 1 negative).

Figure 19. Angiomotin-like protein 1 is critical for growth factor mediated migration.

(A)siRNA knockdown of Amot in MAE cells was confirmed using Western blot.

(B) Mouse aortic endothelial cells (MAE) were allowed to migrate towards VEGF, bFGF or serum in the Boyden chamber assay. MAE cells transfected with AMOTL1 siRNA have inhibited basal migration in the Boyden chamber assay and do not exhibit a migratory response to VEGF, bFGF or serum.

Figure 20 - Amino acid sequence of angiomotin like protein 1 (AMOTL-1)

Figure 21 - Nucleotide sequence of angiomotin like protein 1 (AMOTL-1)

Figure 22 - Amino acid sequence of angiomotin p80 splice variant

Figure 23 - Nucleotide sequence of angiomotin p80 splice variant

Figure 24 - Amino acid sequence of angiomotin p130 splice variant

Figure 25 - Nucleotide sequence of angiomotin p130 splice variant

Example 1 - Angiomotin anaysis

Materials and methods

Tissue culture

AB-1 murine embryonic stem (ES) cells were grown on a layer of mitomycin-C treated murine embryonic fibroblast (MEF) feeder cells. ES cells were grown in ES-medium containing Dulbecco's modified Eagle's medium (DMEM, Sigma) supplemented with 10% KSR knock-out replacement serum (Gibco), 1 mM Na- pyruvate, 2 mM L-glutamine, 0.1 mMb-mercaptoethanol, 1 % penicillin- streptomycin, 1x non-essential amino acids (Sigma) and 50 units/mL of ESGRO (Chemicon International). Cells were grown at 37 ⁰ C with 6% CO2. MEFs were grown in ES-medium without ESGRO and treated with 10 μg/mL mitomycin C (Sigma) for two hours. To induce differentiation ESGRO was omitted from the medium and 1200 cells were allowed to aggregate on the lid of a non-adherent tissue culture dish. The medium was supplemented with VEGF where indicated.

Antibodies and reagents

Peptides corresponding to the 24 C-terminal amino acids were used to generate the C-terminal antibodies against Amot (Levchenko et al. 2003). Biotinylated isolectin-B4 (Sigma) was used to visualize the endothelium. Alexa Fluor 594- conjugated Phalloidin (Molecular probes) was used to visualize actin. The Golgi apparatus was visualized using an antibody against the Golgi-specific protein p1 15 (BD Pharmingen; a kind gift from E. Raschberger, LICR, Stockholm, Sweden). Other antibodies used were von Willenbrant factor (vWF, DAKO), PECAM (BD Pharmingen), FITC-mouse monoclonal anti-α smooth muscle actin antibody (R&D) paxillin (BD Transduction Laboratories), VEGFR2 (BD Transduction Laboratories) and β-actin (Sigma). The rabbit anti-Rich-1 antibody has been described before (Richnau and Aspenstrom 2001 ). Secondary antibodies: anti-rabbit-HRP (GE healthcare), Alexa Fluor antibodies (Molecular probes), anti-rabbit-FITC (DAKO). The nucleus was visualized using mounting medium containing DAPI (Vector Laboratories). VEGF was purchased from Peprotec Ltd.

Zebrafish analysis

Adult zebrafish were grown in the fish facility with a 14 hour light/12 hour dark cycle.

Zebrafish embryos harboring the Tg (fli:EGFP)y1 transgene (Lawson and Weinstein 2002) were injected at the 1-4 cell stage with morpholinos and raised at 28 ⁰ C in standard E3 water supplemented with 0.003% PTU (phenyls- thiourea). BLAST searches on Ensembl were done to identify an amot zebrafish ortholog. The following morpholinos were purchased from Gene Tools LLC: 5'- CCACTGACACAACTACCACCAAGTG-3' (Amot exon 2), and 5'- CCTTACTTGACCTATTGAGGAGCAG-3' (Amot exon 3) were used to inactivate the zebrafish amot. A mismatch morpholino 5'-

CCACTcACACAAgTACgACgAAcTG-3' (Amot exon 2 Ctrl) was used as a negative control. All morpholinos were injected at 250 μM diluted in 0.3X Danieau's solution with 0.05% phenol red included as a tracer. Human p80-Amot and mouse AMOTL1 mRNAs were synthesized from linearized templates in vitro using Ambion's Message Machine kits. For the rescue experiments, mRNAs were mixed with amot morpholinos and injected at approximately 100 pg/embryo. RT- PCR was used to confirm knockdown, using the primers 5' ACTCGATGGTCCCACACATT 3' and 5' GAATGACCCATTGGTGGAAG 3'. Phenotypic consequences of morpholino gene knock down were assessed from 30-60 hpf on a Leica MZ16 stereomicroscope equipped with epifluouresence. Specific vessels were identified using the interactive atlas of zebrafish vascular anatomy (Isogai et al. 2001 ).

In situ hybridization and PECAM whole mount staining

Non-radioactive whole mount hybridization was carried out as described (Bostrom et al. 1996). Two different probes were used to confirm staining. First, a 679 bp fragment from mouse Amot cDNA was amplified using PCR and primers sense

(5'-3'): GAGCTCCTCCGGGAGAAG, reverse (5'-3'):

TGGACCAGCCATTGGAGC. The purified fragment was cloned into the PCR- vector pTAdv (Clontech) and then subcloned into pSPT18 and 19 (Roche) to generate antisense and sense probes. The second probe (600 bp) was generated using a PCR-probe containing the T7 sequence (5'-3'):

TGATTAATACGATCACTATAGG at one end. The primers used were, sense (5 ¹ - 3'): TGGCTACTAGTGGAGTTAAAG and reverse (5'-3'):

CTGGCTGCTGCTGCGCTG. The specificity to Amot was confirmed by BLAST search. PECAM whole mount staining was preformed on embryos dissected between E9.5-11 as previously described (Levchenko et a/. 2003).

Immunofluorescence

Cells: cultured cells were plated in chamber slides and allowed to grow and adhere. The cells were fixed in 4% PFA for 10 min at room temperature and permeabilized in 0.1 % triton X- 100 (Sigma) for one min. Nonspecific reactivity was blocked by incubating with 5% horse serum in PBS for one hour before addition of primary antibody in blocking buffer for one hour. Antibody binding was detected with fluorescent-labeled secondary antibodies.

Sections: embryos were fixed in 4% paraformaldehyde (PFA) over night, dehydrated and embedded in paraffin. 5 μm sections were deparaffinized, rehydrated and boiled for 10 min in citric acid (1 ,8 mM sodium citrate; 8,2 mM citric acid; pH6,0) to unmask the antigen. After washing in PBS for 5 min, the anti- Amot antibody and isolectin-B4 were added and the sections were incubated over night at room temperature. Antibody binding was detected with fluorescent- labeled secondary antibodies and isolectin was detected using the fluorescent streptavidin kit (Vector laboratories Inc. CA).

Mouse breeding and genotyping

The Amot-deficient strain has been described (Shimono and Behringer 2003). Mice and embryos used in this study have been backcrossed up to six generations into the C57/B6 strain. Mouse tail tips, yolk sacs or embryos were used for genotyping by PCR as described.

Growth and immunofluorescence of embryoid bodies

Conducted as previously described by Jakobsson et al. 2006. Briefly, embryoid bodies were grown as hanging drops without leukemia inhibitory factor (LIF) and on day 4 the embryoid bodies were placed in collagen I suspension and

supplemented with VEGF where indicated. At day 10 the embryoid bodies were fixed in 4% paraformaldehyde, blocked and permeabelized, followed with sequential overnight incubations with primary and secondary antibodies.

Western blot

Cell lysates were analyzed by SDS-PAGE and proteins were transferred to nitrocellulose membrane. Non-specific binding was blocked for one hour in 10% dried milk in PBS containing 0.1 % tween (PBS-T). The filter was incubated over night at 4° C in 5% dried milk in PBS-T with primary antibody. The secondary antibody (anti-rabbit-HRP, anti-mouse-HRP or anti-rat-HRP) was diluted 1 :10000 (GE health care) and incubated for one hour at room temperature. Signal was detected using the luminol reagent from Santa Cruz.

Laser-induced CNV

CNV was generated by krypton laser-induced rupture of Bruch ^' s membrane, as previously described (Berglin et al. 2003). Briefly, three krypton laser photocoagulation burns (50 μm spot size, 0.1 sec duration, 120 mW power) were induced in each eye of C57BL/6J mice using a handheld contact lens (647 nm, Spectra-Physics 265 Exciter, Lasertek, Helsinki, Finland).

Mice received an intravitreal injection of 20 pmoles (1 μL) of Amot siRNA (smart pool from Dharmacon) at 0, 3 and 6 days after laser treatment. Ten mice were treated in each group in two independent experiments. Eyes were enucleated 10 days after krypton laser and fixed in 4% paraformaldehyde for 30 min, the cornea and lens were removed and the entire retina was carefully dissected from the eyecup.

The eyecups were rinsed in PBS, permeabilized in 0.5% Triton X-100 and blocked with 3% goat serum in PBS/Triton X-100. The eyecups were incubated with anti-PECAM antibodies overnight at 4°C, followed by an incubation with an Alexa Flour secondary antibody. Radial cuts were made from the edge of the eyecup to the equator and the eyecup was flattened and mounted with antifade medium (Vectashield Mounting Medium, Vector Laboratories, Burlingame, CA, USA) with the sclera facing down and the choroid facing up.

Flat mounts were examined with a fluorescence microscope (Axioskop 2, Carl Zeiss, Jena, Germany) where after images were captured with a digital camera (Carl Zeiss, Jena, Germany) and further analyzed using the AxioVision LE software (Carl Zeiss, Jena, Germany). The vascularization of each lesion was estimated from the PECAM staining by quantifying the number of PECAM- positive pixels per plaque, and data from each animal was treated as a single statistical point.

Immortalization of ECs derived from embryoid bodies

As described (Balconi et al. 2000). Briefly, eleven days old embryoid bodies were collected and disaggregated by 1 .5 mg/mL collagenase A (Sigma). After disaggregating, 25 μg/mL DNase was added. Cells were allowed to grow for two days before infection with PmT virus harvested from GgP+E cells (kindly provided by Dr. Elisabeta Dejana, Mario Negri Institute for Pharmacological Research, Milan, Italy). Wt ECs were selected with 800 μg/mL G418 (Sigma).

Real-time PCR

Total RNA was extracted from 8 days old embryoid bodies and PmT-ECs using the RNeasy mini kit (Qiagen). Contaminating genomic DNA was removed with DNase (Qiagen) and 3 μg of total RNA was used for first-strand cDNA synthesis using oligo dT primers and the Superscript III First-Strand kit (Invitrogen). The reaction was performed in duplicate with the 7500 real-time PCR system (Applied Biosystems). The expression levels of the EC-specific genes in Amot- embryoid bodies and PmT-ECs were normalized against actin and compared to the expression level of wt bodies and cells. A ratio of one corresponds to equal expression level of wt and Amot- cells. Primers used: $-actin sense: (5'-3') CACTATTGGCAACGAGCGG; antiseπse: (5'-3') TCCATACCCAAGAAGGC. PECAM sense: (5"-3') TACTGCAGGCATCGGCAAA; antisense: (5'-3') GCATTTCGCACACCTGGAT. VEGFR-1 sense: (5'-3')

GGGCAGACTCTTGTCCTCAACT; antisense: (5 -3 ¹ )

CAGCTCATTTGCACCCTCGT.

VEGFR-2 sense: (5'-3 ¹ J TACAGACCCGGCCAAACAA; antisense (5'-3') TTTCCCCCCTGGAAATCCT. VE-cadherin: sense: (5'-3')

AGGACAGCAACTTCACCCTCA 3', antisense (5'-3')

AACTGCCCATACTTGACCGTG.

NP-V. sense (5'-3") TCCGCAGCG ACAAATGTG; antisense (5'-3') GGTAACCGGGAGATGTGAGGT.

Focal adhesion measurements

For focal adhesion measurements, the cells were stained with the paxillin-specific antibody, photographed and analyzed using the software Lucia G. Threshold values were set between 0.2 and 20 μm to exclude background.

Cell migration assays

Boyden chamber migration assay was performed as described (Levchenko et al. 2003).

Briefly, 30 000 cells in medium containing 0.1 % serum were loaded in each well and allowed to migrate towards 50 ng/mL VEGF for five hours. Non-migrating cells were removed and remaining cells were fixed in ice-cold methanol and stained with Giemsa stain (VWR). The total number of migrated cells per field was counted at 2Ox magnification; each sample was tested in quadruplicates in three independent experiments. Wound healing was performed as described (Ernkvist et al. 2006). Briefly, cells were seeded onto a chamber slide and allowed to reach confluence. Subsequently a wound was made using a 10 μL tip and photos were taken at indicated time points.

Proliferation assay

50 000 cells were seeded onto 25 mm plates and starved over night in 0.5% FCS. 50 ng/mL VEGF was added and cells were counted every day. The assay was done in triplicates in three independent experiments.

Time-lapse

Cells were seeded on 3-cm cell plates and images were collected every 6 min. The images were used to create a movie using the Ulead GIF animator 5, where every second corresponds to 10 frames.

Polarization analysis

Ceiis were plated subconfluently on chamber slides and allowed to attach and spread for five hours. Thereafter, the cells were fixed and stained for the Golgi apparatus, actin fibers, and the nucleus as described above. If the Golgi apparatus was located within a 120° sector, the cell was defined as polarized, and if it was located within more than a 120° sector, it was defined as unpolarized.

Rac pull down assay

Cells were serum-starved overnight and then stimulated by growth medium containing 10% FBS for 30 minutes. After the treatment, cells were harvested, washed with PBS, and lysed in 1x Mg2+ Lysis Buffer (Fig. 125mM HEPES, pH7.5, 75OmM NaCI, 5% lgepal CA-630, 5OmM MgCI2, 5mM EDTA and 10% glycerol). Clarified lysates were incubated with PAK-1 PBDagarose (Upstate) for 90 minutes at 4 ⁰ C, washed for three times with 1x Mg2+ Lysis Buffer.

Rac1 was detected by immunoblot analysis.

Amot knockdown in cell lines

For silencing endogenous Amot in 293T cells, a 19-nt sequence corresponding to bases 2806- 2824 of homo sapiens amot mRNA (NM_133265.2) was selected as the target. A small hairpin RNA (shRNA) incorporating this sequence was cloned into pLenti-lox3.7 (pLL3.7) as previously described (Rubinson et al. 2003). The resulting vector was transfected together with packaging vectors VSVG and D8.9 into 293T cells and viral supernatant was collected after 48 and 72 hrs. This supernatant was then used to infect fresh 293T cells and GFPpositive infected cells were selected by flow cytometry. For silencing Amot in BCE cells a smart

pool from Dharmacon was used and transfected according to the manufacturer's instruction. Western blot analysis was used to confirm knockdown of both isoforms of Amot.

Statistical analyses

The statistical significance was determined using the Student t-test, where ^*** , PO.001 , ^** , PO.01 and ^* , PO.05.

Results

Vascular defects in Angiomotin-knockdown zebrafish embryos

The hallmarks of vasculogenesis and angiogenesis in mammals are recapitulated in zebrafish and the zebrafish embryo has emerged as a useful model to study the vertebrate cardiovascular development and physiology (Weinstein 2002; Goishi and Klagsbrun 2004).

Thus, to elucidate the role of Amot during embryonic angiogenesis we carried out morpholino-mediated knockdown of amot in zebrafish (AmotKD). The zebrafish amot ortholog was identified by a BLAST search on Ensembl and two antisense morpholinos were synthesized that were designed to block amot mRNA splicing at exon 2 and 3 respectively.

The efficacy of the knockdown was confirmed by RT-PCR (data not shown) and identical results were obtained using both antisense morpholinos.

To specifically visualize the developing blood vessels, we injected the morpholinos into transgenic zebrafish embryos expressing the enhanced green fluorescent protein (EGFP) under the endothelial promoter fli-1, Tg (fli1 :EGFP)y1 (Lawson and Weinstein 2002; lsogai et al. 2003). Macroscopic inspection of AmotKD embryos revealed overall normal development up to 60 hours post fertilization (hpf). More detailed analysis using fluorescence microscopy at 36 hpf showed that in approximately 72% of the embryos injected with amof antisense morpholinos the primordial midbrain and hindbrain channels (pMBC and pHBC, respectively; (lsogai et al. 2001 ) were dilated, whereas the other cranial vessels

appeared normal (Fig. 1A, B, E). This defect was transient and by 60 hpf both vessels in the AmotKD embryos were indistinguishable from embryos injected with the mismatch control morpholino (Fig. 1 E).

AmotKD embryos showed severe defects in the formation of trunk vessels.

During zebrafish development, the intersegmental vessels (ISVs) originate bilaterally from the dorsal aorta (DA) and the posterior cardinal vein (PCV) beginning around 20 hpf, and are thought to form by a process of angiogenesis (Childs et al. 2002; lsogai et al. 2003; Siekmann and Lawson 2007). ISV formation is marked by a stereotyped pattern of sprouting and dorsal migration of individual ECs at each somite boundary. ISV ECs show extensive filopodial extension and retraction as they follow a path initially along the somite boundary and later between notochord and somitic tissue (Leslie et al. 2007). In the control embryos, each ISV forming cell reached the dorsolateral roof of the neural tube and anastomosed with neighboring ISVs to form the dorsal longitudinal anastomotic vessel (DLAV) (Fig. 1 C). In contrast, the sprouting of the ISVs in the AmotKD embryos halted midway resulting in truncated ISVs that failed to form the DLAV (Fig. 1 D). Truncated ISVs displayed a 'hammerhead' appearance and showed a fivefold reduction in number of filopodia per cell (Fig. 1 F). The ISV defect remained at 60 hpf and identical phenotypes were observed using both amot morpholinos.

Thus, morpholino-mediated knockdown in zebrafish revealed that Amot function is required for correct vessel formation and EC migration.

Delayed rescue of the AngiomotinKD phenotype by AMOTL-1

To demonstrate the specificity of the observed defect, we co-injected 100 pg of human amot mRNA with the antisense morpholino, leading to a transient expression of hAmot. This resulted in an almost complete rescue demonstrating that the vascular defects observed in AmotKD embryos are due to loss of amot function (Fig. 1 E).

Amot binds to and co-localizes with the related AMOTL-1 (unpublished observation and S Fig. 1 ) and to assess overlapping functions and potential

rescue, murine amotl-1 mRNA was co-injected with the antisense morpholinos and phenotypes of the embryos were analyzed at 36 and 60 hpf. At 36 hpf, we observed no rescue of the defects, but at 60 hpf there was a partial rescue where only 26% of the amotl-1 rescued embryos displayed the ISV phenotype (Fig. 1 E).

This result suggests functional redundancy between the two family members, but also implies a qualitative difference between Amot and AMOTL-1 function.

Angiomotin is expressed in embryonic blood vessels

To extend these findings, we further investigated the function of Amot in mice.

First we analyzed the amot expression pattern during embryogenesis. Whole mount in situ hybridization analysis (Fig. 8) using amof-specific probes showed that amof is expressed in the intersomitic vessels as early as E8.5 (Fig. 8). Positive signal was also observed in part of the midbrain, the epithelium of the branchial arches, and the limb buds (Fig. 8). No expression of amot was detected in the aorta after E8.5 or in the heart.

lmmunofluorescent staining using anti-Amot specific antibodies confirmed protein expression in blood vessels in the somitic region (Fig. 8) and the epithelium of the branchial arches (Fig. 9). We also detected expression of Amot in capillaries in the brain and neural tube (Fig. 9).

Furthermore, Amot was not detected in the larger vessels, such as the cardinal vein, aorta or the heart (Fig. 8, and data not shown). Positive staining for Amot was found in the fetal blood vessels (Fig. 8) and giant cytotrophoblasts of the placenta (Fig. 9). Amot is also expressed in the epithelial cells forming Rathke's pouch (Fig. 9), which will later form the anterior pituitary.

Angiomotin-deficient mice die in utero

We have previously shown that the disruption of the amot gene in mice (Amot-) results in complete elimination of Amot expression, but also that the penetrance of the amot mutant phenotype is sensitive to genetic background (Shimono and Behringer 2003). Embryos from crosses of mice with a 129/SvEv-background show to morphological defects and 70% of the Amot-defϊcient embryos die

around E7.5. Backcrossing once into the C57/B6 (Fig. 129B6 mixed) background resulted in approximately 50% dead embryos at the same time point (E7.5) (Shimono and Behringer 2003). In this study, we have backcrossed the mice further into the C57/B6 background (up to six backcrosses) and in contrast to previous findings, we found that Amot-deficient embryos survived at a near Mendelian ratio after E7.5 (E8.5-10.5; Table 1 ).

Table 1

Offspring from +/-x -/O ¹ and +/- x +/0 matings genotyped between E8.5 and E10.5

¹ amot is located on the X-chromosome. Thus, males are either wt (+/0) or deficient for amot (-/0).

Table 2

Offspring from +/- x +/0 matings genotyped at weening ¹ .

After 4-6 backcrosses into C57/B6.

Table 3

Offspring from +/- x +/0 matings genotyped between E11.5 and E17.5

Table 4

Offspring from +/- x -/0 matings at E11.

Nevertheless, the number of live Amot-deficient pups was only approximately 25% of the expected ratio, indicating that a majority die in utero (Table 2). To determine at what stage the Amot- embryos die we established timed matings and analyzed embryos between E11.5-17.5. We detected similar rates of survival in these embryos as we did in the Amotpups (Table 3). At E11 , Amot- embryos survived at a close to a Mendelian ratio (Table 4 and 1 1 ). This argues that the embryos survive past E7.5 in a C57/B6- background, but that a majority of the Amot- embryos die between E1 1 and E11.5 (Table 5).

Table 5

Summary of the survival of Amot embryos and mice at different time points

Angiomotin-negative yolk sacs display vascular defects

A primitive vascular network in the yolk sac is first established through vasculogenesis, followed by the development of a more mature network through angiogenesis around E8-8.5 in mice. Angiogenesis in the embryo proper takes place around the same time, but blood cells are not present in the vessels until E8.5 when the embryonic and yolk sac vascular systems amalgamate and the nucleated red blood cells formed in the yolk sac enter the embryo. We analyzed the vascular network of the yolk sacs by immunofluorescent staining of the ECs using an antibody against the endothelial marker PECAM. The yolk sacs from wild type (wt) or heterozygote embryos at E10.5 exhibited branched structures and a highly organized vascular bed (Fig. 2D, F), whereas Amot- yolk sacs displayed an abnormal vascular patterning (Fig. 2E, G). The Amot- yolk sacs contained vessels with increased diameter and some vessels anastomosed to form lagoon-like structures (Fig. 2G). Erythropoiesis is not dependent on Amot since we could observe nucleated blood cells in the Amot- yolk sacs. A functional blood flow between the yolk sac and the embryo proper was also established as we detected red blood cells in the embryo before embryonic erythropoiesis at E11.

Angiomotin is necessary for normal blood vessel formation

We found that at E10.5 the Amot- embryos and placentas were 85 and 75% the size of their littermates respectively (Fig. 10, B, D), indicating that the growth of both embryonic and extraembryonic tissues was affected by the Amot deficiency. Immunofluorescence analysis showed that Amot was expressed in both blood vessels and cytotrophoblasts of the placenta (Fig. 8Q and 9G). We also

examined the possibility that the observed growth retardation was due to placental deficiency. Histological analysis of the wt and Amot- placentas did not reveal any obvious defects. Thus we concluded that the establishment of the placental layers and function was normal in the Amot-deficient placentas (Fig. 10C).

At E 10.5 the wt embryo has developed an intricate system with vessels supporting the developing brain and somites. Since Amot is expressed in the capillaries at this time point we decided to focus our analysis on the cephalic and somitic regions. The blood vessels of the embryos were visualized by whole mount staining of the ECs using the PECAM antibody. The vessel network in the wt brain was well developed and consisted of larger vessels that branched into smaller capillaries (Fig. 2H, J, L). In contrast, the Amotdeficient embryos displayed a less structured network that formed lagoon-like structures with an accumulation of erythrocytes, comparable to the ones detected in the yolk sac (Fig. 21, K, M). The branching was not affected in the Amot- embryos, but the vessel diameter was twice the size compared to wt embryos (Fig. 11 and data not shown). The intersomitic blood vessels in the wt embryos formed an organized network, whereas the interconnecting paraneural blood vessels were missing in the Amot- embryos and the remaining vessels appeared dilated and shorter (Fig. 2N, O). Similar differences in vessel morphology in the brain and the somitic region were also observed at E11 (Fig. 1 1 C-F, I).

Amot was also expressed by a subset of epithelial cells, but we did not observe any obvious defects in the giant cytotrophoblasts of the placenta (Fig. 9K), or in the patterning of the branchial arches or limb buds (Fig. 2I), suggesting that Amot is not essential for establishing these epithelial structures.

Taken into account the growth retardation observed in the Amot- embryos, we also compared Amot- embryos at E10.5 with the vascular network in the brain and the somitic region of wt stage matched embryos at E9.5. At E9.5 the wt embryos displayed a less developed, but nonetheless structured vascular network in the brain, that did not resemble the structures found in the E10.5

Amot- embryos (Fig. 1 1 A, B). The intersomitic vessels were more dilated in the Amot- embryos at E10.5 compared to wt embryos at E9.5 (Fig. 11 G, H).

We did not observe any apparent increase of EC apoptosis in Amot- yolk sacs or embryos by PECAM and DAPI stainings. Thus, the observed defects show that Amot is dispensable for vasculogenesis, but plays an important role in angiogenesis.

VEGF-induced tubulogenesis is impaired in embryoid bodies lacking Angiomotin

To investigate the role of Amot in tubulogenesis, we assessed the tube-forming potential of wt and Amot- ES cells. To this end we used the embryoid body assay, where ES cells are allowed to differentiate and form a primitive vascular plexus (Vittet et al. 1996). This assay mimics vasculogenesis and angiogenesis, and is widely used to investigate the different steps involved. Amot is located on the X- chromosome in mice, and the ES cells used were male, thus providing us with an Amot-deficient cell after only one recombination event. Wild-type embryoid bodies expressed both isoforms of Amot, both by ECs and by epithelial-like cells (Fig. 12A, D-G), whereas the Amot- embryoid bodies lacked Amot (Fig. 12A).

We used a model for invasive angiogenesis to analyze the developing vascular structures in the wt and Amot- embryoid bodies. The embryoid bodies were cultured in threedimensional collagen I gels and stimulated with VEGF to induce outgrowth of tubes. In agreement with our findings from the Amot- embryos, we observed no defect in the ability of the Amot- ES cells to differentiate into ECs (Fig. 3, see also fig 14 and 15).

In the presence of VEGF the wt ECs invaded and extended tubes covered with perivascular smooth muscle cells into the collagen matrix (as previously described, (Jakobsson et al. 2006) (Fig. 3). In contrast, the Amot- ECs did not respond to VEGF, and only one fifth of the amount of PECAM-positive tubes was observed invading the matrix (Fig.

3). Smooth muscle cells invaded the matrix in the absence of VEGF and were not affected in the Amot- embryoid bodies (Fig. 3C, D), showing the EC-specific effect of Amot-deficiency.

To exclude the possibility that the observed difference was due to a deficit of ECs in the Amot- embryoid bodies, we measured the expression of five different

endothelial markers (PECAM, VEGFR1 , VEGFR2, VE-cadherin and NP1 ) using real-time PCR. There was no significant difference in the expression level of the EC-specific markers between wt and Amot- embryoid bodies (Fig. 12B). Neither was the proliferative response of the ECs towards VEGF affected (Fig. 12C).

The difference between the relatively mild in vivo defects and the severe phenotype observed in the embryoid bodies with major defects in migration and tube-forming potential of the ECs, suggests that there is a developmental pressure compensating the lack of Amot with other genes leading to a less severe phenotype in vivo. To test the influence of Amot on the angiogenic process in vivo without the possible developmental upregulation of compensatory signaling pathways, we used the choroidal neovascularization (CNV) model where choroidal vessels in the eye are stimulated to grow into the subretinal space. This process is driven by VEGF and shares similar features with the angiogenic sprouting observed in the embryoid body system (Kvanta 2006). The CNVs were generated by krypton laserinduced rupture of the Bruch's membrane (the outer most layer of the retina), as described in material and methods. Control and Amot siRNA were injected intra ocularly at 0, 3 and 6 days post treatment. We verified the efficacy of Amot siRNA knockdown in mouse ECs (Fig. 12H). The angiogenic response was calculated using the PECAM area of the plaque. As shown in figure 4, the angiogenic response in the Amot siRNA-treated CNVs were half of the control siRNA-treated CNVs. Taken together, the embryoid body and CNV results indicate that there is a pressure to control the expression of other angiogenic-related genes during mouse development.

We investigated the expression patterns of AMOTL- 1 and -2 during mouse development, and found that AMOTL-1 , but not AMOTL-2 overlaps with that of Amot (Fig. 81 H-P and 9 E-L). Western blot analysis showed that AMOTL-1 expression is not upregulated in embryos that pass the restriction point at E1 1 (data not shown). However, since co-injection of amotl-1 mRNA with the antisense amot morpholinos partially rescue the phenotype in AmotKD zebrafish embryos (Fig. 1 E); it is conceivable that signaling from AMOTL-1 might rescue the effect of Amot-deficiency in the mouse.

These studies demonstrate that Amot does not influence the differentiation or proliferation of ECs, but is important for proper migrational and tube forming response to growth factors, such as VEGF.

Angiomotin is important for proper focal adhesion formation

To investigate the cellular mechanism behind the observed defects, we immortalized ECs from the embryoid bodies by polyoma middle T (PmT) virus. PmT selectively transforms embryonic ECs in a single step (Williams et al. 1988; Kiefer et al. 1994). Amot expression in the immortalized ECs (PmT-ECs) was analyzed by Western blot. As shown in figure 5A the wt PmT-ECs expressed both the p80 and p130 isoform of Amot, whereas the Amot- PmT-ECs did not. We verified by real-time PCR, FACS and immunoblotting, that the isolated cells were ECs and expressed equal amount of the endothelial-specific markers (Fig. 5B and S Fig. 6A, B).

We and others have shown that Amot localizes to tight junctions, however it is not essential for their formation (Bratt ef al. 2005; Wells et al. 2006). As expected, both the wt and Amot- PmT-ECs formed tight junctions as indicated by the tight junction specific marker ZO-1 , although only the wt PmT-ECs express Amot (Fig. 5C).

To analyze whether cells lacking Amot differ in cytoskeletal organization compared to the wt cells, we stained subconfluent cells for actin and the focal adhesion marker paxillin. We found a significant difference between wt and Amot- PmT-ECs with respect to actin fiber organization and paxillin staining. Wild-type cells displayed a regular pattern of actin fibers with paxillin staining at the actin fiber ends (Fig. 5D). The Amot- cells on the other hand revealed a disorganized pattern of actin fiber organization and smaller and shorter focal adhesions, which were not ordered along the actin fibers (Fig. 5D). To quantify the difference, we measured the area and the length of the focal adhesions. As shown in figure 5E and F, Amot-deficient cells contained about 20% shorter and smaller focal adhesions compared to wt cells. Similar staining patterns were also obtained using the focal adhesion kinase (FAK) and vinculin specific antibodies (Fig. 13C).

Actin and focal adhesions are important components in controlling migration and loss of Amot is thus likely to affect EC migration.

Angiomotin is involved in EC migration

We evaluated the migratory capacity of the PmT-ECs using two independent migration assays, the Boyden chamber assay in which single cells are allowed to migrate towards a chemoattractant and the in vitro wound healing assay in which a wound is made to a confluent cell layer and cells are subsequently allowed to migrate to fill the wound. The latter assay measures spontaneous migration of a sheet of cells and the two assays are known to evoke different responses in the Rac/Rho pathway (Aepfelbacher et al. 1997; Nobes and Hall 1999).

The Amot- PmT-ECs displayed a threefold reduction in the basal migrating capacity compared to the wt PmT-ECs in the Boyden chamber assay (Fig. 6A). Consistent with the results from the embryoid body assay, the Amot- PmT-ECs did not exhibit a chemotactic response to VEGF, or to basic fibroblast growth factor (bFGF) (Fig. 6A).

Furthermore, the Amot- PmT-ECs showed a similar lack of response to VEGF in the wound healing assay (Fig. 6C). However, there was no difference in the ability of the two cell lines to migrate towards serum, in which lysophosphatidic acid is a major chemoattractant (Fig. 6A, D). To verify our data in another cell system, we performed knockdown of Amot in bovine capillary endothelial (BCE) cells using amor specific siRNA. Both isoforms of Amot were efficiently knocked down (Fig. 6G) and the result was similar to what we observed in the PmT-ECs, with a decrease in both basal and bFGF-induced migration rate in the Amot siRNA transfected cells (Fig. 6H).

Our data demonstrate the importance of Amot for proper migratory response towards growth factors during both single cell migration and sheet-like movement. The observed defect was not due to a general defect in the VEGFR2-signaling, since the wt and Amot- cells displayed a similar proliferative response upon VEGF-treatment (Fig. 6E, F) and expressed equal levels of VEGFR1 , VEGFR2 and NP-1 (Fig. 5B).

We studied the dynamics of sheet-like or single cell migration by time-lapse photography. During single cell migration, the wt PmT-ECs developed one large lamellipodia in the direction of migration (Fig. 7A ). In contrast the Amot- PmT- ECs exhibited an unpolarized phenotype, with several peripheral lamellipodia and cell protrusions that expanded and contracted rapidly (Fig. 7A ). The number of cell protrusions was increased fourfold in the Amot- PmT-ECs compared to the wt cells (Fig. 13D), and moreover, the expansion and contraction of the protrusions was faster in the Amot- cells (Fig.7A). Still, the Amot- PmT-ECs migrated inefficiently in the Boyden chamber assay (Fig. 6C) and the migration distance of these cells was only 20% of wt (Fig. 13E, F).

This indicates that the Amot- PmT ECs are defective in the response towards growth factors, such as VEGF and bFGF, but not towards chemoattractants present in serum.

In addition, the Amot- cells display a defect in organizing lamellipodia formation, leading to excess cell protrusions.

Figure 19 also shows the results of studies of endothelial migration using the Boyden Chamber assay (see review by Chen for protocols - Chen (2005) Methods MoI Biol. 294 pp15-22).

We evaluated the involvement of AMOTL1 in endothelial migration using the Boyden chamber assay in which single cells are allowed to migrate towards a chemo-attractant. AMOTL1 was knocked down in Mouse aortic endothelial (MAE) cells using a pool of four AM OTL 1 -specific siRNA molecules. Control cells were transfected with a pool of non-specific siRNAs. Efficient knock-down (over 80% less than control) was analyzed by western blot 48 hours after transfection (Figure 19 A). Lowering AM0TI1 levels in MAE cells resulted in reduction in the basal migration and completely inhibited VEGF and bFGF induced migration. It was not possible to overcome the anti-migratory effect of AMOTL1 by adding 2% fetal calf serum (Figure 19B).

Angiomotin-deficient cells show a defect in polarization

Cell polarization and the establishment of functionally specialized domains have been shown to play a pivotal role in the directional movement of cells in a chemotactic gradient (Franca- Koh and Devreotes 2004). The observation that Amot- PmT-ECs displayed a defect in migration, led us to investigate the ability of these cells to polarize. In a migrating cell, the Golgi apparatus is oriented in the direction of migration, where it is responsible for the polarized delivery of vesicles and their subsequent insertion at the plasma membrane (Singer and Kupfer 1986). The Golgi apparatus is thus used as an indicator for the degree of polarization of a cell, where a cell is defined as polarized if the Golgi apparatus is located within one 120° sector relative to the nucleus. We stained subconfluent cells with a Golgispecific antibody, divided each cell in three sectors, and measured the extent of polarization.

Staining of wt PmT-ECs revealed that a majority of the cells had their Golgi apparatus located within one sector, whereas only about half of the Amot- PmT- ECs showed the same pattern (Fig. 7B, C).

This indicates that the Amot- cells have a defect in polarization, which might be one of the causes for the observed migratory-deficiency in these cells.

Rad-activity is increased in Angiomotin-deficient and Angiomotin-knockdown cells

Cdc42 is important in the maintenance of cell polarity (Nobes and Hall 1995) and since the Amot- PmT-ECs were less polarized compared to wt cells, we investigated the levels of activated Cdc42.

The superfluous protrusions and lamellipodia formed by the Amot- PmT-ECs led us to study the levels of activated Rac1 , an important modulator of lamellipodia formation. Experiments revealed an increase of activated Rad in the Amot- PmT-ECs compared to wt PmT-ECs (Fig. 7D), consistent with the observed increase in protrusions and lamellipodia (Fig. 7A ). Rac activity is controlled by guanine nucleotide exchange factors (GEFs) and GTPase-activating proteins (GAPs). One such GAP is Rich-1 , shown to be a GAP for both Rad (Richnau

and Aspenstrom 2001 ; Wells ef al. 2006). Recent findings have shown that Amot acts as a scaffolding protein that associates with Rich-1 (Wells et al. 2006). Thus, we were interested in determining if the expression of Rich-1 was altered in the Amot- cells. Western blot using anti-Rich-1 antibodies showed that there are equal expression levels of Rich-1 in both cell lines (Fig. 7F). The localization of Rich-1 in subconfluent cells revealed that in addition to a cytoplasmic distribution, Rich-1 was also expressed at the edge of lamellipodia in wt cells, whereas that pattern was lacking in the Amot- cells (Fig. 7G). This might indicate that there is a local dysregulation of Rac1 activity at the edge, leading to increased lamellipodia formation and an upregulation of GTP-Rac1.

As a step in immortalizing ECs PmT recruits and activates Src family tyrosine kinases, which stimulates signaling to both the MAP kinase and PI3K pathways (Ong et al. 2001 ). To verify that the observed difference in Rac1 activation between the wt and Amot- PmT-ECs was not caused by a synergistic effect from the PmT virus and the lack of Amot, we depleted Amot in 293T cells using siRNA. These cells are of epithelial origin and express both isoforms of Amot endogenously. The Amot knockdown cells showed an increase in Rac1 activity consistent with the Amot- PmT-EC data (Fig. 7E), indicating that the result from the immortalized ECs was not an effect of PmT-induced subversion of Rac1 activity. In addition, the experiment shows that the Amot control of Rac1 activity is not EC specific.

Discussion

By silencing Amot expression in zebrafish, mouse and ECs, we have and provided evidence for an important role of Amot in vascular reorganization and EC migration. Amot was primarily expressed by ECs of intersomitic vessels as well as brain capillaries. No detectable mRNA or protein expression could be detected in larger vessels such aε the aorta or the cardinal vein. Amot was also detected in epithelial structures of the branchial arches and Rathke's pouch. We have previously shown that Amot is upregulated in tumor endothelium of both human and mouse tumors but not detectable in vessels of normal tissues (Troyanovsky et al. 2001; Holmgren et al. 2006). Furthermore, Amot is expressed in ECs during normal physiological angiogenesis of the mouse retina (Bratt ef a/.

2005). Taken together these findings indicate that Amot is primarily expressed in angiogenic tissues.

Angiomotin in angiogenesis

We have previously shown that ectopic expression of Amot promotes migration of ECs.

Furthermore, angiostatin (that binds Amot) or Amot antibodies inhibit EC migration in the Boyden chamber assay (Troyanovsky et al. 2001 ; Bratt et al. 2005; Holmgren et al. 2006; Levchenko et al. 2007). Our in vivo data support the notion that Amot plays a positive role in neovascularization and more specifically EC migration. Knockdown of Amot protein expression in zebrafish resulted in a distinct migratory defect of the ISVs. Capillary sprouting from the DA or PCV was not affected but the vessels failed to migrate and form functional DLAVs. Interestingly, AmotKD phenocopies the loss of VEGFR2 kinase activity in zebrafish (Covassin et al. 2006). The partial formation of arteries by the VEGFR mutant was most likely due to cooperation between the multiple VEGFRs present in zebrafish. The closely related AMOTL-1 protein directly associates to Amot and overlaps with Amot in the expression pattern. This is supported by the finding that injection of murine amotl-1 mRNA can partially rescue (although with a delay as compared to amot mRNA) the ISV effect observed with antisense amot morpholinos. In addition to the ISV phenotype, we also observed that the pHBC and pMBC vessels were dilated at 36hpf . This was reversible as the dilation was no longer detected at 60 hpf.

Genetic ablation of amot in mouse embryogenesis yielded a phenotype quite similar to that of AmotKD in zebrafish. Deletion of amot resulted in approximately 75% lethality occurring between E11 and E11.5. Affected embryos suffered from insufficient capillary formation in the somitic region and defects in vascular integrity (dilation) in brain capillaries. A defect in cell migration and tube formation was suggested by the findings that Amot-deficient stem cells showed a severe impairment in the formation of vascular sprouts in the collagen assay. Furthermore, intraocular injection of amot siRNA in wt mice inhibited choroid vessel infiltration into the retina in the laser-induced CNV model. No obvious defects were detected in the surviving Amot-deficient mice which were fertile and

of similar size to that of wt littermates. However, defects in pathological angiogenesis cannot be excluded.

Angiomotin links polarization and migration

The effect of Amot deficiency was also analyzed at a cellular level using ES- derived ECs or siRNA-mediated knockdown of Amot in BCE cells. Amot deficiency did not alter the VEGFstimulated differentiation of EC in vitro. Furthermore, Amot does not appear to affect VEGFstimulated cell proliferation or affect cell survival. However, loss of Amot renders the cells unresponsive to VEGF or bFGF-stimulated chemotaxis. An interesting aspect of this observation would be that Amot is required for the translation of a growth factor mediated signaling into cell movement. Efficient chemotaxis results from coordinated chemo-attractant gradient sensing, cell polarization and cellular motility. There is increasing evidence that polarity proteins such as Par-6, PKC, Patj are not only involved the establishment of basolateral polarity but also important for front rear polarity during migration (Shin et al. 2007).

Time-lapse analysis showed that wt cells polarize with a leading edge whereas Amot-deficient cells extended protrusions in an un-polarized manner. Consistent with these findings, we could show that Amot- PmT-ECs exhibited loss of polarization of the Golgi apparatus. Recent results indicate that Amot functions as a scaffold for polarity proteins as well as the Rac1/Cdc42 GAP, Rich-1 in epithelial cells (Wells et al. 2006). Rich-1 is also expressed in ECs where it localizes to the leading edge of migrating cells. In Amot- PmT-ECs, Rich-1 failed to localize to lamellipodia whereas ectopic expression of Amot promotes localization (Fig. 7). The subcellular localization of Amot to lamellipodia and the ability to bind cell polarity proteins as well as GTPase regulating proteins indicate a function in the local orchestration of GTPase activity required for cytoskeletal reorganization during migration.

In summary, we have shown that Amot and Amotl plays a specific role in regulating vascular formation through endothelial migration and cell polarity. Finally, these findings suggest that these are attractive molecules to specifically target endothelial migration without affecting differentiation and survival pathways.

Example 2

Proof-of-concept: Antibodies targeting angiomotin inhibit angiogenesis in vivo.

Material and methods

Generation of human recombinant antibodies to Amot

Single chain human antibody fragments with specificity for human Amot were selected from the single-chain fragment-variable (scFv) n-CoDeR phage display library, essentially as described earlier (Soderlind, et al., 2000).

In brief, human p80 Amot was expressed in eukaryotic cells and binding assays were performed with either intact cells, cell lysate or in a purified form, in 3 consecutive rounds of selection. Selected scFv were screened for specific Angiomotin binding in an automated system with an ELISA format with luminescence as the readout.

The scFv identified as being specific for Angiomotin were also screened for cross reactivity to murine Amot with an ELISA format with luminescence as the readout.

The recombinant scFv contained a C-terminal His tag that allowed for IMAC purification using Ni-NTA sepharose (Qiagen). The purity of the preparations exceeded 95 %, as determined by SDS-PAGE. Fab fragments and full length

IgG antibodies were produced by cloning into modified pcDNA3 vectors followed by transient transfections into HEK293 cells with Lipofectamin (Invitrogen).

Fab fragments and full length IgG antibodies were purified on a MabSelect protein A column (Amersham Biosciences). The purity of the preparations exceeded 98 %, as determined from SDS-PAGE. The binding specificity of the scFv was tested using luminescence-based ELISA where dilutions of the scFv antibody fragments were incubated in test plate wells coated with purified Angiomotin.

The Fab fragments utilised for PEGylation contained a C-terminal cystein allowing for a single site specific PEGylation using a 20 kD PEG-maleimid compound

(Nectar Therapeutics). 2-Mercaptoethylamine HCL (Fluka), 5 mM, pH 7.0 at 37 ⁰ C for 90 min, was used for selective reduction of the C-terminal cysteins of the Fab fragments and the molar ratio (PEG:Fab) in the PEGylation reaction was 5:1. The PEGylation reaction was conducted at 25 ⁰ C, pH 7 under nitrogen for 2 h. MabSelect protein A column chromatograpy was used to purify the Fab fragments from unreacted PEG-maleimid. PEGylated Fab could be separated from non-PEGylated Fab using size exclusion chromatography (GE Healthcare, Sweden).

Cell culture

Spontaneously immortalised mouse aortic endothelial (MAE) cells (Bastaki, et al., 1997) transfected with angiomotin or vector (Troyanovsky, et al., 2001 ) were maintained in Dulbecco's modified Eagle's medium (DMEM, Sigma, Sweden) containing 10 % fetal bovine serum (FBS, Gibco, Sweden), 1 % penicillin and 1 % glutamine. The TUBO cell line, kindly provided by Dr. Guido Fomi (Department of Clinical and Biological Sciences, University of Turin, Orbassano, Italy), was derived from a spontaneous mammary tumour that arose in a BALB NeuT transgenic mouse expressing a transforming rat neu oncogene (Rovero, et al., 2000). Cells were cultured in Iscoves Modified Dubecco's Medium (IMDM ₁ Sigma, Sweden) with 10 % FBS. hTERT -immortalised bovine capillary endothelial (BCE) cells (Veitonmaki, ef a/., 2003) were grown in DMEM (Sigma, Sweden) with 10 % fetal bovine serum, 2 ng ml ^*1 FGF-2 (Peprotech, London, UK), 1 % glutamine and 1 % penicillin/streptomycin.

FACS analysis

For evaluation of cell surface binding by flow-cytometry, MAE cells were incubated with individual scFv clones at a concentration of 10 μg mr ¹ in PBS (Invitrogen) containing 0.5 % w/v BSA (DPBS-B) for 1 ,5 hr. Detection of scFv binding was achieved by incubation with anti-flag-biotin (Sigma, Sweden) followed by Streptavidin-Alexa 647 Fluor (Molecular probes). Living cells were defined as negative for SYTOX Green Nucleic Acid Stain (Molecular probes). All incubations were performed on ice.

Tissue cross reactivity

Antibodies were tested at concentrations of between 5 and 20 μg ml ^"1 towards acetone fixed frozen sections (8 μm) of normal human tissues (placenta from 3 individuals, liver, kidney, heart, pancreas, lymph node and cerebrum from 2 individuals). The human IgGs were pre-incubated in tubes with a biotinylated monovalent goat Fab anti-human IgG fragment (Jackson, cat. No. 109-067-003) prior to the staining procedure.

Staining was performed using the avidin-biotin complex (ABC) method. Slides were evaluated under light microscope (Nikon, Labophot-2) and photos were taken with a Leica DMR microscope. Monoclonal mouse anti-human CD34 class II, clone QBEnd (Dakocytomaiton, code no. M7165) was used as a positive control antibody while an nCoDeR derived lgd directed towards FITC was used as a negative control.

Migration assay

Migration assays were performed in a modified Boyden chamber using a 48-well chemotaxis chamber (Neuroprobe Inc., Gaithersburg, MD) as described earlier

(Kundra, et al., 1995). Briefly, 8 μm Nucleopore polyvinylpirrolidine-free polycarbonate filters were coated with 100 μg ml ^"1 of collagen type 1 (Cohesion,

Palo Alto, CA) overnight. hTERT-BCE cells were starved in 0.2 % FCS-DMEM for 16 h. The cells were trypsinised, resuspended in DMEM containing 0.1 % bovine serum albumin (BSA) and 30,000 cells were added with or without B06 scFv or control scFv to each well of the upper chamber.

FGF-2 (Peprotech EC Ltd, Rocky Hill, NJ) at 30 ng ml ^"1 or VEGF at 50 ng ml ^"1 (Peprotech EC Ltd, Rocky Hill, NJ) were used as chemo-attractants in the lower chambers. The chemotaxis chambers were incubated for 3-5 h at 37 °C with 10 % CO ₂ to allow cells to migrate through the collagen-coated polycarbonate filter. Non-migrating cells on the upper surface of the filter were removed and the filter was stained with Giemsa Stain (VWR International Ltd, West Chester, PA).

The total number of migrated cells per field was counted at * 20 magnification; each sample was tested in quadruplicates in at least three independent experiments.

Matriqel assay in vitro

150 μl of liquid Matrigel (Becton Dickinson, Biosciences) was added to each well of 8-well chamber slides (BD Falcon ™) and incubated at 37 ⁰ C for 30 min to allow the gel to polymerise. 1 ,5 x 10 ⁵ mouse aortic endothelial cells (MAE) maintained in serum-free DMEM with 0,5 % BSA were pre-treated for 16 hours with 5 μg ml ^"1 of single chains (Bioinvent International, AB) and seeded on a layer of polymerised matrigel as previously described (Troyanovsky, et a/., 2001 ). After 24 hours the changes in cell morphology were examined using a phase-contrast microscope.

Matriqel plug assay in vivo

Matrigel plug assays were performed as described previously with some modifications (Passaniti, et al., 1992). BALB/c mice were anaesthetised with lsofluran (Forene ^® , Abbot, Sweden) and injected with matrigel mixture subcutaneously in the abdominal midline.

In the FGF-2-induced angiogenesis model, every matrigel plug contained 0,5 ml of Matrigel (Becton Dickinson, Biosciences), 200 ng ml ^"1 FGF-2 (Peprotech, London, UK), 250 μg of control antibodies in one group and 250 μg of B06 scFv in another.

Two control groups of mice were injected either with matrigel containing only FGF-2 or matrigel alone. Matrigel plugs were excised seven days after implantation, photographed and processed for histological studies.

In the tumour-induced angiogenesis model every matrigel plug contained 75,000 TUBO cells and 500 μg of single chain antibodies in 0,5 ml of liquid Matrigel (Becton Dickinson, Biosciences). On day seven after gel implantation matrigel plugs were removed and prepared for immunohistochemical examination. Monoclonal rat anti-mouse CD31 (BD Pharmingen™) and anti-rat-FITC- conjugated (Jackson, Immunoresearch) antibodies were used to visualise vascularisation of matrigel plugs.

To analyse the microvessel density three images from every matπgel plug were taken with a Zeiss Axioplan 2 fluorescence microscope The vessels were counted under the microscope at 20 x magnification

Retinal angiogenesis assay

Intraocular injections were performed as described previously (Gerhardt, et al , 2003) Briefly, pups (P4) were anaesthetized by isofluran inhalation Injections (0 5 μl of -5 μg μl ¹ ScFv BO6 or ScFv CT17 in PBS) were performed using 10 μl gastight Hamilton syringes equipped with 34 gauge needles attached to a micromanipulator Three litters of C57BI6/J mice were treated, 10 pups per group The uninjected eyes served as additional control After 24h, pups were euthanised, eyes collected and fixed in 4 % PFA, and retinas were dissected and treated as described previously (Gerhardt, et al , 2003)

For immunohistochemistry, endothelial cells and microglial cells were visualised with biotinylated isolectin B4 (Bandeiraea simplicifolia, Sigma-Aldπch) followed by streptavidin Alexa Fluor 488 (Molecular Probes, Invitrogen) Retinas were flat- mounted and analysed using a Zeiss SV1 1 fluorescence stereomicroscope equipped with an Axiocam HRc The distance from optic nerve to vascular front was measured using Axiovision 4 5 software (Zeiss, Imaging Associates Ltd)

Images for filopodia analysis were taken on a Zeiss LSM 510 confocal microscope using 40 x 1 2 NA lens (settings pinhole 1 airy unit, 1024 x 1024 pixel) 10 z-ιmages were collected at 0 4 μm interval and presented as projection Projection-images were converted to grey scale and inverted to facilitate filopodia visibility The outline of the vessels at the migration front was measured and filopodia counted using ImageJ 1 36b (NIH ₁ public domain software) Data were analysed by unpaired two-tailed T-test and graphed using

Prism 4 software (GraphPad)

Laser-induced CNV in Mice CNV was generated by krypton laser-induced rupture of Bruch's membrane, as previously described (Berglin, et al , 2003) Briefly, three krypton laser

photocoagulation bums (50 μm spot size, 0.1 s duration, 120 mW power) were induced in each eye of C57BL/6J mice using a handheld contact lens (647 nm, Spectra-Physics 265 Exciter, Lasertek, Helsinki, Finland).

Mice received IP injections with 400 μg PEG FabB06 or CT17 every second day with the first injection given one day before laser treatment. Eyes were enucleated 10 days after laser treatment and fixed in 4 % paraformaldehyde for 30 min, the cornea and lens were removed and the entire retina was carefully dissected from the eyecup. The RPE-choroid-sclera eyecups were rinsed in PBS, permeabilised in 0.5 % Triton X-100 and blocked with 3 % goat serum in PBS/Triton X-100.

The eyecups were incubated with biotinylated isolectin B4 (1 :100 dilution, lectin from Griffonia simplicifolia, Sigma, CA, USA) and anti-CD31 (BD Biosciences Pharmingen, CA, USA) overnight at 4 ⁰ C, followed by incubation with Texas Red Streptavidin (1 :100 dilution, Vector Laboratories, Burlingame, CA, USA) and Alexa 488 goat anti-rat (1 :100, Molecular Probes, OR, USA). Radial cuts were made from the edge of the eyecup to the equator and the eyecup was flattened and mounted with antifade medium (Vectashield Mounting Medium, Vector Laboratories, Burlingame, CA, USA) with the sclera facing down and the choroid facing up. Flat mounts were examined with a fluorescence microscope (Axioskop 2, Carl Zeiss, Jena, Germany), images were captured with a digital camera (Carl Zeiss, Jena, Germany) and further analysed using AxioVision LE software (Carl Zeiss, Jena, Germany).

Lesions were manually measured in a masked fashion and data from each lesion was treated as a single statistical point. The outline of isolectin B4 staining was used to estimate the total plaque area and the vascularisation was estimated from the PECAM staining by quantifying the number of PECAM-positive pixels per plaque.

Results

AMOT antibodies have been screened for anti-migratory activity in vitro. We have now isolated antibodies that also inhibits tumor and bFGF-induced angiogenesis in the matrigel plug assay (Figure 16). The leading candidate also efficiently

inhibits laser induced angiogenesis in the mouse retina (in collaboration with Anders Kvanta, St Eriks Ogonsjukhus) which is a model system for age-related macular degeneration. Finally, we have shown (in collaboration with Holger Gerhardt, London Research Institute) that the AMOT antibodies inhibit vessel migration in the model of retinal angiogenesis of neo-natal mice (Figure 16).

These data provide direct proof that it is possible to mimic the effect of angiostatin using angiomotin as a target.

Example 3

Validation of AMOTL1 as an anti-angiogenic target.

Expression pattern:

We have mapped the expression pattern of AM0TL1 together with the protein atlas consortium (KTH and Uppsala University), lmmunohistochemical analysis of tissues from over 50 normal human tissues and 20 different types of cancers have been performed.

A typical example of lmmunohistochemical staining methodology provides a brown-black staining due to the specific binding of an antibody to its corresponding antigen. The tissue section can be counterstained with hematoxylin to enable visualization of microscopical features. Hematoxylin staining is unspecific and results in a blue coloring of both cells and extracellular material.

Tissue microarrays provide the possibility to immunohistochemically stain a large number and variety of normal and cancer tissues. The tissue microarrays used include samples from 48 different normal tissue types and 20 different types of cancer. For each antibody, protein expression patterns in normal tissue can be viewed as triplicate samples and in cancer tissue as duplicate samples. Normal tissues are sampled from 144 (48 x 3) different individuals and cancer tissues are derived from 216 (216 x 2) unique tumors. Normally, a fraction (<5%) of the 576 images are missing for each antibody due to technical issues. Samples of normal

and cancer tissue have been collected from anonymized paraffin embedded material of surgical specimens, in accordance with approval from the local ethics committee.

Since specimens are derived from surgical material, normal is here defined as non-neoplastic and morphologically normal. It is not always possible to obtain fully normal tissues and thus several of the tissues denoted as normal will include alterations due to inflammation, degeneration and tissue remodeling. In rare tissues, hyperplasia or benign proliferations are included as exceptions. It should also be noted that within normal morphology there exists inter-individual differences and variations due to primary diseases, age, sex etc. Such differences may also effect protein expression and thereby immunohistochemical staining patterns.

Samples from cancer are also derived from surgical material. The inclusion of tumors has been based on availability and representativity. Due to subgroups and heterogeneity of tumors within each cancer type, included cases represent a typical mix of specimens from surgical pathology. However, an effort has been made to include high and low grade malignancies where such is applicable. In certain tumor groups, subtypes have been included, e.g. breast cancer includes both ductal and lobular cancer, lung cancer includes both squamous cell carcinoma and adenocarcinoma and liver cancer includes both hepatocellular and cholangiocellular carcinoma etc. Tumor heterogenity and inter-individual differences is also reflected in diverse expression of proteins resulting in variable immunohistochemical staining patterns.

A cell microarray has been used to enable immunohistochemical staining of a panel of cell lines and cell samples. Duplicates from 47 cell lines and 12 samples of primary blood cells renders a total of 118 cell images per antibody. Normally, a small fraction of the images are missing due to technical issues. All cells were fixed in 4% paraformaldehyde prior to parafin embedding and immunohistochemical staining.

The cell microarray enables a wide representation of lymphoma/leukemia as well as other cell types that are difficult to include in a large scale atlas due to limited availability. Certain phenotypes not present in the tissue microarrays are also

included, e.g. sarcoma, choriocarcinoma, small cell lung carcinoma. In addition, several solid tumors included in the tissue microarrays are represented by a cell line derived from corresponding tumor type. The simultaneous staining using the same protocol for both tissue and cell arrays allows for comparison of IHC staining between cell lines and tissue.

The immunohistochemical protocols used result in a brown-black staining, localized where an antibody has bound to its corresponding antigen. The section is furthermore histochemically counterstained with hematoxylin to enable visualization of microscopical features. Hematoxylin staining is unspecific, and results in a blue coloring of both cells and extracellular material.

In general, AMOTL1 is only expressed in human placenta (cytotrophoblasts and endothelium). However, endothelial expression could be detected in the majority of cancers analyzed (Figure 17-A). We have also mapped the expression pattern during mouse embryogenesis. AMOTL1 and AMOT could be detected in sprouting capillaries of the brain and in other neural tissues. We therefore conclude that AMOTL 1 is up-regulated in angiogenic endothelium.

Cellular localization and protein binding:

lmmunofluorescent stainings of mouse endothelial cells shows that AMOTL1 and AMOT overlap in their sub-cellular localization. Both proteins are found exclusively in the cell-cell adhesion contacts (Figure 17-B). Furthermore, these proteins do not only colocalize but also interact to form a protein complex as shown by co-immunoprecipitation analysis (Figure 17-C). lmmunoprecipitation lmmunoprecipitation was performed as described before (Bratt et al). Briefly, cells were transfected and 24 hours later extracted with lysis buffer and centrifuged. The supernatant were incubated with antibody for 2 hours and thereafter with sepharose A for 2 hours. Pellets were washed four times with lysis buffer and resuspended in 1x sample buffer and analyzed by SDS-PAGE.

Western blot Western blot was performed as described before (Emkvist et al). For total retinal proteins extract, eyes were briefly fixed for 3-5 min at RT in 4% PFA and rinsed in

PBS. Retinas were then dissected as previously described (Chang ling et al., 1990) and homogenized in lysis buffer (1 % Triton X-100, 1% deoxycholic acid, 50 mM Tris (pH 8.0), 100 mM NaCI, 5 mM EDTA). The blot was also probed with D- actin antibody as a loading control (Clone AC-15, Sigma, #A5441 ). Bands were quantified using Image J 1.33 software (NIH, USA).

Experience from the AMOT KO project: Silencing if the AMOT gene results in embryonic lethality of 70-90 % of the embryos due to vascular insufficiency. Although this clearly shows that AMOT is essential it also indicates the presence of genes that can be upregulated and that can rescue the observed phenotype. The expression pattern and the interaction of AMOT/AMOTL1 indicate a functional overlap.

AMOTL1 DNA vaccination.

DNA vaccination using human AMOTL1 inserted into the pCDNA3 vector has been tested using the protocol established for AMOT (Holmgren et al. 2006). Animals were vaccinated twice at -21 and -7 days before challenge with the TSA mouse breast cancer tumor cell-line. This cell-line does not express AMOTL 1 or AMOT thus excluding that the effect is targeting the tumor cells. Preliminary data showed that a lower tumor take was observed in AMOT, AMOTL1 or mice treated with a combination of both genes (Figure 18). Furthermore, combination treatment with both AMOT and AMOTL1 showed T/C (treated over control) of over 90 % as estimated by tumor volume (note that tumor diameter is plotted in Figure 18).

Example 4 - Preferred pharmaceutical formulations and modes and doses of administration.

The polypeptides, polynucleotides and antibodies of the present invention may be delivered using an injectable sustained-release drug delivery system. These are designed specifically to reduce the frequency of injections. An example of such a system is Nutropin Depot which encapsulates recombinant human growth hormone (rhGH) in biodegradable microspheres that, once injected, release rhGH slowly over a sustained period.

The polypeptides, polynucleotides and antibodies of the present invention can be administered by a surgically implanted device that releases the drug directly to the required site. For example, Vitrasert releases ganciclovir directly into the eye to treat CMV retinitis. The direct application of this toxic agent to the site of disease achieves effective therapy without the drug's significant systemic side- effects.

Electroporation therapy (EPT) systems can also be employed for administration. A device which delivers a pulsed electric field to cells increases the permeability of the cell membranes to the drug, resulting in a significant enhancement of intracellular drug delivery.

Polypeptides, polynucleotides and antibodies of the invention can also be delivered by electroincorporation (El). El occurs when small particles of up to 30 microns in diameter on the surface of the skin experience electrical pulses identical or similar to those used in electroporation. In El, these particles are driven through the stratum corneum and into deeper layers of the skin. The particles can be loaded or coated with drugs or genes or can simply act as "bullets" that generate pores in the skin through which the drugs can enter.

An alternative method of administration is the ReGeI injectable system that is thermosensitive. Below body temperature, ReGeI is an injectable liquid while at body temperature it immediately forms a gel reservoir that slowly erodes and dissolves into known, safe, biodegradable polymers. The active drug is delivered over time as the biopolymers dissolve.

Polypeptides, polynucleotides and antibodies of the invention can be introduced to cells by "Trojan peptides". These are a class of polypeptides called penetratins which have translocating properties and are capable of carrying hydrophilic compounds across the plasma membrane. This system allows direct targeting of oligopeptides to the cytoplasm and nucleus, and may be non-cell type specific and highly efficient (Derossi et al., 1998, Trends Cell Biol., 8, 84-87).

Preferably, the pharmaceutical formulation of the present invention is a unit dosage containing a daily dose or unit, daily sub-dose or an appropriate fraction thereof, of the active ingredient

The polypeptides, polynucleotides and antibodies of the invention can be administered by any parenteral route, in the form of a pharmaceutical formulation comprising the active ingredient, optionally in the form of a non-toxic organic, or inorganic, acid, or base, addition salt, in a pharmaceutically acceptable dosage form. Depending upon the disorder and patient to be treated, as well as the route of administration, the compositions may be administered at varying doses.

in human therapy, the polypeptides, polynucleotides and antibodies of the invention can be administered alone but will generally be administered in admixture with a suitable pharmaceutical exipient diluent or carrier selected with regard to the intended route of administration and standard pharmaceutical practice.

The polypeptides, polynucleotides and antibodies of the invention can also be administered parenterally, for example, intravenously, intra-arterially, intraperitoneally, intra-thecally, intraventricularly, intrasternally, intracranially, intra-muscularly or subcutaneously, or they may be administered by infusion techniques. They are best used in the form of a sterile aqueous solution which may contain other substances, for example, enough salts or glucose to make the solution isotonic with blood. The aqueous solutions should be suitably buffered (preferably to a pH of from 3 to 9), if necessary. The preparation of suitable parenteral formulations under sterile conditions is readily accomplished by standard pharmaceutical techniques well-known to those skilled in the art.

Formulations suitable for parenteral administration include aqueous and nonaqueous sterile injection solutions which may contain anti-oxidants, buffers, bacteriostats and solutes which render the formulation isotonic with the blood of the intended recipient; and aqueous and non-aqueous sterile suspensions which may include suspending agents and thickening agents. The formulations may be presented in unit-dose or multi-dose containers, for example sealed ampoules and vials, and may be stored in a freeze-dried (lyophilised) condition requiring only the addition of the sterile liquid carrier, for example water for injections, immediately prior to use. Extemporaneous injection solutions and suspensions may be prepared from sterile powders, granules and tablets of the kind previously described.

Generally, in humans, oral or parenteral administration of the polypeptides, polynucleotides and antibodies of the invention is the preferred route, being the most convenient.

For veterinary use, the polypeptides, polynucleotides and antibodies of the invention are administered as a suitably acceptable formulation in accordance with normal veterinary practice and the veterinary surgeon will determine the dosing regimen and route of administration which will be most appropriate for a particular animal.

The formulations of the pharmaceutical compositions of the invention may conveniently be presented in unit dosage form and may be prepared by any of the methods well known in the art of pharmacy. Such methods include the step of bringing into association the active ingredient with the carrier which constitutes one or more accessory ingredients. In general the formulations are prepared by uniformly and intimately bringing into association the active ingredient with liquid carriers or finely divided solid carriers or both, and then, if necessary, shaping the product.

Preferred unit dosage formulations are those containing a daily dose or unit, daily sub-dose or an appropriate fraction thereof, of an active ingredient.

A preferred delivery system of the invention may comprise a hydrogel impregnated with a polypeptides, polynucleotides and antibodies of the invention, which is preferably carried on a tampon which can be inserted into the cervix and withdrawn once an appropriate cervical ripening or other desirable affect on the female reproductive system has been produced.

It should be understood that in addition to the ingredients particularly mentioned above the formulations of this invention may include other agents conventional in the art having regard to the type of formulation in question.

Example 5 - Exemplary pharmaceutical formulations

Whilst it is possible for a polypeptides, polynucleotides and antibodies of the invention to be administered alone, it is preferable to present it as a pharmaceutical

formulation, together with one or more acceptable carriers. The carriers) must be "acceptable" in the sense of being compatible with the compound of the invention and not deleterious to the recipients thereof. Typically, the carriers will be water or saline which will be sterile and pyrogen-free.

The following examples illustrate pharmaceutical formulations according to the invention in which the active ingredient is a polypeptides, polynucleotides and /or antibody of the invention.

Example 5A: Ophthalmic Solution

Active ingredient 0.5 g

Sodium chloride, analytical grade 0.9 g

Thiomersal 0.001 g Purified water to 100 ml pH adjusted to 7.5

Example 5B: Capsule Formulations

Formulation A

A capsule formulation is prepared by admixing the ingredients of Formulation D in Example C above and filling into a two-part hard gelatin capsule. Formulation B (infra) is prepared in a similar manner.

Formulation B mq/capsule

Active ingredient 250

Lactose B. P. 143

Sodium Starch Glycolate 25 Magnesium Stearate 2

420

Formulation C mq/capsule

Active ingredient 250 Macrogol 4000 BP 350

600

Capsules are prepared by melting the Macrogol 4000 BP, dispersing the active ingredient in the melt and filling the melt into a two-part hard gelatin capsule.

Formulation D mq/capsule

Active ingredient 250 Lecithin 100

Arachis Oil 100

450

Capsules are prepared by dispersing the active ingredient in the lecithin and arachis oil and filling the dispersion into soft, elastic gelatin capsules.

Formulation E (Controlled Release Capsule)

The following controlled release capsule formulation is prepared by extruding ingredients a, b, and c using an extruder, followed by spheronisation of the extrudate and drying. The dried pellets are then coated with release-controlling membrane (d) and filled into a two-piece, hard gelatin capsule.

mα/capsule

Active ingredient 250

Microcrystalline Cellulose 125

Lactose BP 125

Ethyl Cellulose 13

513

Example 5C: Injectable Formulation

Active ingredient 0.200 g Sterile, pyrogen free phosphate buffer (pH7.0) to 10 ml

The active ingredient is dissolved in most of the phosphate buffer (35-40 ^° C), then made up to volume and filtered through a sterile micropore filter into a sterile 10 ml amber glass vial (type 1 ) and sealed with sterile closures and overseals.

Example 5D: Intramuscular injection

Active ingredient 0.2O g

Benzyl Alcohol 0.10 g Glucofurol 75 ^® 1.45 g

Water for Injection q.s. to 3.00 ml

The active ingredient is dissolved in the glycofurol. The benzyl alcohol is then added and dissolved, and water added to 3 ml. The mixture is then filtered through a sterile micropore filter and sealed in sterile 3 ml glass vials (type 1 ).

Example 5E: SVγUD Suspension

Active ingredient 0.2500 g

Sorbitol Solution 1.500O g

Glycerol 2.0000 g

Dispersible Cellulose 0.0750 g

Sodium Benzoate 0.0050 g

Flavour, Peach 17.42.3169 0.0125 ml

Purified Water q.s. to 5.0000 ml

The sodium benzoate is dissolved in a portion of the purified water and the sorbitol solution added. The active ingredient is added and dispersed. In the glycerol is dispersed the thickener (dispersible cellulose). The two dispersions are mixed and made up to the required volume with the purified water. Further thickening is achieved as required by extra shearing of the suspension.

Example 5F: Suppository mq/suppository

Active ingredient (63 μm)* 250 Hard Fat, BP (Witepsol H15 - Dynamit Nobel) 1770

2020

The active ingredient is used as a powder wherein at least 90% of the particles are of 63 μm diameter or less.

One fifth of the Witepsol H15 is melted in a steam-jacketed pan at 45 ^° C maximum. The active ingredient is sifted through a 200 μm sieve and added to the molten base with mixing, using a silverson fitted with a cutting head, until a smooth dispersion is achieved. Maintaining the mixture at 45 ^° C, the remaining Witepsol H15 is added to the suspension and stirred to ensure a homogenous mix. The entire suspension is passed through a 250 μm stainless steel screen and, with continuous stirring, is allowed to cool to 4OO. At a temperature of 38 ^° C to 40 ^° C 2.02 g of the mixture is filled into suitable plastic moulds. The suppositories are allowed to cool to room temperature.

Example 5G: Pessaries mq/pessary

Active ingredient 250

An hydrate Dextrose 380 Potato Starch 363

Magnesium Stearate 7

1000

The above ingredients are mixed directly and pessaries prepared by direct compression of the resulting mixture.

Example 5H: Creams and ointments

Described in Remington, The Science and Practise of Pharmacy, 19 ^th ed., The Philadelphia College of Pharmacy and Science, ISBN 0-912734-04-3.

Example 5I: Microsphere formulations

The compounds of the invention may also be delivered using microsphere formulations, such as those described in Cleland (1997, Pharm. Biotechnol. 10:1 - 43; and 2001 , J. Control. Release 72:13-24).

References

Aepfelbacher, M., M. Essler, E. Huber, M. Sugai, and P. C. Weber. 1997. Bacterial toxins block endothelial wound repair. Evidence that Rho GTPases control cytoskeletal rearrangements in migrating endothelial cells. Arterioscler Thromb Vase Biol 17: 1623-9.

Balconi, G., R. Spagnuolo, and E. Dejana. 2000. Development of endothelial cell lines from embryonic stem cells: A tool for studying genetically manipulated endothelial cells in vitro. Arterioscler Thromb Vase Biol 20: 1443-51.

Berglin, L., S. Sarman, I. van der Ploeg, B. Steen, Y. Ming, S. Itohara, S. Seregard, and A. Kvanta. 2003. Reduced choroidal neovascular membrane formation in matrix metalloproteinase-2-deficient mice. Invest Ophthalmol Vis Sci 44: 403-8.

Bostrom, H., K. Willetts, M. Pekπy, P. Leveen, P. Lindahl, H. Hedstrand, M. Pekna, M. Hellstrom, S. Gebre-Medhin, M. Schalling, M. Nilsson, S. Kurland, J. Tornell, J. K. Heath, and C. Betsholtz. 1996. PDGF-A signaling is a critical event in lung alveolar myofibroblast development and alveogenesis. Ce// 85: 863-73.

Bratt, A., O. Birot, I. Sinha, N. Veitonmaki, K. Aase, M. Ernkvist, and L. Holmgren. 2005. Angiomotin regulates endothelial cell-cell junctions and cell motility. J Biol Chem 280: 34859-69.

Bratt, A., WJ. Wilson, B. Troyanovsky, K. Aase, R. Kessler, E.G. Van Meir, and L. Holmgren. 2002. Angiomotin belongs to a novel protein family with conserved coiledcoil and PDZ binding domains. Gene 298: 69-77.

Carmeliet, P., V. Ferreira, G. Breier, S. Pollefeyt, L. Kieckens, M. Gertsenstein, M. Fahrig, A. Vandenhoeck, K. Harpal, C. Eberhardt, C. Declercq, J. Pawling, L. Moons, D. Collen, W. Risau, and A. Nagy. 1996. Abnormal blood vessel development and lethality in embryos lacking a single VEGF allele. Nature 380: 435-9.

Childs, S., J.N. Chen, D. M. Garrity, and M. C. Fishman. 2002. Patterning of angiogenesis in the zebrafish embryo. Development 129: 973-82.

Covassin, L. D., J.A. Villefranc, M. C. Kacergis, B. M. Weinstein, and N. D. Lawson. 2006. Distinct genetic interactions between multiple Vegf receptors are required for development of different blood vessel types in zebrafish. Proc Natl Acad Sci U S A 103: 6554-9.

Ernkvist, M., K. Aase, C. Ukomadu, J. Wohlschlegel, R. Blackman, N. Veitonmaki, A. Bratt, A. Dutta, and L. Holmgren. 2006. p130-angiomotin associates to actin and controls endothelial cell shape. Febs J 273: 2000-11.

Ferrara, N., K. Carver-Moore, H. Chen, M. Dowd, L. Lu, K.S. O'Shea, L. Powell- Braxton, K.J. Hillan, and M. W. Moore. 1996. Heterozygous embryonic lethality induced by targeted inactivation of the VEGF gene. Nature 380: 439-42.

Franca-Koh, J. and P.N. Devreotes. 2004. Moving forward: mechanisms of chemoattractant gradient sensing. Physiology (Beth esd a) 19: 300-8.

Gerhardt, H., M. Golding, M. Fruttiger, C. Ruhrberg, A. Lundkvist, A. Abramsson, M. Jeltsch, C. Mitchell, K. Alitalo, D. Shima, and C. Betsholtz. 2003. VEGF guides angiogenic sprouting utilizing endothelial tip cell filopodia. J Cell Biol 161 : 1 163- 77.

Gluzman-Poltorak, Z., T. Cohen, Y. Herzog, and G. Neufeld. 2000. Neuropilin-2 is a receptor for the vascular endothelial growth factor (VEGF) forms VEGF-145 and VEGF-165 [corrected]. J Biol Chem 275: 18040-5.

Goishi, K. and M. Klagsbrun. 2004. Vascular endothelial growth factor and its receptors in embryonic zebrafish blood vessel development. Curr Top Dev Biol 62: 127-52.

Holmgren, L., E. Ambrosino, O. Birot, C. Tullus, N. Veitonmaki, T. Levchenko, L. M. Carlson, P. Musiani, M. lezzi, C. Curcio, G. Forni, F. Cavallo, and R. Kiessling. 2006. 32 A DNA vaccine targeting angiomotin inhibits angiogenesis and suppresses tumor growth. Proc Natl Acad Sci U S A 103: 9208-13.

Hynes, R. O. 2002. A reevaluation of integrins as regulators of angiogenesis. Nat Med 8: 918- 21.

Isogai, S., M. Horiguchi, and B. M. Weinstein. 2001. The vascular anatomy of the developing zebrafish: an atlas of embryonic and early larval development. Dev Biol 230: 278-301.

Isogai, S., N.D. Lawson, S. Torrealday, M. Horiguchi, and B. M. Weinstein. 2003. Angiogenic network formation in the developing vertebrate trunk. Development 130: 5281-90.

Jakobsson, L., J. Kreuger, K. Holmborn, L. Lundin, I. Eriksson, L. Kjellen, and L. Claesson- Welsh. 2006. Heparan sulfate in trans potentiates VEGFR-mediated angiogenesis. Dev Cell 10: 625-34.

Jones, E.A., F. Ie Noble, and A. Eichmann. 2006. What determines blood vessel structure? Genetic prespecificatioπ vs. hemodynamics. Physiology (Bethesda) 21 : 388-95.

Kiefer, F., S.A. Courtneidge, and E.F. Wagner. 1994. Oncogenic properties of the middle T antigens of polyomaviruses. Adv Cancer Res 64: 125-57.

Lawson, N.D. and B. M. Weinstein. 2002. In vivo imaging of embryonic vascular development using transgenic zebrafish. Dev Biol 248: 307-18.

Leslie, J. D., L. Ariza-McNaughton, A. L. Bermange, R. McAdow, S. L. Johnson, and J. Lewis. 2007. Endothelial signalling by the Notch ligand Delta-like 4 restricts angiogenesis. Development. 2007;134:839-44

Levchenko, T., K. Aase, B. Troyanovsky, A. Bratt, and L. Holmgren. 2003. Loss of responsiveness to chemotactic factors by deletion of the C-terminal protein interaction site of angiomotin. J Cell Sc/ 116: 3803-10.

Levchenko, T., A. Bratt, J. L. Arbiser, and L. Holmgren. 2004. Angiomotin expression promotes hemangioendothelioma invasion. Oncogene 23: 1469-73.

Levchenko, T., N. Veitonmaki, H. Gerhardt, Y. Ming, A. Kvanta, K. Berggren, R. Carlsson, and L. Holmgren. 2007. Therapeutic antibodies targeting angiomotin inhibit physiological and pathological angiogenesis in vivo. FASEB J. 2008 22:880-9.

Nobes, CD. and A. Hall. 1995. Rho, rac, and cdc42 GTPases regulate the assembly of multimolecular focal complexes associated with actin stress fibers, iameliipodia, and filopodia. Ce// 81 : 53-62.

Nobes, CD. and A. Hall -. 1999. Rho GTPases control polarity, protrusion, and adhesion during cell movement. J Cell Biol 144: 1235-44.

Olsson, A.K., A. Dimberg, J. Kreuger, and L. Claesson-Welsh. 2006. VEGF receptor signalling - in control of vascular function. Nat Rev MoI Cell Biol 7: 359- 71.

Ong, S. H., S. Dilworth, I. Hauck-Schmalenberger, T. Pawson, and F. Kiefer. 2001. ShcA and Grb2 mediate polyoma middle T antigen-induced endothelial transformation and Gab1 tyrosine phosphorylation. Embo J 20: 6327-36.

Richnau, N. and P. Aspenstrom. 2001. Rich, a rho GTPase-activating protein domaincontaining protein involved in signaling by Cdc42 and Rac1. J Biol Chem 276: 35060- 70.

Risau, W. 1997. Mechanisms of angiogenesis. Nature 386: 671-4.

Rubinson, D.A., CP. Dillon, A.V. Kwiatkowski, C. Sievers, L. Yang, J. Kopinja, D. L. Rooney, M. M. Ihrig, MT. McManus, F. B. Gertler, M. L Scott, and L. Van Parijs. 2003. A lentivirus-based system to functionally silence genes in primary mammalian cells, stem cells and transgenic mice by RNA interference. Nat Genet 33: 401-6.

Shimono, A. and R.R. Behringer. 2003. Angiomotin regulates visceral endoderm movements during mouse embryogenesis. Curr Biol 13: 613-7.

Shin, K., Q. Wang, and B. Margolis. 2007. PATJ regulates directional migration of mammalian epithelial cells. EMBO Rep 8: 158-64.

Siekmann, A.F. and N. D. Lawson. 2007. Notch signalling limits angiogenic cell behaviour in developing zebrafish arteries. Nature 445: 781 -4.

Singer, S.J. and A. Kupfer. 1986. The directed migration of eukaryotic cells. Annu Rev Cell Biol 2: 337-65.

Soker, S., S. Takashima, H. Q. Miao, G. Neufeld, and M. Klagsbrun. 1998. Neuropilin-1 is expressed by endothelial and tumor cells as an isoform-specific receptor for vascular endothelial growth factor. Cell 92: 735-45.

Troyanovsky, B., T. Levchenko, G. Mansson, O. Matvijenko, and L. Holmgren. 2001.

Angiomotin: an angiostatin binding protein that regulates endothelial cell migration and tube formation. J Cell Biol 152: 1247-54.

Weinstein, B. M. 2002. Plumbing the mysteries of vascular development using the zebrafish. Semin Cell Dev Biol 13: 515-22.

Wells, CD., J.P. Fawcett, A. Traweger, Y. Yamanaka, M. Goudreault, K. Elder, S. Kulkami, G. Gish, C. Virag, C. LJm, K. Colwill, A. Starostine, P. Metalnikov, and T. Pawson. 2006. A Rich1/Amot complex regulates the Cdc42 GTPase and apical- polarity proteins in epithelial cells. Cell 125: 535-48.

Williams, R.L., S. A. Courtneidge, and E. F. Wagner. 1988. Embryonic lethalities and endothelial tumors in chimeric mice expressing polyoma virus middle T oncogene. Cell 52: 121 -31.

Vittet, D., M. H. Prandini, R. Berthier, A. Schweitzer, H. Martin-Sisteron, G. Uzan, and E. Dejana. 1996. Embryonic stem cells differentiate in vitro to endothelial cells through successive maturation steps. Blood 88: 3424-31.

Yancopoulos, G. D., M. Klagsbrun, and J. Folkman. 1998. Vasculogenesis, angiogenesis, and growth factors: ephrins enter the fray at the border. Cell 93: 661-4.