AU2012244351A1 - Antibodies binding to the extracellular domain of the Receptor Tyrosine Kinase ALK - Google Patents

Abstract

C \NRPonbI\DCC\SCGI713% -I DOC-1 III02012 The present invention concerns an antibody specific for human ALK (Anaplastic Lymphoma Kinase), in particular a scFv, a nucleic acid sequence encoding it, its production and its use as a pharmaceutical or for diagnostic purposes. Said antibody is suitable for the local treatment of tumors, in particular glioblastoma.

Description

AUSTRALIA PATENTS ACT 1990 DIVISIONAL APPLICATION NAME OF APPLICANTS: Delenex Therapeutics AG ADDRESS FOR SERVICE: DAVIES COLLISON CAVE Patent Attorneys I Nicholson Street Melbourne, 3000 INVENTION TITLE: Antibodies binding to the extracellular domain of the Receptor Tyrosine Kinase ALK The following statement is a full description of this invention, including the best method of performing it known to us: C.\NRPortbP\DCC\SCG\4714342 1 DOC - 10/1/12 C:\RjonblDCC\SCG7139)I DOC-31/10/2012 Antibodies binding to the extracellular domain of the Receptor Tyrosine Kinase ALK Related Information This is a divisional of Australian Patent Application No. 2007246144, the entire contents of which are incorporated herein by reference. The application claims priority to U.S. provisional patent application number 60/795,831, filed on April 28, 2006, the entire contents of which are hereby incorporated by reference. The contents of any patents, patent applications, and references cited throughout this specification are hereby incorporated by reference in their entireties. Technical Field The present invention concerns an antibody specific for human ALK (Anaplastic Lymphoma Kinase), in particular a scFv, a nucleic acid sequence encoding it, its production and its use as a pharmaceutical or for diagnostic purposes. Said antibody is suitable for local treatment of tumors or cancer, in particular glioblastoma. Background Art ALK (Anaplastic Lymphoma Kinase; CD246) is a member of the receptor tyrosine kinase (RTK) family. As a typical member of this family, it is a type-I transmembrane protein essentially consisting of three domains: the extracellular ligand-binding domain (aal9-1038), which contains one LDL receptor class A domain and two MAM domains (MAM: Meprin, AS antigen, protein tyrosine phosphatise p), a transmembrane domain (aa1039-1059) and a cytoplasmic domain (aa 1060-1620), containing the tyrosine kinase domain. A signal peptide is present at the N- WO 2007/124610 PCTI/CH2007/000202 2 terminus of the nascent protein (aa 1-18) , which is cleaved upon secretion. The full-length human and mouse ALK were cloned in 1997 by two independent groups (Iwahara 1997; Morris 1997) . ALK is highly similar to the RTK called Leukocyte Tyrosine Kinase (LTK) and .belongs to the insu lin receptor superfamily. ALK exhibits 57% aa identity and 71% aa similarity with LTK in their regions of over lap (Morris 2001) . ALK is highly N-glycosylated and con tains 21 putative N-glycosylation .sites. Amino acids 687. to 1034 have significant similarity (50% aa identity) to LTK. However, the N-terminus proximal 686 aa sequence shows no homology to any known proteins with the excep tion of a very short sequence also found in the LDL re ceptor. (Duyster 2001/SWISSPROT) . In addition, it contains two MAM domains at aa264-427 and aa478-636 (Meprin, A -5 antigen, protein tyrosine phosphatase -p) . These domains are thought to have an adhesive function, as they are widespread among various adhesive proteins implicated in cell-to-cell interaction (De Juan 2002). Furthermore, there is a binding site for the ALK putative ligands cor responding to amino acids 396-406 (Stoica 2001; see be-, low) . The amino acid sequence of the kinase domain of murine ALK shows 98% aa-ident-ity to human ALK, 78% iden tity to mouse LTK, 52% to mouse ros, 47% to human insu lin-like growth.factor receptor and 46% to human insulin receptor (Iwahara 1997;.Ladanyi 2000). No splice variants of ALIR have' been described to date. However, ALK is often associated with chromosomal translocations (see below) The ALK gene spans about 315kb .and has 26 ex ons. Much of the gene consists of two large introns that span about 1'70kb. The ALK transcript is 6.5kb of length WO 2007/124610 PCT/CH2007/000202 3 (Kutok 2002) . According to Morris, the cDNA spans 6226bp (Morris 2001) ALK expression in mice starts during embryo genesis around the development stage Ell and is persist ing in the neonatal periods of development where it is expressed in the nervous system. In the adult, its physiological expression is restricted to certain neu ronal .(neural and glial cells and probably endothelial cells) regions of the CNS at low levels (Morris 1997; Duyster 2001; Stoica 2001) Actually, the abundance of ALK decreases in the postnatal period (Morris 2001). Based on its expression pattern, a role for the receptor iri brain development is suggested (Duyster 2001) . The neural-restricted expression. of ALK suggests that it serves as a receptor for neurotrophic factors (see later) . Consistent with this, its expression pattern overlaps with the genes encoding the TRK family of neuro trophin receptors (Morris 2002). However, ALK knockout mice do not show any obvious phenotype (unpublished data), which might be due to some functional redundancy with TRK family members or other neurotrophin receptors. Notably,' hematopoietic tissues show no detectable expres sion of ALK (see later) (Morris 2001). Two potential ligands for ALK have recently been described, "pleiotrophin" (PTN) and "midkine" (MK) (Stoica 2001; Duyster 2001; Stoica 2002). The PTN-ALK in teraction was identified by using purified human plei otrophin protein to screen a phage display peptide li brary. By this method, a sequence of ALK present in its extracellular domain (aa 396-406) was identified. Impor tantly, this sequence is not shared with LTK, the RTK most closely related to ALK. This ligand-binding region is also conserved in the potential homologue of ALK in WO 2007/124610 PCT/CH2007/000202 4 Drosophila (Loren 2001). ALK is phosphorylated rapidly upon PTN binding (Bowden 2002). Moreover, ALK has been shown to be stimulated by pleiotrophin in cell culture. This makes the pleiotrophin/ALK interaction particularly interesting in the light of the pathological implications pleiotrophin has (Stoica 2001). Cell lines that lack ALK expression also fail.. to show a growth response to plei otrophin and vice versa (Stoica 2001) . In vivo, elevated pleiotrophin levels in the serum of patients suffering from various solid tumors have been demonstrated, and animal. studies have suggested a. contribution of pleiotro phin to tumor growth (Stoica 2001) The role of PTN as rate-limiting angiogenic factor in tumor growth is well established in animal models (Choudhuri 1997) . In 1996 Czubayko et al. (Czubayko 1996) demonstrated the. impor tanceof PTN in tumor angiogenesis,' in prevention of apoptosis and metastasis by modulating PTN levels with'a ribozyme targeting approach. Serum level measurements of PTN in mice demonstrated a. clear correlation with the size.of the tumor. PTN plays a significant role in some of the most aggressive human cancer. types such as mela noma and pancreatic cancer thus giving interesting per spectives for potential further applications of an ALK inhibitor (Weber 2000; Stoica 2001) . In human patients, elevated serum pleiotrophin levels were found in patients with pancreatic cancer (n = 41; P<..0001) and colon cancer (n = 65; P = .0079) . In healthy individuals, PTN is ex pressed in a tightly regulated manner during perinatal organ development and in selective populations of neurons and glia in the adult. Co-expression of PTN and ALK, as found in several cancer cell lines, indicates that they could form an autocrine loop of growth stimulation (Stoica 2001). In WO 2007/124610 PCT/CH2007/000202 5 spite of all these data, literature tells that is not yet clear if the effects of PTN are mediated by ALK alone and/or by other unidentified PTN receptors (Duyster 2001). At least two other potential receptors of PTN have been suggested: the receptor tyrosine phosphatase RPTPP and the heparan sulfate proteoglycan N-syndecan. However, RPTPP might act as a signalling modulator of PTN/ALK sig nalling and N-syndecan as a chaperone for the ligand. (Bowden 2002). Recently, another secreted growth factor re lated to pleiotrophin called midkine (MK) has been iden tified as a second ligand for ALK. Similarly to PTN, binding and activating functions (e.g. induction of soft agar colony formation in cell cultures) of MK can be blocked by the same antibody raised against the ALK-ECD (Stoica 2001) . Like pleiotrophin, midkine is upregulated in many tumors, although its physiological expression is very restricted in adult normal tissues (Stoica 2002). Analysis of 47 bladder tumor samples revealed that MK ex pression is significantly (about four times) enhanced as compared to normal bladder tissue. Furthermore, pro nounced overexpressi'-on correlates with poor patient sur vival (O'Brien 1996) However, the affinity of MK for ALK is about 5 times lower than the one 'of pleiotrophin (Stoica 2002). Interestingly, as with pleiotrophin, inhibition of ALK via ribozymes also inhibits the effects of MK in cell culture (Stoica 2002) . The authors of these studies also come to the conclusion that inhibition of the PTK/MK/ALK pathway opens very attractive possibilities for the treatment of various diseases, some of them having very limited treatment options so far, such as, for example, glioblastoma and pancreatic cancer. (Stoica 2002) WO 2007/124610 PCT/CH2007/000202 In healthy individuals, ALK mRNA expression peaks during the neonatal period and persists in adults in a few selected portions of the nervous system. Re-. cently, expression of the ALK protein was also detected in endothelial cells that were associated to neuronal and glial cells. Evidence that at least a part of the malig nant activities described for pleiotrophin are mediated through ALK came from experiments in which the expression of ALK was depleted by a ribozyme targeting approach. Such depletion of ALK prevented pleiotrophin-stimulated phosphorylation of the anti-apoptopic protein Akt and led -to a prolonged survival of mice that had received xeno grafts. Indeed, 'the number of apoptopic cells in the tu mor grafts was' significantly increased, when ALK expres sion was depleted (Powers 2002). Evidence that malignant activities described for MK are mediated through ALK came from experiments with monoclonaI antibodies directed against- the ALK ECD. Addition of a 1:25 dilution of hybridoma cell supernatant from two anti-ALK ECD antibodies leads to a significant' decrease in colony formation of SW-13 cells in soft agar (Stoica 2002). Analysis of ten different cell lines re vealed that the ability for a growth response to PTN per fectly correlated with the expression of ALK mRNA (the following cell lines responded to PTN and were found to express ALK mRNA: HUVEC, NIH3T3, SW-13, Colo357, ME-180, U87, MD-MB 231; Stoica 2001). Interestingly, in some can cer cell lines (Colo357 pancreatic cancer, Hs578T breast cancer and U87 glioblastoma), PTN and ALK are co expressed, indicating that PTN and ALK form an autocrine loop of growth stimulation (Stoica 2001). Interestingly, both PTN and MK have been shown to cause transcriptional up-regulation of the anti- WO 2007/124610 PCT/CH2007/000202 7 apoptotic bcl-2 protein (Stoica 2002). In addition acti vated Akt (which is a crucial downstream target of aber rant ALK signalling) phosphorylates the pro-apoptotic factor called bad, thus leading to dissociation from bcl xl, which, when liberated from bad, can suppress apop tosis by blocking the release of cytochrome c (see Bowden 2002 for references). Aberrant expression of ALK might be involved in the development of several cancers. However, it was' first associated with a subgroup of high-malignant Non Hodgkin lymphomas (NHLs) , the so-called Anaplastic Large Cell Lymphomas (ALCLs). Non-Hodgkin lymphomas represent .clonal neoplasias originating from. various cells of lym-. phatic origin. Most patients with the primary systemic clinical subtype of ALCL have the-t2,5 translocation, ex pressing a fusion protein that joins the N-terminus of nucleophosmin (NPM) to the C-terminus of ALK. The fusion consists of aa 1-117 of NPM fused to aa 1058-1620 of ALK* and the chromosomal breakage is located in an intron lo cated between the exons encoding the TM and juxtamembrane domain of ALK '(Duyster 2001). NPM-ALK is a transcript containing an ORF of 2040bp encoding a 680aa protein (Morris 2001). This corresponds to a breakage in intron 4 of NPM, which spans.911 bp and intron 16 of ALK which spans 2094bp (Kutok 2002). Most likely the ALK sequence in this fusion protein is the minimal sequence required for the protein to lead to ALCL (Duyster 2001). 'The in verse fusion (ALK-NPM) is not expressed, at least not in lymphoid cells (Kutok 2002) . The wild-type NPM protein demonstrates ubiquitous expression and functions as a carrier of proteins from the cytoplasm into the nucleo lus. As a matter of fact, NPM is a 38kDa nuclear protein WO 2007/124610 PCT/CH2007/000202 8 encoded on chromosome 5 that contains a NLS, binds nu clear proteins and engages in cytoplasm/nuclear traffick ing (Duyster 2001). NPM is one of the most abundant nu cleolar proteins and is normally present as a hexamer (Morris 2001) . Most importantly NPM normally undergoes self-oligomerization (hexamers) as well as hetero cligomerization with NPM-ALK (Duyster 2001). The 2;5 translocation brings the ALK gene portion encoding the tyrosine kinase on chromosome 2 under the control of the strong NPM promoter on chromosome 5, producing permanent expression of the chimeric NPM-ALK protein (pBO) (Duyster 2001). Hence; ALK kinase is deregulated and ectopic, both in terms of cell. type (lymphoid) and cellular compartment (nucleus/nucleolus and cytoplasm). (Ladanyi 2000). The lo calization (.cytoplasm or nucleus) of .NPM seems not to af fect its effect on lymphomagenesis (Duyster 2001). The resultant aberrant tyrosine kinase activity triggers ma lignant transformation via constitutive phosphorylation of intracellular targets. Various other less common ALK fusion proteins are associated with ALCL, All variants demonstrate linkage of the ALK tyrosine kinase domain to an alternat ive promoter that regulates 'its expression.. Full-length ALK has been reported to be also expressed in about 92% of primary neuroblastoma cells and in some rhabdomyosarcomas (Lamant 2000). However, no. cor relation between ALK expression and tumor biology has been demonstrated so far. This fact, taken together with the lack of evidence regarding significant levels of endogenously phosphorylated ALK in these tumors, suggest that ALK expression in neuroblastoma reflects its normal expression in immature neural cells rather than a primary oncogenic role and ALK in these tumors is not constitu tively phosphorylated thus questioning an important role WO 2007/124610 PCT/CH2007/000202 9 for ALK in these tumors (Duyster 2001; Pulford 2001) Nevertheless, ALK signalling might be important in at least some neuroblastomas, as suggested by Miyake et al., who found overexpression and constitutive phosphorylation of ALK due to gene amplification in neuroblastoma-derived cell lines (Miyake 2002). However, other neuroblastoma derived cell 'lines do not show constitutive activation of AIK, thus arguing against a general pathological involve ment of ALK (Dirks 2002; Pulford 2004). Most interestingly, ALK seems to be important for'growth of glioblastoma multiforme, a highly malignant - brain tumor that offers very limited therapeutic options (Powers 2002). Multiple genetic alterations have been shown to occur in .these devastating tumors' including loss or mutations of PTEN, p53 and INK4a-ARF.- In addition, RTK. signalling plays a particularly important role in-growth and development of these tumors, which overexpress vari ous growth factors such as PDGF, HGF, NGF and VEGF sug gesting autocrine RTK signalling loops. Powers and col leagues have shown mRNA and protein expression of ALK in glioblastoma patient tumor samples, whereas the signals were not detectable in normal adjacent brain' tissue (Pow ers' 2002). Furthermore, 'human U87MG glioblastoma cells (which are derived from a patient and represent a'well characterized model system to study tumorigenesis and signalling in glioblastoma) show* ALK-dependent anti apoptotic behaviour in xenograft studies. When ALK is de pleted in these tumor cells by the use of ribozymes, mice injected with these tumor cells survive at least twice as long as when injected with wild-type tumor cells, and . these tumor cells show drastically increased apoptosis. Thus, ALK and its ligand(s) provide an essential survival signal that is rate-limiting for tumor growth of U87MG WO 2007/124610 PCT/CH2007/000202 10 cells in vivo (Powers 2002). These finding indicate that inhibition of ALK signalling could be a promising ap proach to improve life expectancy of glioblastoma pa tients. Glioblastoma multiforme is by far the most common and malignant primary glial tumor with an inci dence of about 2 /100'000/y (about 15'000 .cases in US and Western Europe per year). It affects preferentially the cerebral hemispheres, but can also affect -the brain stem (mainly in children) or the spinal cord. The. tumors may manifest de novo (primary glioblastoma) or may develop from lower grade astrocytomas (secondary glioblastoma). Primary and secondary glioblastomas show little molecular overlap and constitute different disease entities on mo lecular level. They both contain many genetic"abnormali ties including affection of p53, EGFR, MDM2, PDGF, PTEN, p16, RB. No significant therapy advancement has oc curred in the last 25 years. Therapies are only pallia tive and can expand the life ekpectance from 3 months to 1 year. Patients usually present with slowly progressive neurological deficit, e.g. motor weakness, intracranial pressure symptoms, e.g., headache, nausea, vomiting, cog nitive impairment, or seizures: Changes in personality can also be early signs. The etiology of glioblastoma is unknown, familial cases represent less than 1%. The only consistent risk factor identified is exposure to petro chemicals. Diagnosis is made mainly by imaging studies (CT, NMR) and biopsy. Completely staging most glioblas tomas is neither practical nor possible because these tu mors do not have clearly defined margins. Rather they ex hibit well-known tendencies to invade locally and spread along compact white matter pathways. The primary reason WO 2007/124610 PCT/CH2007/000202 why no curative treatment is possible is because the tu mor is beyond the reach of local control when diagnosed. The primary chemotherapeutic agents are carmustine (an alkylating agent) and cisplatinum but only 40% of pa tients show some response. Although there .are quite some uncertainties regarding the role of ALK in glioblastoma, this disease offers various approaches for ALK-directed drugs. In .fact, for this devastating disease even a small improve ment of current therapy options would serve an enormous medical need. It is important to note 'that since glioblastoma cells express the full-leigth ALK, for treating this cancer ALK could be considered as a target not only for small molecule kinase inhibitors but also for antibodies and/or antibody fragments such as scFvs i.e. to induce apoptosis of tumor cells. The strict lo calization of glioblastoma to the CNS supports the use of scPvs, if they can be delivered efficiently'to the CNS. (no rapid clearance due to compartimentalization, but better tumor penetration compared to IgGs due to their smaller size) . Antibodies 'and/or -antibody fragments- could be directed against the ligand-binding. sequence of- ALK (aa 396-406) or against other parts of the extracellular parts of the receptor. The very limited expression of ALK in healthy tissues under physiological conditions -indicates that tu mors expressing ALK might be an excellent target for dis ease treatment using radioactive or toxin-labelled anti bodies and/or antibody fragments, irrespective' of whether ALK is involved in the pathogenesis of these tumors or not. In addition to glioblastoma cells, ALK expression' has been found with high significance in melanoma cell lines and breast carcinoma cell lines (without being con- WO 2007/124610 PCT/CH2007/000202 12 stitutively phosphorylated) (Dirks 2002). The fact that a large portion of the extracellular domain of'ALK seems to be rather unique in the human proteome should make this approach highly specific. WO9515331/US5529925 discloses the cloning and sequencing of the human nucleic acid sequences, which are rearranged in the t(2;5) (p23;q35) chromosomal transloca tion event which occurs in human t(2;5) lymphoma. The re arrangement was found to bring sequences from the nucleo lar phosphoprotein gene (the NPM gene) on chromosome 5q35 to those from a previously unidentified protein tyrosine kinase gene (hereinafter the ALK gene) on chromosome 2p23. The sequence of the fusion gene and fusion protein (NPM/ALK fusion gene or protein, respectively) were also disclosed. The full-length ALK sequence is patented in US5770421, entitled "Human ALK Protein Tyro.sine Kinase." Furthermore, the- patent US6174674B1 entitled "Method of detecting a chromosomal rearrangement involving a break point in the AIK or NPM gene", discloses primers for de tecting the NPM-AIK fusion sequence in patient samples-, In another patent application, US6696548 entitled "ALK protein tyrosine kinase/receptor and ligands thereof", the. use- of ALK for detection of ALK ligands and antibod ies binding to specific sequences of ALK is disclosed. It also discloses a method of identifying an agent capable of binding to the isolated ALK polypeptide, W00196394/US20020034768 discloses ALK as receptor of pleictrophin. US20040234519 discloses anti-oleiotrophin. antibodies, and W02006020684 describes the detection of pleitrophin. Disclosure of the Invention WO 2007/124610 PCT/CH2007/000202 13 Hence, it is a general object of the inven tion to provide a stable and soluble antibody or antibody derivative, which binds the human ALK protein in vitro and in vivo. Most preferably, the antibody is specifi cally targeted against the ligand-binding domain of ALK (amino acids 396-406) and hence will block both the bio-. logic effects of MK, which has a Kd for ALK of about 170 pM, as well as the biologic ef f ects of PTN, which has a Ed for ALK of about 20 -30 pM (Stoica 2002; Stoica 2001) . In a preferred embodiment said antibody or antibody frag ment is a scFv antibody or a Fab fragment. In the follow ing the term antibody comprises full-length antibodies as well as other antibody derivatives. Now, in order to -implement these and still f-uar-t-her-ebjects-jof the invention, which will become more readily apparent as the description proceeds, said anti body is manifested by the features that it comprises a varia-ble heavy chain CDR3 of a sequence of at least 50% identity to the sequence SEQ. ID. No, 2. Preferably, .the sequence identity is at least 60%, 6.5%, 75%, 85%, or more preferably at least 92%.. Most preferably, said antibody has a VH CDR3 of the sequence SEQ. ID. No.2. In one embodiment, the 'antibody or antigen binding portion thereof of the invention specifically binds to a particular epitope of the ALK protein. Such epitopes reside, for example, within amino acids 1-50, 50-100, 100-150, 150-200, 200-250, 250-300, 300-350, 350 400, 400-450, 450-500, 500-550, 550-600, 600-650, 650 700, 700-750, 750-800, 800-900', 900-1OOD, 1000-1100, 1100-1200, 1200-1300, 1300-1400, 1400-1500, or 1500-1620 of the ALK protein, or any interval, portion or range thereof. In one embodiment, the antibody or antigen binding portion thereof specifically binds to an epitope WO 2007/124610 PCT/CH2007/000202 14 comprising, essentially consisting of or a fragment of the region spanning amino acid residues 391 t3 and 406 3 (SEQ. ID No: 91 shows amino acid residues 388 to 409 of the human ALK protein), preferably amino acids 391-406 (SEQ. ID.No: 1) of the ALK protein (SEQ ID NO: 1). It is understood that the indicated range is not to be consid ered as having sharp boundaries, but that the antibody or antigen binding portion thereof may bind or partially bind in a region closely situated -to or within the ligand-binding domain of ALK. Preferably, the antibodies or antibody-derivatives bind to an epitope of the ALK proteig'f 10 to 20 amino acids in length. In another embodiment the antibody or antigen binding portion thereof can be characterized as specifi cally binding to an ALK protein with a KD of less than about 10'x 10'' M. In a particular embodiment, the anti body or antigen binding portion thereof specifically binds to an. ALK protein (or fragment thereof) with a KD of at least about 10 x 10- M, at least about 10'x 10-8 K, at least about 10 x*.0~ M, at least about 10 x 10-1' M, at least about 10 x 10~" M, or at least about 10 x 10~1 M or a KD even more favorable. In various other embodiments, the antibody or anti gen binding portion thereof includes a variable heavy chain region comprising an amino acid sequence at least B0%, 85%, 90%, 95%, 98%, or more preferably at least 99% identical to a variable heavy chain region amino acid se quence as set forth in SEQ ID NO: 4. In other embodiments, the antibody or antigen binding portion thereof includes a variable light chain region comprising an amino acid sequence at least 80%, B5%, 90%, 95%, 98% or more preferably at least 99% iden- WO 2007/124610 PCT/C H2007/000202 15 tical to a variable light chain region amino acid se quence as set forth in SEQ ID NO: 5. In still other embodiments, the antibody or antigen binding portion thereof includes both a variable heavy chain region comprising an amino acid sequence at least 80%, 85%, 90%, 95%, 98% or more preferably at least 99% identical to a variable heavy chain region amino acid sequence as set forth in SEQ ID NO: 4 and a variable light chain region comprising an amino acid sequence at least 80%, 85%, 90%, 95%, 98% or more preferably at least 99% identical to a variable light chain amino acid se quence as set forth in SEQ ID NO: 5. In certain other embodiments, the antibody or antigen binding portion thereof specifically bind to an epitope that overlaps with an epitope bound by an anti body or antibody derivative of ESBA521 (Seq. ID. No. 19) and/or competes for binding to an ALK protein, or portion thereof, with an antibody or antibody derivative of ESBA521. In a related embodiment, the antibody or anti gen binding portion thereof specifically binds.to an epi tope comprising residues 391-406 (SEQ ID NO: 1) of an ALK protein, or portion thereof. The variable heavy and light chain regions of the antibodies or antigen binding portions thereof typi cally include one or more complementarity determining re gions (CDRs). These include the CDR1, CDR2, and CDR3 re gions. In particular embodiments, the variable heavy chain CDRs are at least 80%, 85%, 90%, 95%, or more pref erably 100% identical to a CDR of the ESBA521 antibody. In other particular embodiments, variable light chain CDRs are at least 80%, 85%, 90%, 95%, or more preferably 100%, identical to a CDR of a variable light chain region of the ESBAS21 antibody.

WO 2007/124610 PCT/CH2007/000202 16 Accordingly, particular antibodies or frag ments of the invention comprise a variable heavy chain region that includes one or more complementarity deter mining regions (CDRs) that are at least 80%, 85%, 90%, 95%, or more preferably 100%, identical to a CDR of a variable heavy chain region of the ESBA521 and a variable light chain region that includes one or more CDRs that are at least-80%, 85%, 90%, 95% or more preferably 100%, identical to a CDR of a variable light chain region of the ESBA521 antibody. The variable heavy chain region of the anti bodies or antigen binding portions thereof can also in clude all three CDRs that are at least 80%, 85%, 90%, 95%, or more preferably 100%, identical to the CDRs of the variable heavy chain region of the ESBA521. antibody and/or all three CDRs that are at least 80%, 85%, 90%, .95% or more preferably 100%, identical to the CDRs of the variable light chain region of the ESBA521 antibody. In another embodiment of the invention, the antibodies or antigen binding portions thereof (a) in cl ude a heavy chain variable region that is encoded-by or derived from (i.e. is the product of) -a human VH gene (e.g., H3 type); and/or (b) include a light chain vari able region that is encoded by or derived from a human V kappa or lambda gene. (e.g., lambdal type). The antibodies of the present invention in clude full-length antibodies, for example, monoclonal an tibodies, that include an effector domain, (e.g., an Fc domain), as well as antibody portions or fragments, such as single-chain antibodies and Fab fragments. The anti bodies can also be linked to a variety of therapeutic agents (e.g., anticancer agents, chemotherapeutics, or toxins) and/or a label (e.g., radiolabel).

WO 2007/124610 PCT/CRH2007/000202 17 In another aspect, the invention features iso lated nucleic acids including a sequence encoding an an tibody heavy chain variable region which is at least 75%, 80%, 85%, 90%, 95%, or more preferably at least 99%, identical to SEQ ID NO: 22. The invention also features isolated nucleic acids that include a sequence encoding an antibody light chain variable region which is at least 75%, 80%, 85%, 90%, 95%, or more preferably at least 99%, identical .to- SEQ ID NO: 21. The invention also features expression vectors including any of the foregoing nucleic acids either alone or in combination (e.g., expressed from one or more vec tors), as well as host 'cells comprising such expression vectors. Suitable host cells for expressing antibodies of the invention include a variety of eukaryotic cells, e.g.:, yeast cells, mammalian cells, e.g., Chinese hamster ovary (CHO) cells, NSO cells, myeloma cells, or plant cells. The molecules of the invention can also be ex pressed in prokaryotic cells, e.g., E. coil. The invention also features methods f6r making the antibodies or antigen binding portions -thereof by ex pressing nucleic acids encoding antibodies in a host cell (e.g., nucleic acids encoding the antigen binding region portion of an antibody). In yet another aspect, the in vention features a hybridoma or transfectoma including the aforementioned nucleic acids, In another embodiment, the invention provides an antigen comprising an epitope of the ALK protein, preferably of ,the PTN ligand binding domain, more pref erably a fragment comprising, essentially consisting of or a fragment of the region spanning amino acid residues 391 i3 and 406 L 3 (see SEQ. ID No: 91 which shows amino WO 2007/124610 PCT/CH2007/000202 18 acid residues 388 to 409 of the human ALK protein), most preferably amino acids 391-4D6 (SEQ. ID.No: 1) . The an tigen can be used for raising, screening, or detecting the presence of an anti-ALK antibody or can be used as an agent in active immunotherapy, i.e. as a vaccine. As a vaccine, the antigen can be used alone or in combination with an appropriate adjuvant or hapten, e.g., mixed or conjugated either chemically or geneti cally. The antigen when used for active immunotherapy can also be used in combination with passive immunother apy, for example, with any of the anti-A1K antibodies disclosed herein, or in combination with a monoclonal or polyclonal preparation of anti-ALK antibodies, e.g., se rum gammaglobulin from a seropositive donor. In another embodiment, the antibody molecules -(or VL and VH binding regions) are fully. human. Treat ment of humans with human monoclonal antibodies offers several advantages. For example, the antibodies are likely to be less immunogenic in humana than non-human antibodies. The therapy is also rapid because ALK inac tivation can occur as soon as the antibody reaches a can cer site (where ALK is 'expressed). Therefore, in a re lated embodiment, the antibody is a scFv antibody, i.e., ESBA521 or an antibody comprising a VL and/or VH re gion(s) (or CDRs thereof; e.g., VL CDR3 (SEQ ID NO:3) and/or VE CDR3 (SEQ ID NO: 2)) of ESBA521. Human antibodies also localize to appropriate sites in humans more efficiently than non-human antibod ies. Furthermore, the treatment is specific for ALK, is recombinant and highly purified and, unlike traditional therapies, avoids the potential of being contaminated with adventitious agents. Alternatively, antibodies and WO 2007/124611) PCT/CH2007/000202 19 antibody-derivatives of the present invention may be pro duced by chemical synthesis. In another embodiment, thie invention provides compositions for treating a cancer (or the, making of a medicament so suited) that can prevent neoplasia in a subject by competing with ligands of.ALK such as midkine (M() and/or pleiotrophin (PTN) and thereby block ALK signaling mediated by such ligands. Such t composition can be administered alone or in combination with in the art recognized anti-cancer agents, for example, meth otrexate, and the like. The antibody of the invention and/or ALK vac cine can be used alone or in combined with a known thera peutic, e.g., an anti-cancer agent, e.g.,' methotrexate and the like. Other features and advantages of the invention will be apparent from the following detailed description, and from the claims. Brief Description of the Drawings The invention will be better understood and objects other-than those set forth above will become ap parent when consideration is given to the following de tailed description thereof. Such description makes refer ence to the annexed drawings, wherein: Figure 1 shows a scheme of the human ALK pro tein used. A 16 amino acid peptide of the PTN binding site (dashed) is used as the epitope in a two hybrid screen for scPv binders. Figure 2 shows the stepwise randomization of VH CDR3, VL CDR 3 and. VH CDR 2 portions to obtain ESBA521 WO 2007/124610 PCT/CH2007/000202 20 as secondary binder and a set of scFvs as tertiary bind ers (see Tab. 1) . X stands for any amino acid residue. Fig. 3 shows an ELISA experiment wherein the binding characteristics of improved scFvs are compared to that of the framework they originate from. Fig. 4 shows immunostaining of transiently transfected HeLa cells with ESBAS21 (left panels) and a polyclonal ALK specific antibody (right panels). Middle panel: same cells visualized by light microscopy. Figure 5 shows an ELISA experiment comparing the ESBA512 to the improved tertiary binders. Detailed Description of the Invention In order that the present invention may be more readily understood, certain terms are first defined. Definitions The term "ALK" and "Alk-1" includes the human ALK protein encoded by the ALK (Anaplastic Lymphoma Ki nase) gene which is a membrane-spanning' protein tyrosine kinase (PTK)/receptor. The term "antibody" refers to whole antibodies and any antigen binding fragment (i.e., "antigen-binding portion," "antigen binding polypeptide, " or "immuno binder") or single chain.thereof. An "antibody" refers to a glycoprotein comprising at least two heavy (H) chains and two light (L) chains inter-connected by disul fide bonds, or an antigen binding portion thereof. Each heavy chain is comprised of a heavy chain variable region (abbreviated herein as VH) and a heavy chain constant re gion. The heavy chain constant region is comprised of three domains, CH1, CH2 and CH3. Each light chain is comprised of a light chain variable region (abbreviated WO 2007/124610 PCT/CH2007/000202 21 herein as VL) and a light chain constant region. The light chain constant region is comprised of one domain, CL. The VH and VL regions can be further subdivided into regions of hypervariability, termed complementarity de termining regions (CDR), interspersed with regions that are more conserved, termed framework regions (FR). Each VH and VL is composed of three CDRs and four FRs, arranged from amino-terminus to carboxy-terminus in the following. order: FRI, CDR1, FR2, CDR2, FR3, CDR3, FR4. The vari able regions of the heavy and light chains contain a binding domain that interacts with an antigen. The con stant regions of the antibodies may mediate the binding of the immunoglobulin to host tissues or factors, includ ing various cells of the immune system (e.g., 'effector cells) and the first component (Clq) of the classical complement system. The term "antigen-binding portion" of an anti body (or simply "antibody portion") refer to one or more fragments of an antibody -that retain the ability to spe cifically bind to an antigen (e.g., ALK). It has been shown that the antigen-binding function of an antibody can be performed by fragments of a full-length antibody. Examples of binding fragments encompassed within the term "antigen-binding portion" of an antibody include (i) a Fab fragment, a monovalent fragment. consisting of the VL, Ve,- CL and CH1 domains; (ii) a F(ab') 2 fragment, a biva lent fragment comprising two Fab fragments linked by a disulfide bridge at the hinge region; (iii) a Fd fragment consisting of the VH and CH1 domains; (iv) a. Fv fragment consisting of the VL and Vg domains of a single arm of an antibody, (v) a single domain or dAb fragment (Ward et al., (1989) Nature 341:544-546), which consists of a VH domain;:and (vi) an isolated complementarity determining WO 2007/124610 PCT/CH2007/000202 22 region (CDR) or (vii) a combination of two or more iso lated CDRs which may optionally be joined by a synthetic linker. Furthermore, although the two domains of the Fv fragment, VL and VH, are coded for by separate genes, they can be joined, using recombinant methods, by a synthetic linker that enables them to be made as a single protein chain in which the VL and Vg regions pair to form monova lent molecules (known as single chain Fv (scFv); see e.g., Bird e't al. (1988) Science 242:423-426;. and Huston et al. . (1988) Proc. Natl. Acad. Sci. USA 85:5879-5883). Such single chain antibodies are also intended to be en compassed within the term "antigen-binding portion" of an antibody. These antibody fragments are obtained using conventional techniques known to those with skill in the art, and the fragments are screened for utility in the same manner as are intact antibodies. Antigen-binding portions can be produced by recombinant DNA techniques, or by enzymatic or chemical cleavage of intact .immu noglobulins. Antibodies can be of different isotype, for example, an IgG (e.g., an IgGI, IgG2,- IgG3, or IgG4 sub type)., IgA1, IgA2, IgD, IgE, or 1gM antibody. The term "frameworks" refers to the art recog nized portions of an antibody variable region that exist between the more divergent CDR regions. .Such framework regions are typically referred to as frameworks 1 through 4 (FRI, FR2, FR3, and FR4) and provide a scaffold for holding, in three-dimensional space, the three CDRs found in a heavy or light chain antibody variable region, such that the CDRs can form an antigen-binding surface. Such frameworks can also be referred to as scaffolds as they provide support for the presentation of the more diver gent CDRs.. Other CDRs and frameworks of the immunoglobu lin superfamily, such as ankyrin repeats and fibronectin, WO 2007/124610 PCT/CH2007/000202 23 can be used as antigen binding molecules (see also, for example, U.S. Patent Nos. 6,300,064, 6,815,540 and U.S. Pub. No. 20040132028). The term "epitope" or "antigenic determinant" refers to a site on' an antigen to which an immunoglobulin or antibody specifically binds (e.g., ALK, for example, amino acid residues 391-406 of human ALK-1 (see e.g., SEQ -ID NO: 1) . An epitope typically includes at least 3, 4, 5,'6, 7, 8, 9, 10, 11, 12, 13, 14 or 15 amino acids in a unique spatial conformation. See, e.g., Epitope Mapping Protocols in Methods in Molecular Biology, Vol. 66, d. E. Morris, Ed. (1996). The terms "specific binding," "selective bind ing," "selectively binds," and "specifically binds," re fer to antibody binding to an epitope on a predetermined antigen. Typically, the antibody binds with an affinity (KD) of approximately less than 10~7 M, such as approxi mately less than 10 ~ M, 10- M or 10-* M or even lower The term "K," refers to the dissociation equilibrium constant of a particular antibody-antigen in teraction. * Typically, the antibodies of the invention bind to ALK' with a dissociation equilibrium constant (KD) of less than approximately 10-" m, such as less than ap proximately 10~6 M, 10- M or 10~1* M or even lower, for example, -as. determined using surface plasmon resonance (SPR) technology in a BIACORE instrument. The terms "neutralizes ALK," "inhibits ALK," and "blocks ALK" are used interchangeably to refer to the ability of an antibody of the invention to prevent ALK from interacting with one or more target ligands and, for example, triggering signal transduction. The term "nucleic acid molecule," refers to -DNA molecules and RNA molecules. A nucleic acid molecule WO 2007/124610 PCT/CH2007/000202 24 may be single-stranded or double-stranded, but preferably is double-stranded DNA. A nucleic acid is "operably linked" when it is placed into a functional relationship with another nucleic acid sequence. For instance, a pro.

moter or enhancer is operably- linked to a coding sequence if it affects the transcription of the sequence. For nucleic acids, the term "substantial ho mology" indicates that two nucleic acids, or designated sequences thereof, when optimally aligned and compared, are identical, with appropriate nucleotide insertions or deletions, in at least about BO% of the nucleotides, usu ally at least about 90% to 95%, and more preferably at least about 98% to 99.5% of the nucleotides. Alterna tively, substantial homology exists when the segments' will hybridize under selective hybridization conditions, to the complement of' the strand. Such hybridization con ditions are know in the art, and described, e.g., in Sam book et al. infra. The percent identity between. two sequences is a function of the number of identical positions shared by the sequences, taking into account the number of gaps, and the length of each gap, which need to be introduced for optimal alignment of the two sequences. The compari son of sequences and determination of percent identity between two sequences can be accomplished using a mathe natical algorithm, as described in' the non-limiting exam ples below. The percent identity between two nucleotide sequences can be determined using the GAP program in the GCG software package, using a NWSgapdna. CMP matrix and a gap weight of 40, 50, 60,.70, or 80 and a length weight of 1, 2, 3, 4,. 5, or 6. The percent identity between two nucleotide or amino acid sequences can also be determined WO 2007/124610 PCT/CH2007/000202 25 using the algorithm of E. Meyers and W. Miller (CABIOS, 4:11-17 (1989)) which has been incorporated into the ALIGN program (version 2.0), using a PAM120 weight resi due table, a gap length penalty of 12 and a gap penalty of 4. In addition, the percent identity between two amino acid sequences can be determined using the Needle man and Wunsch (J. Mol. .Biol. (48):444-453 (1970)) algo rithm which has been incorporated into the GAP program in the GCG software package, using either a Blossum 62 ma trix or'a PAM250 matrix, and a gap weight of 16, 14, 12, 10, 8, 6, or 4 and a length weight of 1, 2, 3, 4, 5, or The nucleic acid and protein sequences of the present invention can further be used as a "query se quence" to .perform a search against public databases to, for example, identify related sequences. Such searches can be performed using the NBLAST and XBLAST programs (version 2.0) of Altschul, et al. (1990) J. Mol. Biol. 225:403-10. BLAST nucleotide searches can be performed with the NBLAST program, score = 100, wordlength =. 12 to obtain nucleotide sequences homologous to the nucleic acid molecules of the invention. BLAST protein searches can be performed with the XBLAST program, score = 50, wordlength = 3 to obtain amino acid sequences homologous to the protein molecules of the invention. To obtain gapped alignments for comparison purposes, Gapped BLAST can be utilized as described in Altschul et al., (i997) Nucleic Acids Res. 25(17) :3389-3402. When utilizing BLAST and Gapped BLAST programs, the default parameters of the respective programs (e.g., XBLAST and NBLAST) can be used. The present invention also encompasses "con servative sequence modifications" of the sequences set WO 2007/124610 PCT/CH2007/000202 26 forth in the SEQ ID NOs of the present invention, i.e., nucleotide and amino acid sequence modifications which do not abrogate the binding of the antibody encoded by the nucleotide sequence or containing the amino acid se quence, to the antigen. Such conservative sequence modi fications include nucleotide and amino acid substitu tions, additions and deletions. For example, modifica tions can be introduced by standard techniques known in the art, such as site-directed mutagenesis and PCR mediated mutagenesis. Conservative amino acid substitu tions include ones in which the amino acid residue is re placed with an amino acid residue having a similar side chain.. Families of amino acid residues having similar side chains have been defined in the art. These families include amino acids with basic side chains (e.g., lysine, arginine, histidine), acidic side chains (e.g.., aspartic acid, glutamic acid), uncharged polar side chains (e.g., glycine, asparagine, glutamine, serine,. threonine, tyro sine, cysteine, tryptophan), nonpolar side chains (e.g., alanine, valine, ..leucine, isoleucine, proline,' phenyla lanine, methionine), beta-branched side chains (e.g., threonine,. valine, isoleucine) and aromatic side chains (e.g., tyrosine, phenylalanine, tryptophan, histidine). Thus, a predicted nonessential amino acid residue in a human anti-ALK antibody is preferably replaced with an other amino acid residue from the same side chain family. Methods of identifying nucleotide and amino acid conser vative substitutions which do not eliminate antigen bind ing are well-known in the art (see, e.g., Brummell et al., Biochem. 32:1180-1187 (1993); Kobayashi et al. Protein Eng. 12(10):879-884 (1999); and Burks et al. Proc. Natl. Acad. Sci. USA 94:412-417 (1997)) WO 2007/124610 PCT/CH2007/000202 27 Alternatively, in another embodiment, muta tions are randomly introduced along all or part of an anti-ALK antibody coding sequence, such as by saturation mutagenesis, and the resulting modified anti-ALK antibod ies can be screened for binding activity. A "consensus sequence" is a sequence formed from the most frequently occurring amino acids (or nucleotides) in a family of re lated sequences (See e.g., Winnaker, From Genes to Clones (Verlagsgesellschaft, Weinheim, Germany 1987). In a fam ily of proteins, each position in the consensus sequence is occupied by the amino acid occurring most frequently at that position in the .family. If two amino -acids occur equally frequently, either can be included in the consen sus sequence. By reference to the tables and figures pro vided herein a consensus sequence for the antibody heavy / light chain variable region CDR(s) can be derived by optimal alignment of the amino acid sequences of the variable region CDRs of the antihdies which are reactive against epitope 390-406 of the human ALK-1 protein. The term. "vector," refers to a nucleic acid molecule capable of transporting another nucleic. acid to which it has been linked. . One type of vector is a "plas mid," which refers to a circular double stranded DNA loop into which additional DNA segments may be ligated. An other type of vector is a viral vector, wherein addi tional DNA segments may be ligated into the viral genome. Certain vectors are capable of autonomous replication in a host cell into which they are introduced (e.g., bacte rial vectors having a bacterial origin of replication and episomal mammalian vectors). Other vectors (e.g., non episomal mammalian vectors) can be integrated into the genome of a host cell upon introduction into the host WO 2007/124610 PCT/CH2007/000202 28 cell, and thereby are replicated along with the host ge nome. The term "host cell" refers.to a cell into which and expression vector has been introduced. Host cells can include bacterial, microbial, plant or animal cells. Bacteria, which are susceptible to transforma tion, include members of the enterobacteriaceae, such as strains of Escherichia coll or Salmonella; Bacillaceae, such as 'Bacillus subtilis;. Pneumococcus; Streptococcus, and Haemophilus influenzae. Suitable microbes include Saccharomyces cerevisiae and P-ichia pastors. Suitable animal host cell lines include CHO (Chinese Hamster Ovary lines) and NSO cells. The terms "treat," "treating, " and- "tr$at ment," refer to therapeutic or preventative measures de scribed herein. The methods of "treatment" employ ad ministration to a subject,* in need of such treatment, an antibody of. the present invention, for example, a subject having an ALK-mediated disorder or a subject who ulti mately may acquire such a disorder, in order to prevent, cure, delay, reduce the severity of, or ameliorate one or more symptoms of the disorder or recurring disorder, or in order to prolong the survival of a subject beyond that expected in the absence of such treatment. The term "ALK-mediated disorder" refers to disease states and/dr symptoms associated with ALK mediated cancers or tumors. In general, the term "ALK mediated disorder" refers to any disorder, the onset, progression or the persistence of the symptoms of which requires the participation of ALK. Exemplary ALK mediated disorders include, but are not limited to, for example, cancer, in -particular, glioblastoma.

WO 2007/124610 PCT/CH2007/000202 29 The term "effective dose" or "effective dos age" refers to an amount sufficient to achieve or at least partially achieve the desired effect. The term therapeuticallyy effective dose" is defined as an amount sufficient to cure or at least partially arrest the dis ease and its complications in a patient already suffering from the disease. Amounts effective for this use will depend upon the severity of the disorder being treated and the general state of the patient's own immune system. The term "subject" refers to any human or non human animal. For example, the methods and compositions of the present invention can be used to treat a subject with a cancer, e.g., glioblastoma. Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. Although methods and materials similar or equivalent to those described herein can be used in the practice or testing of the present invention, suitable methods and materials are described below. In case of conflict, the present specification, including definitions, will control. In addition, the materials, methods, and examples are illustrative only and not intended to be limiting. Various aspects of the invention are described in further detail in the following subsections. It is understood that the various embodiments, preferences and ranges may be combined at will. Further, depending of the specific embodiment, selected definitions, embodiments or ranges may not apply.

WO 2007/124610 PCT/CH2007/000202 30 The present invention provides in a first as pect an antibody binding the human ALK protein, said an tibody comprising a variable heavy chain CDR3 of a se quence with at least 50% sequence identity to the se quence SEQ. ID. No. 2. Preferably, the sequence identity is at least 60%, 70%, 75%, 80%, or more preferably at least 90%. Most preferably, the CDR3 has the precise se quence of SEQ. ID. No. 2. In .a preferred embodiment of the present in vention the antibody binds specifically to the human ALK protein, i.e. it does not bind to the mouse ALK protein, whose PTN binding site differs in only 2 amino acid resi dues compared to the PTN binding site (SEQ. ID. No. 1) of the human ALK protein. The human isoleucine at position 3 is a valine and the aspartate is an alanine in the cor responding mouse sequence. The antibody of the present invention can be a full-length antibody, but also an antibody fragment, such as for example a scFv or a Fab fragment. Antigen binding fragments are well known in the art. Preferably, a scFv antibody is used.. The heavy chain and the light chain are com posed of framework sequences, each comprising three CDRs, CDR1, CDR2 and CDR3,. which are predominantly involved in antigen binding. The antibody of the present invention comprises VH domain of the H3 type and a VL domain of the lambdal type. The VE and VL framework of the antibody of the present invention is stable and soluble so as to be functional in an intracellular reducing environment. Preferably it is the framework 4.4. that has previously been isolated by a yeast screening system referred to 'as the "Quality control system" (Auf der Maur et al., 2001; WO 2007/124610 PCT/CH2007/000202 31 Auf der Maur et al. 2004). The sequence of the framework can be deduced for example from SEQ. ID. No. 20 (see be low), where the framework portions are represented by non-underiined and straight letters, while the CDR se quences are underlined and the linker sequence is in italics. The antibody of the present invention is able to bind a 16-amino acid ALK epitope peptide of a sequence that's at least 75%, preferably 80%, 85%, 90%, 95%, or most preferably 100%, identical to the.sequence of SEQ. ID. No. 1. This sequence is also referred to as PTN binding site and is a unique sequence in the entire human genome. The corresponding mouse sequence varies in-2 out of 16 amino acids, i.e., V at position 3 and A at posi tion 7 instead of I or D, respectively. Preferably, the antibody of the present invention binds human but not mouse ALK, i.e., is specific for the human protein. The ALK epitope comprising about residues 391-406 or consist ing of these residues is uniquely suited for selecting an antibody or antigen binding fragment that can specifi cally bind ALK and block or inhibit ALK-mediated activ ity. This epitope is also suitable for screening or raising antibodies that specifically block-ALK activity. Thus, this epitope, especially an epitope comprising about residues 391-406 or consisting of about these resi dues, is uniquely suited for use as an active immuno therapeutic agent or vaccine as further described herein. The antibody of the present invention has an affinity for the ALK epitope peptide with a Kd.of 30 nM or less, preferably 10 nM or less, most preferably below 3 nM. In another embodiment of the present inven tion,. the antibody comprising a variable light chain CDR3 WO 2007/124610 PCT/CH2007/000202 32 of a sequence with at least 50% sequence identity to the sequence SEQ.ID. No. 3. Preferably, the sequence iden tity is at least 60%, 70%, 80%, 85%, more preferably at least 90%. Most preferably, CDR3 is identical to SEQ. ID. No. 3. Again, this antibody binds a 16-amino-acid ALK epitope peptide of a sequence with at least 75%, prefera bly 80%, 85%, 90%, 95%, or most preferably 100%, amino acid identity to the sequence SEQ.ID. No. 1. Also, the antibody has an affinity for the ALK epitope peptide with a Kd of less than 10 nM, preferably less than 7nM. -In a. preferred embodiment of the present in vention the antibody comprises a VH sequence of SEQ.ID. No. 4 and a VL sequence of SEQ.ID, No. 5. Additionally, it can comprise at least one mutation in at least one of the CDRs resulting in a higher affinity characterized by a Kd of. less than about.3 nM. Said at least one mutation is preferably in CDR1 or CDR2 of VH and/or VL, most pref erably in CDR2 of VH. In another preferred embodiment of the pre sent invention the antibody comprises a variable heavy chain CDR2 comprising a sequence selected from the group of SEQ. ID. No. 7, SEQ.ID, No. 8, SEQ. ID. No. 9, SEQ.ID. No. 10, SEQ'.ID. No. 11, SEQ.ID. No. 12, or SEQ.ID. No. 13. Preferably these defined CDR2 sequences are preceded by the amino acids residues Al and followed by the se quences of SEQ.ID. No. 17, so that the entire CDR2 is de fined. A preferred antibody of the present invention comprises a VH sequence. of SEQ. ID. No. 4 and a VL se quence of SEQ. ID. No. 5. In scFv antibodies, the domain structure can NH 2 -VL-linker-VH-COOH or NH 2 -VH-linker-VL-. COOH; preferably, the liner has the sequence SEQ. ID. No. 16. Alternatively, the variable regions represented WO 2007/124610 PCT/CH2007/000202 33 by SEQ ID NOS: 4 and 5 can be engineered into a full length antibody, e.g., IgG or IgM. Constant regions suitable for combining with the variable regions of the invention are known in the art. Within the scope of the present invention is the use of the antibody or antibody derivative as a me dicament or as a diagnostic tool. Preferably, the produc-. tion of a medicament for the treatment of cancers or tu mors is envisaged. For this purpose an antibody can be radiolabelled using radionuclides or radiometal labeling. This is particularly valuable for tumor targeting, imag ing and biodistribution studies. Also, recombinant DNA technology makes it possible to genetically fuse coding regions of variable V genes to modified toxin domains. For example, a scFv-toxin' fusion wherein the scFv is spe cific for a tumor marker protein can target the toxin to the tumor, where the toxin causes cytotoxicity. Such targeted therapy results in the selective concentration of cytotoxic agents or radionuclides in tumors and should lessen the toxicity to normal tissues. In a preferred embodiment of the present in vention the antibody is used for a treatment of cancers or tumors, preferably neuroblastoma, glioblastoma, rhab domyosarcoma, breast carcinoma, melanoma,' pancreatic can cer, B-cell non-Hodgkin's lymphoma, thyroid carcinoma, small cell lung carcinoma, retinoblastoma, Ewing sarcoma, prostate cancer, colon cancer, or pancreatic cancer, preferably glioblastoma, neuroblastoma and rhabdomyosar coma. ALK expression and protein has been detected in many soft tissue tumors (Li et al., 2004). Full-length ALK has been found in these human tumors.'Furthermore, the antibody is preferably used for local treatments. Most preferred is local treatment of glioblastoma.

WO 2007/124610 PCT/CH2007/000202 34 Another aspect of the present invention is to provide a DNA sequence encoding the antibody of the pre sent invention. A suitable prokaryotic expression vector for ESBA512 (SEQ.ID.No: 19) is pTFT74 (see SEQ.ID.No: 90 for the sequence including the ESBA512 coding sequence). Therein, the ESBA512 coding sequence is under the control of the T7-promotor and the recombinant gene product is usually purified over inclusion bodies. Another preferred prokaryotic expression vector is pAK400, wherein the ESBA512 sequence is his-tagged for simplified purifica tion (see SEQ.ID.No: 89 for the sequence including the ESBA512 coding sequence). The gene product is secreted by the host cell into the periplasm. In addition, an expression vector comprising said DNA sequence and a suitable host cell transformed with said expression vector is provided. Preferably, said host cell is an E. coli cell. Yet another aspect of the present invention is the production of the antibody 6f the present inven tion, comprising culturing the host cell that is trans formed with the expression vector for said antibody, un der conditions that allow the synthesis of said antibody and recovering it from said culture. Another aspect of the present invention is to provide an ALK epitope, comprising or consisting essen tially of residues 391-406 of SEQ ID NO: 1. Said epitope is suitable for identifying, screening, or raising anti ALK antibodies or fragments thereof. Preferably, the an tibody or antigen binding fragment thereof that is capa ble of specifically binding residues 391-406 (SEQ ID NO:1) of an isolated ALK protein or fragment thereof. More preferably, the antibody is- a single chain antibody (scFv), Fab frag-ment, IgG, or IgM.

WO 2007/124610 PCT/CH2007/000202 35 In a further aspect, an ALK vaccine compris ing an isolated ALK protein or a fragment thereof, or a nucleic acid encoding an epitope of ALK is provided, Preferably, the vaccine comprises residues 391-406 of an isolated ALK protein. Said vaccine is preferably formu lated with a carrier, adjuvant, and/or hapten to enhance the immune response. The sequences of the present invention' are the following ones: SEQ.ID. No. 1: GRIGRPDNPFRVALEY Human ALK epitope peptide (amino acids 391 406 of the ALK protein); underlined residues are differ ent in the mouse homologue. SEQ.ID. No. 2: RDAWLDVLSDGFDY ESBA521 CDR 3 of VH. Residues obtained after randomization are underlined. SEQ.ID. No. 3: ATWDNDKWGVV ESBAS21 CDR 3 of VL. Residues obtained after randomization are underlined. SEQ.ID. No. 4: EVQLVESGGGLVQPGGSLRLSCAASGFTFSSYAMSWVRQAPGKGLEWVSAISGSGGS TYYADSVKGRFTISRDNSKNTLYLQMNSLRAEDTAVYYCARDAWLDVLSDGFDYWGQ GTLVTVSS VH of ESEAS21. CDRs are underlined. SEQ.ID. No. 5: QSVLTQPPSVSAAPGQKVTISCSGSTSNIGDNYVSWYQQLPGTAPQLLIYDNTKRPS

GIPDRFSGSKSGTSATLGITGLQTGDEADYYCATWDNDKWGVVFGGGTKLEVLG

WO 2007/124610 PCT/CH2007/000202 '36 VL of ESBA521. CDRs are underlined. SEQ.ID. No. 6: AISGSGGSTYYADSVKG VH CDR 2 of ESBAS21 SEQ.ID. No. 7: AINMKGNDRYYADSVGK VH'CDR 2 of scFv 265.1 SEQ.ID. No. 8: AIRTNSKEYYADSVKG VH CDR 2 of scFv 43.2 SEQ.ID. No. 9 AIKTDGNHKYYADSVKG VHI CDR 2 of scFv 100.2 SEQ.ID. No. 10: RTDSIEQYYADSVKG VH CDR 2 of scPv 2.11 SEQ.ID. No. 11: ETSSGSTYYADSVKG' VH CDR 2 of sc'v 28. 11 SEQ.ID. No. 12: 'NTGGGSTYYADSVICG VH CDR 2 of scFv 33.11 SEQ.ID. No. 13: NTRGQNEYYADSVKG VH CDR 2 of scFv 4.12 WO 2007/124610 PCT/CH2007/000202 37 SEQ.ID. No. 16 GGGGS GGGGSGGGGSS GGGS Linker connecting VL and VH SEQID. No. 17 YYADSVKG C-terminal half of CDR2 of ESBA521 and its derivatives. SEQ.ID. No. 18 DAGIAVAGTGFDY VH CDR3 of FW4.4 SEQ.ID. No. 19 QSVLTQPPSVSAAPGQKVTISCSGSTSNIGDNYVSWYQQLPGTAPQLLIYDNTKRPS GIPDRFSGSKSGTSATLGITGLQTGDEADYYCATWDNDKWGVVFGGGTKLEVLGGGG GSGGGGSGGGGSSGGSEVQLVESGGGLVQPGGSLRLSCAASGFTFSSYAiSWVRQA PGKGLEWVSAISGSGGSTYYADSVKGRFTISRDNSKNTLYLQMNSLRAEDTAVYYCA RDAWLDVLSDGFDYWGQGTLVTVSS scFv ESBA521, CDRs underlined; linker in italics SEQ. ID. No. 20 QSVLTQPPSVSAAPGQKVTISCSGSTSNIGDNYVSYQQLPGTAPQLLIYDNTKRPS GIPDRFSGSKSGTSATLGITGLQTGDEADYYCGTWDSSLSGVVPGGGTKTVLGGG GSGGGGSGGGGSSGCGSEVQLVESGGGLVQPGGSLRLSCAASGFTFSSYAMSWVRQA PGKGLEWVSAISGSGGSTYYADSVKGRFTISRDNSIKNTLYLQMNSLRAEDTAVYYCA RDAGIAVAGTGFDYWGQGTLVTVSS FW4.4, CDRs are underlined, linker in italics SEQ ID No. 21 WO 2007/124610 PCT/CH2007/000202 38 cagtctgtgctgacgcagccgccctcagtgtctgcggccccagga cagaaggtcaccatctcctgct ccggaagcacctccaacattggcgataattatgta t cctggtaccaacaactcccaggaacagccccccaactcctcatttatgacaatact aaacgaccctcagggattcctgaccggttctctggctccaagtctggcacgtcagcc accctgggcatcaccggactccagactggggacgaggccgattattactgcgcgacc tgggataatgataagtggggtgtggttttcggcggagggaccaagctcgaggtcc taggt Nucleic acid sequence- of ESBAS21 VL CDRs are underlined SEQ ID No. 22 gaggtgcagctggtggagtccgggggaggcttggtacagcctggg gggtccctgagactctcctgtgcagcctctggattcacctttagcactatgccatg agctgggt ccgccaggctccagggaaggggctggagtgggtctcagctattagtggt agtggtggtagcacatactacgcagactccgtgaagggccggttcaccatctccaga gacaattccaagaacacgctgtatctgaaaatgaacagcctgagagccgaggacacg gccgtatattactgcgcgcgtgatgcgtggttggatgtgcttticggatggetttgac tactggggccagggaaccctggtcaccgtctcctcg Nucleic acid sequence of ESBA521 VH CDRs are underlined SEQ ID No. 23 cagtctgtgctgacgcagccgccctcagtgtctgcggccccagga cagaaggtcaccatctcctgctccggaagcacctccaacattggcgataattatgta tcctggtaccaacaactcccaggaacagccccccaactcctcatttatgacaatact aaacgaccctcagggattcctgaccggttctctggctccaagtctggcacgtcagcc accctgggcatcaccggactccagactggggacgaggccgattattactgcgcgacc tgggataatgataagtggggtgtggttttcggcggagggaccaagctcgaggtccta ggtggtggtggtggt tatggtggtggtgttctggcgcggcgge tccatggtggt ggatccgaggtgcagctggtggagtccgggggaggcttggtacagcctggggggtcc ctgagactctcctgtgcagcctctggattcacctttagcagctatgccatgagetgg gtccgccaggctccagggaaggggctggagtgggtctcagctattagtggtagtggt ggtagcacatactacgcagactccgtgaaqqqccggttcaccatctccagagacaat WO 2007/124610 PCT/CH2007/000202 39 tccaagaacacgctgtatctgcaaatgaacagcctgagagccgaggacacggccgta tattactgcgcgcgtgatgcgtggttggatgtgctttcggatggctttgactactgg ggccagggaaccctggtcaccgtctcctcg Nucleic acid sequence of ESBAS21 CDRs are underlined, linker in italics The invention is now further described by means of examples: Materials and Methods In general, the practice of the present invention employs, unless otherwise indicated, conventional. tech niques of chemistry, molecular biology, recombinant DNA technology, immunology (especially, e.g., antibody tech nology) , and standard techniques in polypeptide prepara tion. See, e.g., Sambrook, Fritsch and Maniatis, Molecu lar Cloning: Cold Spring Harbor Laboratoy 'Press (1989); Antibody Engineering Protocols (Methods in Molecular Bi ology), 510, Paul, S., Humana Pr' (1996); Antibody Engi neering: .A Practical Approach (Practical Approach Series, 169), McCafferty, Ed., Irl Pr (1996); Antibodies: A' Labo ratory Manual, Harlow et al., C.S.H.L. Press, Pub. (1999); and Current Protocols in Molecular Biology, eds. Ausubel et al., John Wiley r Sons (1992). Experiment 1: Screening to identify Alk binding scPvs In a wealth of structural studies on anti body-antigen interactions it was found that residues in the complementarity-determining region 3 (CDR-H3) of the heavy chain generally contribute the most substantial contacts to the antigen (Chothia and Lesk, 1987; Chothia WO 2007/124610 PCT/CH2007/000202 40 et al., 1985; Padlan, 1994). We applied our recently de scribed yeast two-hybrid antigen-antibody interaction screening technology to directly isolate antigen-binding scFvs by screening of four scFv libraries of randomized synthetic CDR-HI3 sequences(Auf der Maur et al., 2002). The four libraries are based on four different stable hu man scFv frameworks in which 7 amino acids within the third CDR loop of the variable heavy chain (VH-CDR3) were randomized. The randomized parts were introduced by stan dard PCR cloning techniques. The scFv libraries were screened against a 16 amino acid peptide derived from the extracellular domain of the human tyrosine receptor -kinase Anaplastic Lymphoma Kinase (ALK) by a yeast screening system called "Quality Control". (Auf der Maur et al., 2001; Auf der Maur et al., 2004). Briefly, the Quality Control technology is an antigen-independent in trabody selection system for identifying from a natural pool of human variable-light (VL) and. variable-heavy (VA) chains those VL and VH combinations with favourable bio physical properties, such as- stability, solubility and expression yield. One promising and specific binder from one of the four scFv libraries was isolated after the first screening round. This particular scFv was derived from the framework FW4.4 library. FW4.4 (SEQ. ID. No. 20) consists of a VL domain (lambdal) connected by a classi cal flexible glycine-serine linker (CGGGS) 4 to a VH 3 do main. The VE CDR3 of FW4,4 comprises 13 amino acids (DAGIAVAGTGFDY; SEQ.ID. No. 18). To construct the li brary, the central part of the VH CDR3 (DAXXXXXXXGFDY) was randomized by standard PCR-cloning methods using a degenerated oligonucleotide. The last two residues (Asp and Tyr) were kept constant, because their structural im portance was demonstrated in many cases (Chothia and WO 2007/124610 PCT/CH2007/000202 41 Lesk, 1987) The remaining residues were not modified in order to keep the complexity of the library in manageable dimensions. The scFv library was cloned in' a yeast ex pression vector (pLibi) as C-terminal fusion to the tran scriptional activation domain of Gal4 (Auf der Maur et al., 2002). The ligand-binding domain (LBD) of human Alk was chosen as antigen for the interaction screen (Stoica et al., 2001) This 16 amino acid peptide was cloned into another yeast expression vector (pBait1) as C-terminal fusion to the DNA-binding protein LexA. The reporter yeast strain TfDE173 (Auf der Msur et al., 2002) containing the stably integrated re porter genes HIS3 and lacZ under the control of six LexA binding sites was transformed with the bait vector ex pressing the Alk LBD fused to LexA together with the ran dom CDR-H3 scFv library fused to the.Gal4 activation do main. Transformed cells were selected on plates lacking histidine and containing 2.5 mM 3-amino-triazole (3-ATY, which is a competitive inhibitor of the HIS3 gene prod uct.'Growing colonies were picked over a period of six days and the library plasmids were isolated. The same re porter strain was transformed with the rescued plasmids to confirm antigen-dependent gene activation. A quantita tive liquid P-galactosidase assay was performed to meas ure binding-strength between the Alk LBD,. i.e. the 16 amino acid ALK peptide, and the selected scFv. The scFv with highest reporter gene activation also demonstrated best affinity (-31nM) for the Alk LBD peptide in ELISA (data not shown). The sequences of other VH CDR3 sequences identified as contributing to ALK binding are provided below in Table la.

WO 2007/124610 PCT/CH2007/000202 clone VH CDR 3 (mutated residues) WT FW4.4 DAGIAVAGTGFDY (SEQ. ID. No. 24) H5 SH 2.1 DAKFMSDGIGFDY (SEQ. ID. No. 25) HS SH 4.1 DAWGWTILSGFDY (SEQ. ID. No. 26) H5 SH S .1 DAAYMIRYEGFDY (SEQ. ID. No. 27) H5 SH 2.3 DAWIYWAREGFDY (SEQ. ID. No. 28) H5 SH 3.3 DACMTYSREGFDY (SEQ. ID. No. 29) H5 SH 5.3 DAWLDVLSDGFDY (SEQ. ID. No. 30) H5 SH 14.3 DAPSVNDREGFDY (SEQ. ID. No. 31) The sequences of other suitable frameworks are provided below in Table 1b. FW Secuence 5.2 EIVLTQSPATLSLSPGERATLSCRASQTLTHYLAWYQQKPGQAPRLLIYD TSKRATGTPARFSGSGSGTDFTLTISSLEPEDSALYYCQQRNSWPHTFGG GTKLEIKRGGGGGSGGGGSGGGGSGGSEVQLVESGGGVAQPGGSLRVS' CAASGFSFSSYAMQWVRQAPGKGLEWVAVISNDGRIEHYADAVRGRFTIS RDNSQNTVFLQMNSLRSDDTALYYCAREIGATGYLDNWGQGTLVTVSS .(SEQ. ID. No. 15) Experiment 2: Affinity maturation In order to obtain an scFv with higher affinity, this primary binder was subjected to a further affinity matu ration process by mutagenesis and a second screening round in yeast. Enabling affinity maturation, the expres sion level of the LexA Alk LBD peptide fusion protein was reduced in order to lessen reporter gene activation driven by' the interaction of the primary binder with the Alk LED peptide. The strong actin promoter on the pBaitl vector was exchanged with the truncated and thus less ac tive version of the ADH promoter (alcohol dehydrogenase) resulting in pBait3. This reduction of the bait expres sion level, in the presence of the primary binder, was sufficient to inhibit growth on plates lacking histidine and containing 5 mIM 3-AT. Mutagenesis of the primary WO 2007/124610 PCT/CH2007/000202 43 binder for affinity maturation was accomplished by ran domizing parts of the CDR3 within the variable light chain. This was performed directly in yeast by homologous recombination (Schaerer-Brodbeck and Barberis, 2004). The VL CDR3 of FW4.4 comprises 11 amino acids (SEQ. 7D. No. 14: GTWDSSLSGVV) . The first two positions were partially randomized, such that the first position either encodes Gly, Ala or Gln, and the second position Thr, Ser or Ala. At the positions 5 to 8 within VL CDR3 all amino acid residues were allowed. The remaining positions were kept constant. Randomization was introduced by PCR. The re sulting PCR product had a size of 356 bp and comprised the randomized CDR cassette with 267 bp upstream and 27 bp downstream framework sequences. This product is the so-called donor -PCR fragment, which bears homologies to the target vector. The target *vector is the yeast plasmid (pLibl) encoding the primary binder fused to the activa tion domain of Gal4. In order to improve efficiency of homologous recombination and to exclude false positives in the subsequent. screening, the CDR-L3 in the target vector wias slightly modified. A unique SacI restriction site was introduced in the middle of VL CDR3, which leads to a frameshift in the scFv encoding part of the fusion protein and results in a truncated protein unable to bind to the Alk LBD. In addition, the SacI site enable's lin earization of the target vector, which enhances recombi nation efficiency in yeast. The screening was launched by pre transformation of the reporter yeast strain YDE173 with the plasmid (pBait3) expressing the LexA Alk LBD from the truncated ADH promoter. This pre-transformed yeast cells were made competent again and co-transformed with the linearized target vector and with the donor PCR fragment, WO 2007/124610 PCT/CH2007/000202 44 which bears homologies upstream and downstream of the VL CDR3. Upon homologous recombination between the PCR prod uct and the target vector, the novel VL CDR3 sequence is integrated into the corresponding site of the target vec tor. As a net result of this event the primordial VL CDR3 gets exchanged with the randomized VL CDR3. This allows reconstitution of a circular plasmid that expresses a fully functional fusion protein with a novel VL CDR3 se quence, which will .activate reporter gene expression and enable growth on selective plates upon interaction with the Alk LBD peptide. A total of 119 clones grew on selective plates over an observation period of 6 days. 'These clones were picked and the library plasmids were isolated and retransformed into the same reporter yeast strain. A quantitative liquid P-galactosidase assay was performed to measure binding strength between the Alk LBD (antigen) and the affinity-matured scFv. 20 clones with highest lacZ activation were also tested in ELISA with Alk LBD peptide. The best candidate revealed a KD.of about 7nM (Fig. 3) and was named ESBA521. The sequences of other VL CDR3 sequences identified as contributing to ALK binding are provided below in Table ic. clone VL CDR 3 WT FW4.4 GTWDSSLSGVV (SEQ. ID. No. 32) 5.3-9.1 AAWDSV1HGVV (SEQ. ID. No. 33) 5.3-21.1 AAWDNSMGVV (SEQ. ID. No. 34) 5.3-22.1 AAWDTMRYGVV (SEQ. ID. No.35) 5.3-25.1 AAWDTTRVGVV (SEQ. ID. No. 36) 5.3-27.1 ASWDTMLKGVV (SEQ. ID. No. 37) 5.3-28.1 ASWDTPTCGVV (SEQ. ID. No. 38). 5.3-29.1 ATWDISRCGVV (SEQ. ID. No. 39) 5.3-46.1 ATWDTVCAGVV (SEQ. ID. No. 40) 5.3-53.1 ATWDVDVFGVV (SEQ. ID. No. 41) WO 2007/124610 PCT/CH2007/000202 45 5.3-57.1 ATWDDVVGGVV (SEQ. ID. No. 42) 5.3-86.1 AAWDSFYNGVV (SEQ. ID. No. 43) 5.3-94.1 ASWDTLIEGVV (SEQ. ID. No. 44) 5.3-J-07.1 ATWDNDKWGVV (SEQ. ID. No. 45) 5.3-112.1 AAWDSTTCGVV (SEQ. ID. No. 46) 5.3-113.1 ATWDMWMKGVV (SEQ. ID. No . 47) 5.3-117.1 GTWDSSLSGVV (SEQ. ID. No. 48) 5.3-118.1 AAWDWVLGGVV (SEQ. ID. No. 49) 14.3-6.1 ATWDNPGQGVV (SEQ. ID. No. 50) 14.3-7.1 ATWDDWVIGVV (SEQ. ID. No. 51) 14.3-8.1 ASWDDQIKWGVV (SEQ. ID. No. 52) 14.3-9.1 ATWDTNREGVV (SEQ. ID. No. 53) 14.3-12.1 ASWDDLHIGVV (SEQ. ID. No. 54) 14.3-13.1 ASWDEEAWGVV (SEQ. ID. No. 55) 14.3-21.1 ATWDYIKIGVV (SEQ. ID. No. 56) 14.3-4B.1 ATWDTFERGVV (SEQ. ID. No. 57) 14.3-49.1 ATWDSNLIGVV (SEQ. ID. No. 58) 5.3-24.,1 ATWDNNTCGVV (SEQ. ID. No. 59) 5.3-3.1 AAWDCDINGVV (SEQ. ID. No. 60) 5.3-8.1 ASWDSMKIGVV (SEQ. ID. No. 61) 5.3-19.1 ATWDCTRAGVV (SEQ. ID. No. 62) Experiment 3: The scFv ESBA521 specifically binds to the transmembrane form of human ALK In order to test whether the newly identified scFv was able to also recognize the transmembrane human ALK protein on the surface of living cells, immunostain ing experiments of transiently transfected HELA cells were performed. ESBAS21 reacts with .the ALK protein in a comparable way as a polyclonal antibody (Fig. 4) . In a control experiment it was shown that the 'framework 4 .4 scPv does not react with human ALK. Surprisingly, ESBA521 only binds to the human Alk protein, but not to the cor responding mouse protein, although the mouse antigenic peptide only differs in two amino acid positions from the human peptide sequence. By contrast, the polyclonal ALK antibody recognizes both human and mouse protein. There fore, binding of ESBA521 is specific for the human ALK protein at the cell surface.

WO 2007/124610 PCT/CH2007/000202 46 Experiment 4: Isolation of improved binders by PCR mutagenesis of VH CDR 2 To further improve antigen binding, ESBA521 was used as the starting ecFv in a further round of af finity maturation using the same two-hybrid approach as described for the first round of affinity maturation, ex cept in this case CDR2 of VH was changed by PCR mutagene sis and transformed into the yeast recipient, wherein ho mologous recombination at CDR2 is enforced in analogous way. Again, a restriction site was introduced in CDR2 to enable linearization of the target plasmid. The mutations introduced in CDR2 are given in Table 2. scFv (performing VH CDR 2 (mutated residues) best) WT (ESBA521) AISGSGGSTYYADSVKG (SEQ. ID. No. 63) 1.1 AI-KTDGQNYYADSVKG (SEQ. ID. No. 64) 17.1 AIRSDGNERYYADSVKG (SEQ. ID. No. 65) 35.1 AINTNGNEKYYADSVKG (SEQ. ID. No. 66) 64.1 AISTNGKERYYADSVKG (SEQ. ID. No. 67) 130.1 AIRTQSQEEYYADSVKG (SEQ. ID. No. 68) 152.1 AIKSRSQEQYYADSVKG (SEQ. ID. No. 69) 167.1 AIKSHSQQQYYADSVKG (SEQ. ID. No. 70) 214.1 AINSEGQQRYYADSVKG (SEQ. ID. No. 71) 225.1 AIKSKGQNKYYADSVKG (SEQ. ID. No. 72) 262.1 AIRTNSEEKYYADSVKG (SEQ. ID. No. 73) 265.1 AINMKGNDRYYADSVKG (SEQ. ID. No. 74) 43.2 AI-RTNSKEYYADSVKG (SEQ. ID. No. 75) 70.2 AIKTESQQRYYADSVKG (SEQ. ID. No. 76) 99.2 AINSNGKQDYYADSVKG (SEQ. ID. No. 77) 100 .2 AIKTDGNHKYYADSVKG (SEQ. ID. No. 78) 109-.2 AIDTKGNGQYYADSVKG (SEQ. ID. No. 79) 146.2 AIRSDSSHXYYADSVKG (SEQ. ID. No. 80) 173.2 AINTKSNEQYYADSVKG (SEQ. ID. No. 81) 19 9. 2 AIRTDSKNSYYADSVKG (SEQ. ID. No. 82) 2.11 AIRTDSKEQYYADSVKG (SEQ. ID. No. 83) 19.11 AIRTNSKEEYYADSVKG (SEQ. ID. No. 84) 28.11 AIETSSGSTYYADSVKG (SEQ. ID. No. 85) WO 2007/124610 PCT/CH2007/000202 47 33.11 AINTGGGSTYYADSVKG (SEQ. ID. No. 86) 4.12 AINTRGQNEYYADSVKG (SEQ. ID. No. 87) 6.12 AISTSG-STYYADSVKG (SEQ. ID. No. 88) Among the isolated scrvs obtained after this procedure, seven turned out to have significantly im proved affinity with a Kd in the range of 2-3 nM (Fig. 5), the best of them being 28.11. Experiment 5: Prevention of tumor-growth upon administration of anti-ALK antibody The progenitors of the antibody ESBA521 were selected to bind- to amino acids 391-406 of ALK, which comprises the amino acids (396-406) tha.t.are known to bind pleiotrophin (Stoica 2001). ESBA521 was obtained by randomizing additional amino acids in the CDRs of its progenitor and by selecting for binders that bind to the 391-406 amino acid stretch contained in its natural con text of the ALK extracellular domain'(ECD). These pro ceedings. resulted in an antibody, which binds the ALK ECD with high affinity at the same site that binds PTN. To our knowledge. This is the first monoclonal antibody that specifically targets the PTN binding site of ALK. Ac cordingly, ESBA521 is predicted to have high affinity to the ALK ECD and efficiently competes with pleiotrophin (PTN) and midkine (MK) for binding to the ALK receptor, and thus, the ESBA521 antibody is suitable for inhibiting both MK and PTN ligand binding to the ALK protein. Because ALK and its ligands are involved in neoplasia, tumor invasion and angiogenesis, inhibition of the interaction between ALK and its cognate ligands dis-. rupts ALK mediated tumorgenesis.

WO 2007/124610 PCT/CH2007/000202 48 The effect of ESBA521 on a specific tumor can be determined by the following two assays described be low. In a first assay, xenograft experiments are prepared in order to determine the cancer growth rate limiting role of ALK (Powers 2002) . Briefly, a UB7MG cell suspension of 20 million cells/ml media supplemented with 10% fetal calf serum are prepared. These are injected into NU/NTJ mice and resultant tumors are measured. Test antibodies, preferably full length antibodies, more pref erably, pegylated antibodies, are introduced and.tumor growth is assessed. While there are shown and described presently preferred embodiments of the invention, it is to be dis tinctly understood that the invention is not limited thereto but may be otherwise variously embodied and prac ticed within the scope of the following claims. References:. Auf der Maur, A., Escher, D., and Barberis, A. (2001). Antigen-independent selection of stable intra cellular single-chain antibodies. FEBS Lett 508, 407-412. Auf der Maur, A., Tissot, K., and Barberis, A. (2004). Antigen-independent selection of intracellular stable antibody frameworks. Methods 34, 215-224. Auf der Maur, A., et al., Direct in vivo screening of intrabody libraries constructed on a highly WO 2007/124610 PCT/CH2007/000202 49 stable single-chain framework. J Biol Chem, .277(47) p. 45075-85, 2002. Bai RY. et al. Nucleophosmin-anaplastic lym phoma kinase associated with anaplastic large-cell lym phoma activates the phosphatidylinositol 3-kinase/Akt an tiapoptotic signalling pathway. Blood 96(13), 4319-4327, 2000. Bowden et. et al. Anti-apoptotic signaling of pleiotrophin through its receptor, anaplastic lymphoma kinase. The Journal of Biological Chemistry 277(39), 35862-35868, 2002. Chothia, C., and Lesk, A. M. (19B7) . Canoni cal structures for the hypervariable regions of iTrnu noglobulins. J. Mol. Biol. 196, .901.-917. Chothia, C., Novotny, J.,* Bruccoleri, R., and Karplus, M. (1985). Domain association in immunoglobulin molecules. The packing of variable domains. J. Mol. Biol. 186, 651-663. Choudhuri et al., An Angiogenic Role for the Neurokines Midkine and Pleiotrophin in Tumorigenesis. Cancer Res. 57, 1814-1819, 1997. Czubayko et al., Melanoma angiogenesis and metastasis modulated by ribozyme targeting of the se creted growth facator pleiotrophin. PNAS 93, 14753-14758, 1996.

WO 2007/124610 PCT/CH2007/000202 50 De Juan C. et al. Genomic organization of a novel glycosylphosphatidylinositol MAM gene expressed in human tissues and tumors. Oncogene 21, 3089-3094, 2002. Delsol G. et al. A new subtype of large B cell lymphoma expressing the ALK kinase and lacking the 2;5 translocation. Blood 89(5), 1483-1490, 1997. Dirks WG. et al. Expression and functional analysis of the anaplastic lymphoma kinase (ALK) gene in tumor cell lines. International Journal of cancer 100, 49-56, 2002. Duyster J. et al. Translocations involving anaplastic lymphoma kinase (ALK) . Oncogene 20, 5623-5637, 2001. Ergin M. et al. Inhibition'of tyrosine kinase activity induces caspase-dependent apoptosis. in anaplas tic large cell lymphoma with NPM-ALK (p80) fusion. pro tein. Experimental Hematology 29, 1082-1090, 2001. Fang et al., Pleiotrophin stimulate' fibro blasts and endothelial and epithelial cells and is ex pressed in human cancer. JBC vol. 267 p. 25889, 1992. Fiorani C. et al. Primary systemic anaplastic large-cell lymphoma (CD30+): advances in biology and cur rent therapeutic approaches. Clinical Lymphoma 2(1), 29 37, 2001.

WO 2007/124610 PCT/CH2007/000202 51 Iwahara T. et al. Molecular characterization of ALK, a receptor tyrosine Icinase expressed specifically in the nervous system. Oncogene 14, 439-449, 199'7. Kutok JL. and Aster JC. Molecular biology of anaplastic lymphoma kinase-positive anaplastic 'large-cell lymphoma. Journal of Clinical Oncology 20(17), 3691-3702, 2002. Ladanyi M. Aberrant AjK tyrosine kinase sig nalling. Different cellular lineages, common oncogenic mechanisms? American Journal of Pathology 157-(2), 341 345, 2000. Lamant et al. Expression of the ALK tyrosine kinase gene in neuroblastoma. Am. J. Pathol. 156, 1711 1721, 2000. Li, X.-Q., Hisaoka, M., Shi, D.-R., Zhu, X Z., and Hashimoto, H. Expression of Anaplastic Lymphoma Kinase in soft tissue tumors: an .immunhistochemical and molecular study of 249 cases. Human pathology 35, 711 721, 2004. Loren CE. et al. Identification and charac terization of ' Dalk: a novel Drosophila melanogaster RTK which drives ERK activation in vivo.. Genes to cells 6(6), 531-544, 2001. Miyake I. et al. Activation of anaplastic lymphoma kinase is responsible for hyperphosphorylation WO 2007/124610 PCT/CH2007/000202 52 of ShcC in neuroblastoma cell lines. Oncogene 21, 5823 5834, 2002. Morris SW. et al. ALK, the chromosome 2 gene locus altered by the t(2;5) in non-Hodgkin's lymphoma, encodes a novel neural receptor tyrosine kinase that is highly related to leukocyte tyrosine kinase (LTK) . Onco gene 14, 2175-2188, 1997. Morris SW. et al. ALK+CD30+ lymphomas: A dis tinct .molecular genetic subtype of non-Hodgkin' s lym phoma. British Journal of Haematology 113, 275-295, 2001. O'Brien et al., The angiogenic factor midkine is expressed in bladder cancer and overexpression corre lates with a poor outcome in patients with invasive~ can cer. Cancer Res. 56, 2515-251B, 1996. Padlan, E. A. (1994) . Anatomy of the antibody molecule. Mol. Immunol 31, 169-217. Powers C. et al. Pleictrophin signalling through anaplastic lymphoma kinase is rate-limiting for glioblastoma growth. The journal of biological chemistry 277'(16), 14153-14158, 2002. Pulford K. et al. Anaplastic lymphoma kinase proteins and malignancy. Current Opinion in Hematology 81, 231-236, 2001. Pulford K. et al. Anaplastic lymphoma kinase proteins in growth control and cancer. J Cell Physiol 199, 330-358, 2004.

WO 2007/124610 PCT/CH2007/000202 53 Schaerer-Brodbeck, C. and A. Barberis, Cou pling homologous recombination with growth selection in yeast: a tool for construction of random DNA sequence li braries. Biotechniques, 37(2): p. 202-206, 2004. Stoica GE. et al. Identification of anaplas tic lymphoma kinase as a receptor for the growth factor pleiotrophin. The Journal of Biological Chemistry 2'76(20), 16772-16779, 2001. Stoica GE. et al. Midkine binds to anaplastic lymphoma kinase (ALK) and acts as a growth factor for different cell types.. The Journal of Biological Chemistry 277 (39), 35990-35998, 2002. Weber D. et al. Pleiotrophin can -be rate limiting for pancreatic cancer cell growth. Cancer Re search 60, 5284-5288, 2000. Wellstein et al., A Heparin-binding Growth Factor Secreted from Breast Cancer Cells Homologous to a Developmentally Regulated. Cytokine. JBC Vol 267, p2582, 1992. SWISSPROT: www. expasy_.org www .emedicine. comn/med/topic2692, htm (glioblastoma) www. enedicine .com/MED/topic3205 .htm (ALCL) - 53A Throughout this specification and the claims which follow, unless the context requires otherwise, the word "comprise", and variations such as "comprises" and "comprising", will be understood to imply the inclusion of a 5 stated integer or step or group of integers or steps but not the exclusion of any other integer or step or group of integers or steps. The reference in this specification to any prior publication (or information derived from it), or to any 10 matter which is known, is not, and should not be taken as an acknowledgment or admission or any form of suggestion that that prior publication (or information derived from it) or known matter forms part of the common general knowledge in the field of endeavour to which this specification relates.

Claims (17)

1. An isolated antibody which inhibits pleiotropin (PTN) binding to human anaplastic lymphoma kinase (ALK).

2. The antibody of claim 1, wherein said antibody inhibits angiogenesis.

3. The antibody of claim 1, wherein said antibody inhibits apoptosis.

4. The antibody according to any one of claims 1 to 3 which binds to an ALK epitope being a fragment of, comprising or essentially consisting of the region spanning amino acid residues 391-406 (SEQ ID No: 1).

5. The antibody according to any one of claims 1 to 4, having an affinity KD for the epitope of claim 4 of 10 nM or less.

6. The antibody according to any one of claims 1 to 5, wherein said antibody is a full-length antibody of the IgG or the IgM type.

7. The antibody according to any one of claims 1 to 5, wherein said antibody is a Fab fragment, a F(ab')2 fragment, a Fd fragment, a Fv fragment or a scFv fragment.

8. The antibody according to any one of claims 1 to 7, wherein said antibody is radio labeled, toxin-labeled or pegylated.

9. The antibody according to any one of claims 1 to 8 for use as medicament or diagnostic tool. C-\NRPonbl\DCC\SCG'471I 60I DOC-I/Il/2012 - 55

10. Use of the antibody according to any one of claims 1 to 8 for the production of a medicament for the treatment of cancers or tumors.

11. A method of treating cancers or tumors, said method comprising administering an effective amount of the antibody according to any one of claims 1 to 8 to a subject with a cancer or tumor.

12. The use of claim 10 or the method of claim 11, wherein the treatment is of neuroblastoma, glioblastoma, rhabdomyosarcoma, breast carcinoma, melanoma, pancreatic cancer, B-cell NHL, thyroid carcinoma, small lung carcinoma, retinoblastoma, ewing sarcoma, prostate cancer, colon cancer, pancreatic cancer, lipoma, liposarcoma or fibrosarcoma.

13. An isolated DNA sequence encoding the antibody of any one of claims 1 to 8.

14. An expression vector comprising the DNA sequence of claim 13.

15. A suitable host cell comprising the expression vector of claim 14.

16. A medicament comprising the antibody of any one of claims 1 to 8.

17. An isolated antibody according to any one of claims 1 to 8, use according to claim 10 or claim 12, a method according to claim 11 or claim 12, an isolated DNA sequence according to claim 13, an expression vector according to claim 14, a host C.NRPonbl\DCCSCG\4796)_I DOC-l/l 112012 - 56 cell according to claim 15, or a medicament according to claim 16, substantially as hereinbefore defined with reference to the Figures and/or Examples.