EP1173608A1

EP1173608A1 - Interaction of vascular-endothelial protein-tyrosine phosphatase with the angiopoietin receptor tie-2

Info

Publication number: EP1173608A1
Application number: EP00927011A
Authority: EP
Inventors: Werner Risau; Gregor Fachinger; Urban Deutsch
Original assignee: Max Planck Gesellschaft zur Foerderung der Wissenschaften eV
Current assignee: Max Planck Gesellschaft zur Foerderung der Wissenschaften eV
Priority date: 1999-04-23
Filing date: 2000-04-20
Publication date: 2002-01-23
Also published as: WO2000065085A1; AU4554300A; EP1046715A1

Abstract

Use of vertebrate vascular-endothelial protein tyrosine phosphatases (i.e.: murine phosphatase VE-PTP or human phosphatase HPTPβ) or portions thereof for the manufacture of an agent for monitoring or β modulating the activity of the angiopoietin receptor-type tyrosine kinase Tie-2.

Description

Interaction of vascular-endothelial protein-tyrosine phosphatase with the Angiopoietin receptor Tie-2

Specification

The present invention relates to a method for monitoring or modulating the activity of the angiopoietin receptor-type tyrosine kinase Tie-2.

A key mechanism in the proliferation and differentiation control of all cells are membrane-located receptors, whose activation in many cases is mediated by external factors via phosphorylation of tyrosine residues. The mutation of a series of endothelial cell specific receptor-tyrosine kinases (RTKs) results in lethal phenotypes early during murine embryonal development (Hanahan, Science 277 ( 1 997), 48 - 50; Risau, Nature 386 (1 997), 671 - 674). The proliferation and differentiation of endothelial cells depends on two receptor tyrosine kinase systems. The vascular endothelial growth factor (VEGF) is a secreted angiogenic factor and promotes vascularization by activation of its high affinity receptors VEGFR-1 (Flt-1 ) or VEGFR-2 (FIk-1 ). The RTKs Tie-1 and Tie-2 are involved in the sprouting and remodelling of the embryonic vascular system. The activity of these kinases is regulated by the recently identified ligands, the angiopoietins.

After ligand binding RTKs are activated by phosphorylation on tyrosine residues. Specific protein-tyrosine phosphatases (PTPs) are involved in the fine-tuning of RTK activity. Several classes of PTPs have been identified. However, the biological functions thereof are presently not understood (Neel & Tonks, Curr. Opin. Cell Biol. 9 ( 1 997), 1 93 - 204; Streuli, Curr. Opin. Cell Biol. 8 ( 1 996), 1 82 - 1 88) .

In a study to identify PTPs in endothelial cells a murine vascular-endothelial protein-tyrosine phosphatase VE-PTP was identified (VE-PTP: a receptor protein-tyrosine phosphatase expressed in vascular endothelium, EMBO- FEBS Workshop on Protein Phosphatases and Protein Dephosphorylation, Oxford, UK, September 21 - 26, 1 997). Indications for a functional interaction between VE-PTP and a receptor-type kinase have not been described, however. Further, the association of PTPs with their substrates is difficult to determine due to the transcient nature of the enzyme substrate association (Flint et al., Proc. Natl. Acad. Sci. U.S.A. 94 ( 1 997), 1 680 - 1 685) .

The experiments underlying the present application discovered that VE-PTP is a homoiog of the human HPTP/? (Krueger et al., E BO J., 9, ( 1 990), 3241 - 3252), and that it is specifically expressed in endothelial cells both during the embryonal development of mice and in brain capillary vessels of newborn animals. Biochemical analyses using VE-PTP trapping mutants show a specific interaction between the C-terminal part of the molecule which contains the catalytic domain and the RTK Tie-2 but not with the vascular endothelial growth factor receptor VEGFR-2. Moreover, a dephosphorylation of Tie-2 could be detected in the presence of VE-PTP in transiently transfected COS-1 cells. These data identify Tie-2 as a specific substrate for VE-PTP and show that it is a specific modulator of Tie-2 activity.

This result is of high clinical relevance, as Tie-2 holds a key position in angiogenetic processes, the formation of the blood vessel system during embryonal development, the healing of wounds as well as in pathological processes, e.g. tumor development. As VE-PTP shows a specific interaction with Tie-2 and can modulate the tyrosine phosphorylation of the latter, the receptor-protein tyrosine phosphatase is a target both for diagnostic monitoring and for therapeutically influencing the said processes.

Thus, a subject matter of the present invention is the use of vertebrate, e.g. mammalian vascular-endothelial protein-tyrosine phosphatases or portions thereof for the manufacture of an agent for monitoring or modulating the activity of receptor-type tyrosine kinase Tie-2.

A further subject matter of the present invention is the use of nucleic acids encoding vertebrate, e.g. mammalian vascular-endothelial protein-tyrosine phosphatases or portions thereof for the manufacture of an agent for monitoring or modulating the activity of receptor-type tyrosine kinase Tie-2.

Still a further subject matter of the invention is the use of ligands for vertebrate, e.g. mammalian vascular-endothelial protein-tyrosine phosphatases for the manufacture of an agent for monitoring or modulating the activity of receptor-type tyrosine kinase Tie-2.

The vascular-endothelial protein-tyrosine phosphatases and nucleic acids coding therefor, e.g. genes or cDNA molecules, are obtainable from vertebrate cells, preferably from mammalian endothelial cells, e.g. murine or human cells. Preferably the vascular-endothelial protein-tyrosine phosphatase is selected from murine phosphatase VE-PTP, human phosphatase HPTP ? or portions thereof, particularly portions comprising the catalytic domain which is located at the C-terminus of the molecule (Fig.

1 a). The nucleic acid sequence and the corresponding amino acid sequence of murine vascular-endothelial protein-tyrosine phosphatase are depicted in

SEQ. ID. NO 1 and 2, respectively. The corresponding sequences of the human protein, which were identified by Krueger et al. (supra) are depicted in SEQ. ID. NO 3 and 4.

The polypeptide or a portion thereof is suitable for monitoring or modulating the activity of receptor-type tyrosine kinase Tie-2. In addition to a phosphatase with unmodified sequence of the catalytic domain also mutants thereof, which show a modified, e.g. enhanced binding to Tie-2, e.g. the trapping mutants as depicted in Fig. 2 are suitable for the present invention. Particularly mutants, which exhibit an enhanced binding to Tie-2 are well suited for diagnostic and therapeutic applications.

The interaction between the vascular endothelial protein-tyrosine phosphatase and its substrate Tie-2 can also be monitored and/or modulated on the nucleic acid level. To this end nucleic acids, e.g. DNA molecules, RNA molecules or artificial nucleic acid analogs such as peptidic nucleic acids may be used. Preferably these nucleic acids comprise at least 1 5, particularly at least 20 nucleotides from murine phosphatase VE-PTP gene, human phosphatase HPTP ? gene or sequences complementary thereto. These nucleic acids are suitable for the determination of the PTP expression by using known hybridization or/and amplification techniques such as PCR. On the other hand, nucleic acids can be used for the modulation of the VE-PTP expression in the form of antisense constructs or as ribozymes.

A still further aspect of the invention is the use of ligands for vertebrate, e.g. mammalian vascular endothelial-protein tyrosine phosphatases. Examples of such ligands are antibodies, e.g. polyclonal or monoclonal antibodies and antibody fragments. Polyclonal antibodies are available according to known protocols by immunization of test animals with purified VE-PTP, HPTPff or partial fragments thereof, which preferably contain the catalytic domain. From these test animals monoclonal antibodies can be generated in a known manner by using the method applied by Koehler and Milstein. The polyclonal or monoclonal antibodies can also be used in the form of fragments which are obtainable by proteolyic treatment or genetic engineering.

One embodiment of the invention concerns the monitoring or detection of the Tie-2 activity. This detection can be carried out by using known methods, e.g. using labelled polypeptides, nucleic acids or antibodies. A further embodiment concerns the modulation of the Tie-2 activity. Thereby a stimulation or a repression of the Tie-2 activity is possible.

Of major importance is the examination or influencing of the interaction between VE-PTP and Tie-2 for angiogenesis. Thus the present invention provides means for inducing or for inhibiting vascular growth or remodelling and blood vessel maturation. Particularly, the present invention provides means for inhibiting tumor growth and formation of tumor metastases, e.g. by repressing Tie-2 activity in target cells.

Moreover, the invention is explained by the following figures and sequence protocols.

Fig. 1 a shows the schematic representation of VE-PTP, its genetically engineered trapping mutants and HPTPβ.

Fig. 1 b and c show Northern blot and RT-PCR analyses of VE-PTP expression in mouse tissues and during mouse embryonic development.

Fig. 2 shows in vivo expression analysis of VE-PTP by in situ hybridization.

Fig. 3 shows biochemical interactions of VE-PTP trapping mutants with Tie-2 protein.

Fig. 4 shows selective dephosphorylation of Tie-2, but not

VEGFR-2 by wild-type VE-PTP.

Fig. 5 shows a sequence comparison of the C-terminus of

HPTP^ with VE-PTP and the translated "mRPTP^" sequence. Known protein domains are depicted: M embrane proxim al FN I l l-d o mai n ( b l ue) , transmembrane domain (red) and catalytic domain

(green). The catalytic center is characterized by a C(x)₅R-motif.

SEQ. ID. NO. 1 and 2 show the nucleotide sequence of VE-PTP cDNA and the corresponding amino acid sequence.

SEQ. ID. NO. 3 and 4 show the nucleotide sequence of HPTPβ cDNA and the corresponding amino acid sequence.

Example 1

A PCR screen of a murine brain capillary cDNA library and reverse transcribed mRNA of bEND5 endothelioma cells to identify endothelial specific members of the protein-tyrosine phosphatase family was performed. For PCR, 1 00 pmol degenerated primers RPTP1 5'-GA(C/T)

TT(C/T) TGG ATG (A/G/T) (G/T)l TGG GA-3' and RPTP2 5'-CCI ACI CGI

GCI (G/C)(A/T)(A/G) CA(A/G) TGI AC-3' in 50 μ\ reactions were used. As templates 1 .25 μg Λ-DNA from mouse P4-1 0 brain capillary-library

(Schnϋrch & Risau, Development, 1 1 9 ( 1 993), 957 - 968) or 3 μ\ of

Superscript cDNA preparation (GIBCO BRL) from bEND5 mRNA were used.

Isolated 370 bp products were cloned into the vector pCRII (Invitrogen), analysed by restriction cleavage and sequenced on an ABI 370 automated sequencer (Applied Biosystems).

One of the identified PCR products encodes a polypeptide, designated as vascular-endothelial protein-tyrosine phosphatase (VE-PTP) which was identified as murine homolog of the previously described receptor-type protein-tyrosine phosphatase HPTP/? (Krueger et al. EMBO J. 9 ( 1 990), 3241 - 3252) . VE-PTP and HPTPff belong to the subclass III of receptor-type PTPs bearing exclusively fibronectin type Ill-like repeats in the extracellular domain and a single catalytic domain in the cytoplasmatic tail (Fig. 1 a) (Brady-Kalnay & Tonks, Curr. Opin. Cell. Biol. 7 ( 1 995), 650 - 657).

Fig. 1 a shows a schematic representation of VE-PTP, its genetically engineered trapping mutants C- > S, R- > A and HPTP/?. Rectangles indicate mutated amino acids in the catalytic core. The location of the degenerated primers used in the PCR screen are indicated by arrows (EC-dom., extracellular domain; FN III fibronectin-type Ill-like repeat; cat. dom., catalytic domain) .

Example 2

A Northern blot and RT-PCR analysis of VE-PTP expression in mouse tissues and during mouse embryonic development were performed. A 751 bp EcoRI-fragment from VE-PTP part 1 , obtained by PCR using primers PrPTPβfor 5'-GGA AGA GGT ACC TGG TGT CCA TCA AGG-3' and PrPTPβrev 5'-GGC CGG TCC CTA CGA ATG CTG AGC CGG GCA G-3' deduced from a partial clone of murine "RPTP/?" (Schepens et al. Mol. Biol. Reports, 1 6 ( 1 992)), and cloned in the vector pBS KS( + )(Stratagene), was labelled with ³²P-dCTP (Amersham Pharmacia Biotech) . For Northern blot analysis 20 μg of total RNA from mouse tissues (Chomczynski & Sacchi, Analyt. Biochem. 1 62 ( 1 987), 1 56 - 1 59) were loaded on a formaldehyde containing agarose gel and blotted. A mouse embryo mRNA Northern blot was obtained from Clontech and hybridization was carried out according to manufacturer's instructions. Autoradiography was performed at -70° C for 1 7 d. For semiquaπtitative PCR 50 μ\ reactions containing 2 μ\ of reverse transcribed cDNA preparations and 20 pmol of primers betaseq2 5'- CCC TCT CCC TTC CTA CCT GG-3' and betarev 5'- GGC CGG TCC CTA CGA ATG CTG AGC CGG GCA GG-3' were used, giving a 41 6 bp fragment. 30 cycles PCR was optimized to detect 1 fg of VE-PTP plasmid DNA. ?-actin RT-PCR was performed as described (Nakajima-lijima et al, Proc. Natl. Acad. Sci. U.S.A. 82 ( 1 985), 61 33 - 61 37) . Northern blot analysis of VE-PTP expression revealed a major transcript of approximately 1 1 kB and two additional transcripts of 7.5 and 6 kB. In the adult mouse VE-PTP mRNA was strongly expressed in brain as well as in lung and heart. Very weak expression was detectable in kidney and liver (Fig. 1 b) . These data were confirmed by semi-quantitative RT-PCR performed with RNA from these organs (Fig. 1 b) . During embryonic development VE-PTP was weakly expressed at embryonic day El l , expression increased at E1 5 reaching a maximum at E1 7 (Fig. 1 c) . Strong expression was detected at E7, which may result from expression in contaminating maternal tissue as expression in the placenta was observed by in situ hybridization analysis as well.

Example 3

An in vivo expression analysis of VE-PTP by in situ hybridization to frozen sections of mouse embryonic tissues was carried out. The results are shown in Fig. 2. Fig. 2a is a darkfield image of an E1 2.5 embryo section hybridized with a VE-PTP antisense probe. (NC: neural crest, DA: dorsal aorta) . Fig. 2b is a darkfield image and Fig. 2c is a brightfield image of a higher magnification of the vessel indicated in a (asteriks). Fig. 2d - h are sagittal sections of E1 5.5 embryos hybridized with antisense VE-PTP probes. Fig. 2d is a darkfield image and Fig. 2e a brightfield image of the lung. Fig. 2f is a darkfield image of the head region. Fig. 2g is an E1 5.5 embryo section hybridized with a VEGFR-2 antisense probe. Fig. 2h - k are vessels in brain sections of P10 mice hybridized with antisense VE-PTP probes. As templates for in vitro transcription pCRII (Invitrogen) VE-PTP-1 (370 bp fragment of VE-PTP coding for protein sequence corresponding to aa 1 786 - 1 91 3 in HPTPβ in pCRII) and pBS VE-PTPpart1 were used. Sectioning of mouse embryos and in situ hybridization were performed as described (Breier et al, Development, 1 14 (1 992), 521 - 532). At the earliest timepoint analysed (E9.5), expression was detectable in the endothelial cell layer lining the dorsal aortae. During the subsequent developmental stages VE-PTP expression was increased throughout the developing vascular system (Fig. 2a) . Strong hybridization signals were visible in endothelial cells forming blood vessels, whereas no specific signals were detected in blood cells or smooth muscle cells surrounding the vessels (Fig. 2b, c) . At E1 5.5 specific signals were detectable in all organs with highest expression in the lung (Fig. 2d,e). Comparison to serial sections hybridized with an antisense probe to VEGFR-2 (Flk-1 ) as an endothelial cell marker, confirmed the vascular endothelial specific expression pattern of VE-PTP (Fig. 2 f,g). In contrast to the uniform expression levels of VEGFR-2 in different types of embryonic endothelial cells, VE-PTP was more strongly expressed in endothelial cells lining larger, smooth muscle cell invested vessels than those of small capillaries and veins. On brain sections of newborn mice, specific expression of VE-PTP was detectable in brain capillaries as well as in larger vessels (Fig. 2h-k) . No specific signals were visible in neuronal or glial cells.

Example 4

The biochemical interactions of VE-PTP with the receptor tyrosine kinases Tie-2 and VEGFR-2 were investigated using bacterial GST-fusion proteins. The results are shown in Fig. 3.

Fig. 3a demonstrates the results of GST-fusion pull down experiments. GST and GST x VE-PTP R/A fusion protein were incubated with lysates from bEND5 cells. Precipitates were blotted with an anti-Tie-2 antibody and reblotted with an VEGFR-2 specific antibody, (tot. lys.: total lysates of bEND5 cells) . pGEX-VE-PTP contains a 1 .1 kB 3' part of EST-clone 552065 (Lennon et al., Genomics 33 (1 996), 1 51 -1 52) coding for the cytoplasmic domain of VE-PTP cloned in pGEX 3T (Amersham Pharmacia Biotech) . GST and GST-fusion proteins were expressed in E.coli strain TOP10 essentially as described (Frangioni & Neel, Anal. Biochem. 21 0 ( 1 993), 1 79 - 1 87). For pull down experiments 10 cm dishes of confluent endothelial cells were pretreated with pervanadate, lysed and incubated with 1 0 μg of GST-fusion protein prebound to glutathion-sepharose as described before (Jallal et al., J. Biol. Chem. 272 (1 997), 1 21 58 - 1 21 63) .

Fig. 3b shows co-immunoprecipitation of VE-PTP trapping mutants (C- > S, R- > A) with Tie-2. COS-1 cells were transfected with FLAG-tagged VE-PTP and trapping mutants together with Tie-2. Immunoprecipitation was performed with anti-FLAG antibody M2. Precipitates were blotted with a Tie-2 specific monoclonal antibody.

pCMV-FLAG VE-PTP wt, C- > S and R- > A contain cDNA sequences coding for a polypeptide stretch corresponding to aa 1 41 8-1 977 in HPTP/? cloned in pCMV-FLAG-1 (Kodak). Trapping mutations C-> S and R- > A were introduced by PCR mutagenesis using primer Prbetamutcs 5'-TCC GTA GTG CAC TCG AGT GCT GGT GTG-3' and primer Prbetamutra 5'-GCT GGT GTG GGC GCC ACA GGG ACG TTC-3'. COS-1 cells (Gluzman, Cell 23 ( 1 981 ), 1 75 - 1 82) were transfected using the modified calcium phosphate method (Chen & Okayama, Mol. Cell. Biol. 7 ( 1 987), 2745 - 2752). For transfection 1 0 μg of pCM V-FLAG derivates and 2 μg of expression plasmids coding for the RTKs were used. As control 0.5 μg of EGFP expression plasmid (Clontech) were cotransfected. Cells were harvested after 2 d of expression. Transfection efficiency was evaluated under fluorescent light and was usually between 30 - 70%.

In mixing experiments of endothelial cell lysates and trapping mutants of the VE-PTP catalytic domain fused to GST, we detected interaction with the Tie-2 receptor, while GST alone did not precipitate Tie-2. The interaction was independent of pretreatment with pervanadate. In these assays coprecipitation of VEGFR-2 was never detectable (Fig. 3a). To test for potential substrate interactions with Tie-2 and VEGFR-2 we coexpressed these RTKs with either a FLAG-tagged version of VE-PTP corresponding to aa 141 8-1 997 of HPTP/?, or the respective trapping mutants (Fig. 1 a) . Physical association was analysed by co- immunoprecipitation using an anti-FLAG-antibody and subsequent blotting of the precipitates with antibodies specific for the repective RTK. In this assay the Tie-2 receptor co-precipitated with both trapping mutants of VE- PTP (C- > S, R- > A) (Fig. 3b). The wild type phosphatase failed to precipitate Tie-2 efficiently, even though the receptor was expressed at comparable levels. This reduced association of PTPs in vitro with their substrates is due to the transient nature of the enzyme substrate association. Unlike Tie-2, VEGFR-2 could neither be co-immunoprecipitated with VE-PTP nor with one of the trapping mutants, even though VEGFR-2 expression was comparable to that of Tie-2.

Example 5

Finally, the phosphorylation state of RTKs was determined in the presence of VE-PTP. Figure 4 shows dephosphorylation of (a) Tie-2 but not (b) VEGR- 2 by wild-type VE-PTP. RTKs were immunoprecipitated with specific antibodies from cotransfected COS-1 cells. Precipitates were blotted with anti-phosphotyrosine antibodies and after stripping reprobed with RTK- specific antibodies.

Tie-2 and VEGFR-2 expression vectors were published previously (Koblizek et al., Curr. Biol. 8 ( 1 997), 529 - 532; Millauer et al., Cell 72 ( 1 993), 835 - 846) . Rat monoclonal antibodies against Tie-2 clones 3g1 and 4g8 (Koblizek et al. ( 1 997) supra) and Flk-1 clone 1 2σ1 (Kataoka et al., Devel. Growth Diff. 39 ( 1 997), 729 - 740) were used. Immunoprecipitations were performed with 5 /yg of the monoclonal antibodies and immunoblotting with 2 μglrc . Polyclonal anti-Flk-1 serum 1 D3 (Sugen) was used in a 1 :5000 dilution. Monoclonal anti-Flag antibody M2 (Kodak), polyclonal antiserum against Tie-2 (Santa Cruz Biotechnology) and monoclonal mouse antibody against phosphotyrosine PY20 (Transduction Labs) were used according to the manufacturer's instructions. Immunoprecipitations and immunoblotting were performed as described before (Esser et al., J. Cell. Biol. 1 40 ( 1 998), 947 - 959); Jallal et al., J. Cell. Biol. Chem. 272 ( 1 997), 1 21 58 - 1 21 62) .

Immunoprecipitates of VEGFR-2 and Tie-2 co-expressed with either the VE- PTP trapping mutants (C- > S, R- > A) or wt VE-PTP were blotted with an a- phosphotyrosine-specific antibody and then reprobed with antibody specific for the RTK. Only for Tie-2, changes in the phosporylation status were observed. In the presence of the trapping mutants (C- > S, R- > A) the receptor was reproducibly more highly phosphorylated than in the controls. This hyperphosphorylation of Tie-2 in the presence of catalytically impaired trapping mutants suggests that physical interaction leads to protection of the receptor from dephosphorylation. In contrast, hypophosphorylation of the Tie-2 receptor was observed in the presence of wt VE-PTP, when compared to vector control (Fig. 4a) . No significant changes were detected in the phosphorylation status of VEGFR-2, irrespective of the presence of VE-PTP or its trapping mutants (Fig. 4b). These findings clearly show that Tie-2 is a specific substrate for the endothelial specific phosphatase VE- PTP.

Claims

1 . Use of vertebrate vascular-endothelial protein-tyrosine phosphatases or portions thereof for the manufacture of an agent for monitoring or modulating the activity of receptor-type tyrosine kinase Tie-2.

2. The use of claim 1 wherein said phosphatase is selected from murine phosphatase VE-PTP, human phosphatase HPTP ?or portions thereof.

3. The use of claim 1 or 2 wherein said portion comprises the catalytic domain.

4. Use of nucleic acids encoding vertebrate vascular-endothelial protein- tyrosine phosphatases or portions thereof for the manufacture of an agent for monitoring or modulating the activity of receptor-type tyrosine kinase Tie-2.

5. The use of claim 4 wherein said nucleic acid comprises at least 1 5 nucleotides from murine phosphatase VE-PTP nucleic acid, human phosphatase HPTP ? nucleic acid or sequences complementary thereto.

6. The use of ligands for vertebrate vascular-endothelial protein-tyrosine phosphatases for the manufacture of an agent for monitoring or modulating the activity of receptor-type tyrosine kinase Tie-2.

7. The use of claim 7 wherein said ligands are selected from antibodies and antibody fragments.

8. The use of any one of claims 1 - 7 for detecting Tie-2 activity.

9. The use of any one of claims 1 - 7 for stimulating Tie-2 activity.

10. The use of any one of claims 1 - 7 for repressing Tie-2 activity.

1 1 . The use of any one of the previous claims for monitoring or modulating angiogenesis.

1 2. The use of any one of the previous claims for inducing vascular growth or remodelling and blood vessel maturation.

3. The use of any one of the previous claims for inhibiting vascular growth or remodelling and blood vessel maturation.

4. The use of any one of the previous claims for inhibiting tumor growth.

5. The use of any one of the previous claims for inhibiting formation of tumor metastases.