JP2002516338A - Agents as inhibitors of cell migration and / or focal adhesion that interfere with the binding of protein tyrosine phosphatase PEST to signal proteins - Google Patents

Abstract

(57) SUMMARY The present invention interferes with the binding of protein tyrosine phosphatase PEST (PTP-PEST) to signaling molecules involved in cell migration, focal adhesion and / or cell proliferation, ie, p130cas and paxillin. To a compound or drug. Such agents can be derived from the minimal sequence found in binding studies. PTP-PEST is a conserved phosphatase required for embryo differentiation. Knockout cells (PTP-PEST − / −) are obtained by immortalization of null embryos, in which abnormalities in cell migration, focal adhesion and cell proliferation are observed. Therefore, an agent capable of interfering with the activity of PEST in an affected target tissue is a potential agent for treating a disease caused by any phenomenon selected from the group consisting of cell proliferation, cell migration, inflammation and angiogenesis. Pharmaceutical composition. Furthermore, the present invention relates to a method for detecting a true enzyme substrate for an enzyme, that is, phosphatase, by a method combining a method for producing a knockout by gene targeting and a substrate trapping technique using a substrate-binding inactivated mutant enzyme. . By using the method of generating knockouts by gene targeting, there is less detection of erroneous results than other techniques for artificially increasing nonspecific hyperphosphorylation levels (eg, methods using vanadate compounds).

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001] Technical Field The present invention is an agent that modulates focal adhesion and cell migration, pharmaceutical agents, especially regulates the phosphorylation of tyrosine residues in proteins.

[0002] Phosphorylation of background tyrosine residues invention, one of the important mechanisms for cell adhesion, cell mitogenic, in biochemical and cellular events, such as differentiation and cell migration to transduce extracellular stimuli (Outline of Document 1). The degree of tyrosine phosphorylation of proteins is regulated by the functions of two protein families: protein tyrosine kinases (PTKs) and protein tyrosine phosphatases (PTP or PTPase). All PTPases are unique motifs [I / V]
It has a conserved catalytic domain characterized by HCxAGxxR [S / T] G. Biochemical and kinetic studies indicate that mutation of a cysteine residue within this unique motif completely abolishes PTPase activity, indicating that this cysteine residue is essential for the catalytic activity of the PTPs. (Reference 2).

[0003] PTP-PEST is a stable soluble PTP and is ubiquitously expressed in embryonic development and in adult tissues of mice (Reference 3). The N-terminal part of PTP-PEST contains a catalytic domain, and the C-terminal part is five proline-rich domains (Reference 4)
And a binding site for the adapter protein Shc (Reference 5). Of the proline-rich domains, two motifs are recognized by the SH3 domain of other signaling molecules. The Pro4 (PPPER) motif has been shown to promote interaction with the SH3 domain of Csk kinase. This same Pro4 motif, along with the Pro1 (PPKPPR) motif, Gr
Interacts with the SH3 domain of the b2 adapter protein. It has been proposed that this interaction facilitates the movement of PTP-PEST towards the activated epidermal growth factor receptor to dephosphorylate a substrate of approximately 120 kDa (6). The Pro1 domain of PTP-PEST is the only known substrate to date, the adapter protein p130cas (crk associated substrata).
te) was actually found to play a major role in binding PTP-PEST (Reference 6). Recently, the present inventors have actually shown that p130cas is constitutively tyrosine phosphorylated in the absence of a PTP-PEST regulator using PTP-PEST-deficient fibroblasts (Reference 7). . This is PTP-P
FIG. 4 shows the physiological relationship between Pro1 of EST and SH3 domain of p130cas.

[0004] Although p130cas was recognized as a direct target for YopH PTPase during Yersinia infection (Reference 8), it was also identified as a major substrate for tyrosine kinase in v-crk transformed cells (Reference 9). . In normal cells,
After activation by integrin-dependent FAK or Src kinase, p130ca
s becomes hyperphosphorylated (References 10 and 11). This phosphorylation is thought to perform a series of tyrosine-dependent signaling. The resulting phenomena involves the remodeling of actin filaments. P130c of PTP-PEST
Effects on as have dramatic consequences on mammalian development as well as on physiopathological events due to the importance of integrin-dependent signaling in the cytoskeleton, cell motility and transformation. The process of cell migration is critical for correct embryonic development in mammals. In adults, cell migration also plays an important role in events such as infiltration of lesions by fibroblasts and epithelial cells and translocation of lymphocytes and neutrophils to inflammatory sites. In cancer, cancer cells need to move in order to reach the circulatory system and spread throughout the living body. Therefore, PTP-P
By providing EST negative cells, PTP-PEST and its substrate p
It is important to assess the physiological significance in cell motility, such as 130 cas. Inhibitors of PTP-PEST can be useful therapeutic agents.

[0005] The power of cell motility is the polymerization / depolymerization of actin under the control of the Rho family of G proteins. The three most studied members of this family are Rho, Rac and Cdc42, which control the formation of focal adhesions, the formation of cell membrane ruffles and the formation of filopodia, respectively. . Cell membrane folds and filopodia consist of actin-rich cell membrane protrusions that the cells themselves use to extend forward. The difference between the two is that the folds are formed by a network of interlaced actin filaments that form them, whereas filopodia is thought to be in the form of hair.

[0006] Focal adhesion is a site of contact between the extracellular matrix and the cytoskeleton via the integrin family of transmembrane proteins (
Reference 12). Focal adhesion involves structural proteins such as tailin, tensin and α-actinin, and FA activated by binding to extracellular matrix.
It contains important tyrosine kinases such as K, Src and Csk. And p
A protein complex is formed that contains many different adapter proteins, such as 130cas, Shc, Grb2, Crk and Nck. The interaction of these proteins plays an important role in focal adhesion in signaling pathways. These proteins also represent the most important sites of intracellular tyrosine phosphorylation such that focal adhesions are stained with anti-phosphotyrosine antibodies after plating cells on the extracellular matrix.

[0007] Even if proteins involved in the formation and disassociation of focal adhesions were rapidly identified, their roles have only begun to be elucidated slowly. Recent studies have shown that there is a direct correlation between FAK activity and the rate of cell migration, a consequence of hyperphosphorylation of p130cas by Src kinase (13). ). V-Src, a virus-type Src, also increased the turnover of focal adhesion in transformed cells, indicating that the catalytic activity of v-Src is required for this turnover (Reference 14).

[0008] Regarding the role of the protein tyrosine phosphatase (PTP) group, generally not much is known, and particularly its role in cell migration is not known. One exception is Yersinia YopH, a bacterial PTP. Once YopH is translocated into cells by Yersinia infection, p130cas and FA
K is dephosphorylated and destabilizes focal adhesion (Reference 8). From this, it is clear that p130cas plays an important role in focal adhesion.

The present inventors have proposed a PTP, PTP-PES, which uses p130cas as a main substrate.
The role of T was studied (Reference 6). PTP-PEST is rich in the catalytic domain of typical phosphatases and several prolines that have been shown to interact with several signaling proteins such as p130cas, Grb2 (4) and paxillin (15). Including the region. Furthermore, the inventors have also reported that the NPLH motif is responsible for constitutive binding between PTP-PEST and the PTB domain of the adapter protein Shc (Reference 5).

[0010] PTP-PEST is known as a cytoplasmic protein, like other members of the PTP family (Reference 3). It was shown to translocate around the cell membrane.

[0011] The identification of physiological substrates for the protein tyrosine phosphatases is an important step in understanding the function of these enzymes. Mutations of invariant amino acids within the catalytic domain of the PTPases produce an enzyme that is inactive, but nonetheless can form a complex with a tyrosine phosphorylated substrate.

As a means for isolating a virtual substrate of the PTP group, a mutant in which the cysteine at position 215 in the catalytic domain of PTP1B is replaced with serine or the aspartic acid at position 181 is replaced with alanine is a substrate trapping intermediate. (Reference 2). Although such mutants are important tools in studying the function of PTPs,
Unfortunately, it is necessary to increase the phosphotyrosine content of the protein source before performing substrate trapping experiments. Two approaches have recently been used as the method. One is a method of physiologically stimulating cells with EGF or the like to induce a cascade involved in tyrosine phosphorylation (Reference 4). However, this method tyrosine phosphorylates only a limited number of proteins. The other is a method in which cells are treated with pervanadate to inhibit the PTPase group in the cells (Reference 6) and tyrosine phosphorylate a very large number of proteins (Reference 6).
However, some of the phosphorylated tyrosines are not physiologically phosphorylated tyrosines, which can lead to incorrect results. Overall, it is unlikely that these two approaches will be able to identify all of the authentic substrates of PTPase, so alternative methods need to be considered.

[0013] Given these circumstances, it is necessary to study the physiological significance of phosphotyrosine phosphatases and their substrates. Furthermore, in order to develop therapeutic agents, it is necessary to elucidate the role of PTP-PEST in cell migration and provide drugs that can interfere with it.

SUMMARY OF THE INVENTION An object of the present invention is to provide a novel therapeutic agent for treating a disease caused by any of cell proliferation, cell migration, inflammation and angiogenesis.

[0015] Such novel therapeutics include complex formation of protein tyrosine phosphatase (ie, phosphotyrosine phosphatase) PEST (PTP-PEST) with signaling molecules involved in cell migration, focal adhesion and cell proliferation. Can be derived from substances related to

In a particular embodiment of the invention, the signaling molecules are p130cas and paxillin, the former being the only substrate known to date for PTP-PEST.

[0017] Studies of the binding capacity of these two substrates have led to the discovery of peptides that retain the ability to bind to the substrate. These peptides are prototypes of drugs that can interfere with the binding of PTP-PEST to the substrate. These peptides are thought to be therapeutic agents that can compete with wild-type PTP-PEST for binding to substrates and other signaling molecules.

It is another object of the present invention to find a true substrate for enzymes such as phosphatases. Fibroblasts that have abolished PTP-PEST expression have a series of tyrosine hyperphosphorylated proteins. This hyperphosphorylated state can be detected by using a ligand such as an anti-phosphotyrosine antibody. Such null mutant cells (
That is, cells in which expression was abolished) were used together with a substrate trapping mutant enzyme in which the catalytic site was inactivated. As a result, in fibroblasts, among the hyperphosphorylated proteins, p130cas was identified as the only substrate of PTP-PEST.

[0019] The following description of specific embodiments of the invention, the present invention will be described with reference to specific embodiments and the accompanying drawings, the purpose is to illustrate the invention, limiting the Absent.

Using a fibroblast cell line from which PTP-PEST was removed by gene targeting, the cell migration ability and the number of focal adhesions in homozygotes and heterozygotes for this mutation were examined. Homozygous cells for removal {hereinafter referred to as “PEST (− / −) type cells”} showed a marked decrease in migration and an increase in the number of focal adhesions. Starting from this fact, the phosphorylation state of important tyrosine phosphorylated proteins found in focal adhesion was confirmed. Was found to be highly phosphorylated. Recently, paxillin has been shown to bind to PEST (Reference 15), but it is not known whether it is a physiological substrate for this phosphatase (ie, PEST). This suggests that PTP-PEST plays an important role in cell migration, probably by directly or indirectly assisting focal adhesion disassociation, via dephosphorylation of p130cas and / or paxillin. And came to the conclusion.

In experiments involving PEST (− / −) type cells, cells with two nuclei, often separated by a dividing furrow, were observed, but we have found that this morphology has been shown to slow down the process of cell division. It was first associated with a migration disorder that increased the proportion of cells in M phase (middle phase). Recently, a protein has been cloned that is also a cleavage groove binding protein and a substrate for PTP-HSCF, another phosphatase similar to PEST (Reference 16). This protein is called PSTPIP and PTP-
It is associated with a domain of HSCF, which domain is also found in PTP-PEST. Based on this fact, the phosphorylation state of PSTPIP in (-/-) type cells was confirmed, and it was found that this protein was also highly phosphorylated. From this, regardless of whether PSTPIP is a direct substrate,
It was concluded that PTP-PEST may also be involved in the reorganization of the cytoskeleton during cell division. This suggests that inhibition of the binding of PTP-PEST to its substrate provides a useful tool for treating disease (anti-inflammatory, anti-angiogenic and / or anti-tumor agents). It reinforces the assumptions made.

Example 1 Preparation of knockout mouse PTP-PEST (-/-) According to the genomic mapping of the PTP-PEST locus, from the fourth exon to the seventh exon of the PTP-PEST gene (an essential domain in the PTPase catalytic domain is A targeting vector with the coding region removed was constructed. Subsequently, when 2 × 10 ⁷ embryonic stem cells J1 (References 13 and 14) were subjected to electroporation, 7 out of 223 clones lost one of the wild-type alleles, and Blot analysis showed a 7.0 kb diagnostic band. When chimeric mice were made using two independently isolated (+/-) type cell lines, the target allele contributed to the germline in both chimeric mice. When the heterozygous mutants were crossed, heterozygous mutants appeared at a ratio of 2 to wild type 1, and no homozygous mutants were found. This indicates that deletion of the PTP-PEST gene is lethal to the embryo. Heterozygous mutants have a normal lifespan and show no pathological symptoms throughout their life. Pregnant female mice of the heterozygous mutant were dissected and the homozygous mutant was found to be abnormal at day 9.5 (9.5 dpc).
) It turned out to happen. Visual observation of homozygous mutant embryos often reveals apparent delay in tail development, failure of neural tube closure, and failure of embryonic turning by day 9.5 . Some embryos continue to develop until day 10.5, but few such embryos and are significantly smaller.

Histological analysis of various embryos revealed the following. 7.5
At day, normal gastrulation is observed with three morphologically normal germ layers (FIG. 3.1). In addition, typical yolk sac formation is observed in all embryos (FIG. 3.2).
b). Signs of developmental abnormalities are seen at day 8 and development of neuroectodermal epithelium is impaired (data not shown). 9.5 When the homozygous embryos are examined,
Clear failures are detected in late neural tube development and segment formation (FIGS. 3.3-3)
. 5). Such failure is significant in the tail. In the embryo that survived to day 10.5 and resulted in embryo folding, remarkable degeneration of the mesenchyme and the neuroectoderm epithelium occurred. Surprisingly, such embryos are observed as deformed embryos with receding tails (FIG. 3.6). This is confirmed from embryo tissue sections. A striking mesenchymal symptom sometimes seen in homozygous mutant embryos is visceral lobe shrinkage / deficiency in the underdeveloped gastrointestinal epithelium (FIG. 3.6). In the mutant embryos, the mesenchyme with the lateral septum is reduced and degenerated (FIG. 3.7). The hepatic cavity (hepatic diverticulum)
No development of fetal liver occurred (Fig. 3.8).

[0024] The present inventors described the mechanism of liver development in particular, because of the observations described last and the fact that PTP-PEST is expressed in the developing liver and adult liver. Was studied. First, the present inventors searched for the presence of mRNA characteristic of hepatocytes by in situ hybridization, and focused on the expression of albumin mRNA. This is because albumin mRNA is expressed during liver development (hepatic specification) and its expression is restricted to hepatocytes in the developmental stage (Reference 41). In wild-type embryos, no more than 9
. On the fifth day, expression of albumin mRNA was observed, whereas PTP-PES
No expression of albumin mRNA was observed in embryos of the T null mutant. This indicates that the development of the liver was disrupted in the very early stages. This is the same for a 10 dpc embryo. At this stage, the liver is one of the major organs in normal embryos (data not shown). The present inventors also performed the same analysis on the expression of the HNF-3a gene. The HNF-3a gene is normally found in the developing liver, gastrointestinal tract, notochord and neural tube floor (42-45). As in the case of albumin, PTP-PEST (-/
HNF-3amRNA was not detected in hepatocyte-like cells among the −) type cells. However, the gastrointestinal tract, notochord and neural tube floor were positive in staining for HNF-3amRNA. This indicates that the development of these organs is relatively unaffected, at least at the superficial level, by the absence of PTP-PEST.

It remains possible that such failures are caused by indirect effects on vascularization. Although the initial vasculature in the homozygous mutant embryo appears normal, the following experiments were performed to more accurately demonstrate the state of the vasculature. Heterozygous mutant animals are treated with endothelial cell-specific t
It was crossed with a transgenic mutant expressing the lacZ reporter gene under the transcriptional control of the ek promoter. The resulting transgene-bearing and embryos that are heterozygotes for PTP-PEST removal are crossed back to carry the transgene lacZ-tek and PTP-PE
An embryo of a PTP-PEST E9 mutant homozygous for ST removal was obtained.
The embryo was stained with Xgal, and the whole and cross section were observed. In this homozygous mutant, the occurrence of vasculature was normal although a slight delay was observed. This confirmed that the specific appearance of PTP-PEST (-/-) type embryos was not caused by angiogenesis.

The appearance of hyperphosphorylated P130cas in (− / −) type fibroblasts is an important indicator of the physiological relevance of the interaction between PTP-PEST and p130cas. The following experiments were performed to demonstrate whether substances that could be p130cas and other substrates could be detected. Wild type and (− / −) type embryos were isolated, dissolved directly in lysis buffer for SDS-PAGE, and subjected to Western blot analysis using anti-phosphotyrosine antibody. The hyperphosphorylated substance was mainly detected as a band having a molecular weight of 130 kDa.

Example 2 In PTP-PEST (-/-) type cells, cell migration and focal adhesion
Materials and Methods Affected by Antibodies Antibody to the C-Terminus of PIP
Polyclonal antibodies are described in reference 75. Mouse anti-cortactin monoclonal antibody is available from J. Virginia University of Charlottesville, VA. Donated by Dr. Thomas Parsons.
The following antibodies were purchased from the listed vendors. Rabbit anti-FAK antibody (A
17) and rabbit anti-p130cas antibody (C20): Santa Cruz Biotechnology, Inc. (Santa Cruz, CA); mouse anti-paxillin antibody: Transduction Laboratories (Tr
ansduction Laboratories) (Lexington, Kentucky); Mouse anti-Vinculin antibody (hVIN-1): Sigma (St. Louis, MO); Horseradish peroxidase (HRP) binding-anti-mouse IgG antibody, HR
P-binding-anti-rabbit IgG antibody and FITC-binding-anti-mouse antibody: Jackson
Jackson Immunoresearch laboratories, Inc.
.) (West Grove, PA); anti-phosphotyrosine antibody: Mouse HRP binding-anti-phosphotyrosine antibody (pY20) manufactured by Transduction Laboratories and mouse anti-phosphotyrosine monoclonal antibody (4G10) were used.

Cell Culture PTP-PEST (+/-) type cell lines and PTP-PEST (-/-) type cell lines are as described previously. These cells contain 10% serum, L
-Maintained in DMEM supplemented with glutamine, penicillin / streptomycin.

Wound-healing migration assay The ability of each cell line to migrate on fibronectin was determined as follows. Falcon Chamber Slide (Becton Dickinson
Kinson) (Franklin Lakes, NJ) was purchased from Fibronectin in phosphate buffered saline (PBS, 10 mM sodium phosphate and 14 mM).
0 mM NaCl-containing, pH 7.4) solution (Fibronectin concentration: 10 μg / m
1) at 4 ° C. overnight. Using this Falcon chamber slide, cells were seeded at 60% confluence in a normal medium (containing 10% serum). After the cells adhered, the cell monolayer was scratched with a sterile plastic 1 ml micropipette tip. Thereafter, each well was washed, and thereafter, a new normal medium was supplied to each well every day. Three days later, the cells were fixed with a PBS solution of 0.2% glutaraldehyde at room temperature for 5 minutes, and the photograph was taken using a low-magnification phase-contrast microscope. Then, the degree of cell migration to the injured site was qualitatively evaluated.

After culturing the cells on the above-mentioned slide coated with fibronectin for 20 minutes or 3 hours, the cells are fixed with 4% (w / v) paraformaldehyde (PFA) in PBS for 5 minutes at room temperature. did. Then, Triton-X100 was added to 0.1% (v / v
) With a 4% PFA solution for an additional 20 minutes to permeabilize the cell membrane. Thereafter, the slides were treated with PBS containing 1% (w / v) bovine serum albumin for 20 minutes to block. After incubating the cells with an anti-vinculin antibody (diluted 1/400 in PBS) for 1 hour at room temperature,
Washed three times with BS, FITC-conjugated anti-mouse antibody (diluted 1/200) and TRI
TC Binding-Phalloidin (Molecular Probes, In
c.) (Eugene, Oreg.) (4 μl per well). After three further washes with PBS, the coverslips were washed with glycerol and 2.
The cells were mounted with PBS containing a 1: 1 mixture of 5% 1,4-diazabicyclo- [2,2,2] -octane (DABCO) (manufactured by Sigma), and the cells were observed using a Nikon fluorescence microscope.

The division furrow was stained as follows. Cells were seeded at low confluence on uncoated glass slides and left overnight in a medium containing 10% serum. Thereafter, fixation of cells, permeabilization of cell membrane and blocking were performed in the same manner as described above. Next, the cells were treated with a rhodamine-phalloidin complex (4 μl per well) for 1 hour, and washed and mounted in the same manner as described above.

Protein immunoprecipitation and immunoblotting The cells in the petri dish were washed with PBS supplemented with 1 mM sodium vanadate, collected, lysed, and lysed.
(Hercules, California)). In the case of immunoprecipitation of paxillin and cortactin, cells were lysed using RIPA buffer (50 mM).
Tris-HCl (pH 7.2), 150 mM NaCl, 0.1% SDS,
0.5% sodium deoxycholate and 1% Nonidet P40). In the case of immunoprecipitation of FAK, PSTPIP and PSTPIP2, cell lysis was performed using HNMETG (50 mM HEPES (pH 7.5), 150 mM NaC
1, 1.5 mM MgCl ₂ , 1 mM EGTA, 1% Triton X-100 and 1
(With 0% glycerol). In the case of immunoprecipitation of vinculin and p130cas, the buffer used for cell lysis was TNE buffer (10 mM Tris-HCl (pH 7.5), 1 mM EDTA, 150 mM
NaCl and 1% Nonidet P40). Each of the buffers used for cell lysis was prepared by adding 1 mM sodium vanadate and a COMPLETE protease inhibitor complex (manufactured by Boehringer Mannheim).

[0033] Each immunoprecipitation was performed as follows. 500 μg of total cell lysate was added to 1 μl
g of antibody and 30 μl of protein G agarose beads (manufactured by Gibco BRL) in a cell lysis buffer (supplemented with sodium vanadate and protease inhibitor) at 4 ° C. for 90 minutes, and the beads are incubated. Washed three times with each cell lysis buffer (final wash at 4 ° C. for 15 minutes). However, in the case of immunoprecipitation of FAK, PSTPIP and PSTPIP2,
HNTG buffer (10 mM HE buffer) instead of HNMETG which is a cell lysis buffer
Washed with PES (pH 7.5), 150 mM NaCl, containing 0.1% Triton X-100 and 10% glycerol. The washed beads were subjected to SDS-PAG.
E was redispersed in 30 μl of the loading dye.

[0034] Western blotting was performed according to the following common procedure. After transferring the protein onto the PVDF membrane, a blocking buffer (1% BSA, 1% (v /
v) blocking with goat serum, PBS containing 0.1% Tween)
Experiments were performed using the above blocking buffer containing the corresponding antibodies. The dilution of the primary antibody used is as follows. 1: 3000 for 4G10; 1: 10,000 for anti-paxillin antibody; 1: 1,000 for anti-vinculin antibody, anti-p130cas antibody, anti-cortactin antibody and HRP-conjugated-pY20; 1: 100 for anti-FAK antibody.

The dilution of the secondary antibody was 1: 10,000. For detection, NEN (registered trademark)
A Renaissance (registered trademark) chemiluminescence kit manufactured by Life Science Products (Albany, NY) was used.

Results PTP-PEST (− / −) cells have reduced motility on fibronectin. Recently, there have been two reports suggesting an indirect correlation between the degree of phosphorylation of p130cas and the rate of cell migration. In these articles, transfection of kinases into cells is performed. In one paper, v-Src was used as a kinase. p130cas is a direct substrate of v-Src (Reference 14). In another article, FAK is overexpressed as a kinase. When activated, FAK binds to both c-Src and its substrate, p130cas (13).

P130cas is a substrate of PTP-PEST in humans (Reference 6) and mouse, and p130cas is a pTP-PEST in cell lines from which PTP-PEST has been removed by gene targeting.
Since 130 cas is constitutively hyperphosphorylated, we believe that this P
It was decided to investigate whether the absence of TP would alter the motility of these cells under conditions where FAK and p130cas were activated.

Two cell lines, heterozygous and homozygous for PTP-PEST deficiency, were used. In order to minimize the dominant negative effects of the targeted allele, a comparison was made between the two cell lines, heterozygous and homozygous.

The simplest way to qualitatively compare cell migration is the wound healing migration assay. Fibroblast monolayers formed on glass slides coated overnight with fibronectin were wounded at 37 ° C. and fixed 36 hours later. The typical appearance of each cell line, obtained from five independent experiments, is shown in FIG. P
TP-PEST (+/-) type cells can migrate to the wound faster than the front end of cells that are pushed out by proliferation (FIG. 4 (a)). On the other hand, PT
P-PEST (-/-) type cells do not show any chemomotility (FIG. 4 (b)) and cannot penetrate the wound.

In PTP-PEST (− / −) type cells, the number and size of focal adhesions are increasing. The cell cytoskeleton plays an important role in cell motility. Thus, the present inventors investigated whether the migration impairment observed in homozygous cells was due to abnormal actin filament construction or to apparent changes in focal adhesion. Cells of each cell line are seeded on fibronectin, actin filaments are stained with rhodamine-bound phalloidin, and mouse anti-vinculin monoclonal antibody and fluorescein-bound-anti-mouse secondary antibody are used to indirectly detect focal adhesions by indirect fluorescence. Visualized by antibody method (hi
ghlighted).

After 25 minutes, folds of cell membrane and filopodia were observed in all cell lines. This suggests that movement disorder is not caused by the inability to polymerize actin or construct the structure of actin. In all cell lines, large immature focal adhesions were observed. This indicates that even in PTP-PEST (-/-) type cells, the initial stage of the pathway for forming these contacts is not affected. (FIG. 5 (a)-
(D))

However, when the cells were left on fibronectin for 3 hours prior to fixation and staining, the homozygous cells remained widespread and many large focal adhesions were scattered throughout the lower surface of the cells. Also, a small actin fiber like hair surrounded the cells. This is a characteristic of cells in the early stages of apolar migration. On the other hand, the PTP-PEST (+/-) type cells were spherical, the edges were clear, the focal adhesion was small, and the cells were concentrated at the periphery of the cells.

Differences in the size and number of focal adhesions between two cell lines differing only in the presence or absence of constitutively highly phosphorylated PTP-PEST of p130cas and paxillin in PEST (− / −) type cell lines For understanding, we examined the tyrosine phosphorylation status of certain focal adhesion proteins (FIG. 6).

The adapter protein p130cas was used as a positive control because it is a substrate for PTP-PEST and has been shown to be hyperphosphorylated.

Other proteins obtained by immunoprecipitation and using an anti-phosphotyrosine antibody as a probe include cortactin and paxillin. It has been shown that paxillin is hyperphosphorylated in FAK knockout cells that are known to be involved in decreased cell mobility and increased focal adhesion (Reference 33). Paxillin phosphorylation has also recently been shown to be necessary for cell spreading and formation of focal adhesions (46). Again, paxillin was found to be in a hyperphosphorylated state. Paxillin has recently been shown to bind to PEST (Ref. 30), but it is still under investigation whether it is a direct or indirect substrate of PTP.

Vinculin is a structural protein that binds to tailin and actin during focal adhesion formation. Vinculin can be tyrosine phosphorylated, but this phosphorylation is not thought to be involved in the formation of focal adhesions. The formation of focal adhesion is thought to be due to a change in the steric structure of vinculin that occurs after the binding of phosphatidylinositol diphosphate (Reference 12). Vinculin tyrosine phosphorylation was not at detectable levels in any of the cell lines, presumably because it did not stimulate the cells (data not shown).

The last component to be investigated as a component involved in focal adhesion is FA
K. The role of p130cas in cell migration has been shown to be in a pathway triggered by FAK (13), which binds to integrins, forms clusters and is activated by transphosphorylation. You. FA
Phosphorylation of tyrosine 397, the Src kinase binding site of K, has been shown to play an important role in cell migration. FIG. 6 shows that FAK is not hyperphosphorylated in PEST knockout cells.

Effect of PSTPIP on high phosphorylation and fission groove formation in PEST (− / −) type cells. In studying the effect of PEST targeting on cell migration, another protein involved in the actin cytoskeleton, PSTPIP, was cloned and shown to be a substrate for PTP HSCF, a phosphatase of the PEST family (Reference 16). The region presumed to be the coiled coil region of PSTPIP is PTP H
Interacts with the C-terminal proline-rich region of SCF. This C-terminal region is
This region is present in all members of the PEST family, including PTP-PEST.

The present inventors performed immunoprecipitation using a polyclonal antibody against PSTPIP, and further performed Western blotting using an anti-phosphotyrosine antibody to examine the tyrosine phosphorylation level of PEST (− / −) type cells. .
FIG. 7 shows that PSTPIP is hyperphosphorylated in knockout cell lines, but it is still under investigation whether PIP is a direct or indirect substrate. Similar experiments show high homology with PSTPIP, but with SH
This was also performed for PSTPIP2, a protein without three domains.
The phosphorylation level of PSTPIP2 was similar between (+/-) and (-/-) type cell lines (data not shown).

PSTPIP has homology with CDC15 of Schizosaccharomyces pombe , which is an actin-binding protein involved in the formation of the dividing cleft during cell division (Reference 47). In an experiment using the PEST (-/-) type cell line, cells thought to have blocked cytokinesis were highly observed. Specifically, the two daughter cells are almost independent, but still attached to each other by actin-rich conjugation. This phenomenon is presumed to be a result of the prolongation of the M phase because actin is not reconstituted during cell division. This phenomenon, although not quantified, was frequently observed in dividing (-/-) cells. FIG. 8 shows actin cleft rings stained by the rhodamine-phalloidin complex.

Discussion The activation of focal adhesion kinase (FAK) and the formation of focal adhesion are closely related phenomena. The order in which these phenomena occur following integrin stimulation has been largely debated until recently. Recent experiments (48 and 49) have shown that FAK activation is the result of focal adhesion formation. Specifically, physical expansion occurs due to the formation of intracellular stress fibers, and under the control of Rho GTPase, the bound integrins form clusters and interact with other structural proteins such as α-actinin and tensin. Combined (Reference 12). Because FAK proteins bind to integrins, they are in close proximity to each other and, like receptor tyrosine kinases, trans-autophosphaorylate following binding of extracellular ligands.
can do. This phosphorylation activates FAK and provides a binding site for other signaling proteins that amplify the signaling pathways triggered by integrins.

However, the formation of focal adhesion and the activation of FAK are not strictly proportional. Some signals transmitted by the FAK are known to feed back information and contribute to the turnover speed of focal adhesion. The presence of such a signal was determined by an increase in the size and number of focal adhesions in the FAK knockout cell line (Reference 33). Recently, FAK has been shown to play a role in cell migration, a phenomenon that requires the formation of focal adhesions. Increased cell migration is a phenomenon that occurs through tyrosine phosphorylation of the adapter protein p130cas (Reference 13).

The present inventors have examined cell lines to which another gene, specifically the cytoplasmic protein phosphatase PEST, has been targeted. In both human and mouse, the main substrate was shown to be p130cas (Reference 6).
. Following the activation of integrin, PEST has recently been shown to migrate to the periphery of the cell membrane, but this migration of PEST reaches the substrate of PEST that has migrated to focal adhesion with FAK activation. (Reference 9
).

The effect of PTP-PEST removal on cell migration and the size and number of focal adhesions was similar to the phenotype of FAK knockout cells. In each case, when fibronectin was used as extracellular matrix, reduced cell migration and the presence of large, immature focal adhesions dispersed on the lower surface of the cells were seen (FIGS. 4 and 5). . Although having p130cas as a common substrate, FAK is a kinase and PTP-PEST is a phosphatase. However, PEST nevertheless appears to enhance rather than antagonize the effects of FAK on cell migration. This suggests the existence of a mechanism that simultaneously causes both the formation path and the decomposition path of the focal adhesion to increase the turnover speed of the focal adhesion structure. If there is no pathway for decomposing the focal adhesion, the contact between the extracellular matrix and the cell becomes very strong, and the cell does not move while being adhered (Reference 50).

Interestingly, FAK knockout cells do not contain a hyperphosphorylated focal adhesion component, but do contain hyperphosphorylated proteins like the PEST knockout cells. FAK knockout cells show hyperphosphorylation of cortactin and paxillin (p130cas not tested). P
In EST knockout cells, together with p130cas, a physiological substrate,
Of these two, only paxillin was hyperphosphorylated.

It has been shown that paxillin phosphorylation is involved in the spread of cells on the extracellular matrix (Reference 46). Paxillin can bind to PEST (Reference 30), but it is not yet clear whether it constitutes a physiological substrate for this phosphatase.

It is not known exactly what role PEST plays in cell migration and formation of focal adhesions. The absence of a difference in the early stages of cell adhesion to fibronectin (FIGS. 6 (a)-(d)) suggests that PEST is not involved in the formation of such a structure. However, a difference was observed in the equilibrium state (FIGS. 3E to 3H). The focal adhesion of the knockout cells was large and arrowhead-shaped. This is a feature of immature focal adhesions. The presence of many such focal adhesions and their distribution below the cell membrane may explain the lack of motility in the wound healing migration assay. On the other hand, the heterozygous cells are rounded, and the focal adhesion like a dot is seen only at the tip of the cell membrane, which is consistent with the migration of the cells. This proves that PEST is involved in the decomposition of focal adhesions.Whether PEST is actively decomposing the focal adhesion structure or provides positive feedback on the formation of the focal adhesion structure. It is not clearly shown here if it is blocked.

Phenylarsine oxide (phenylar), an inhibitor of tyrosine phosphatase
Studies with sine oxide) showed that even after 16 hours of starvation, cell treatment with this compound was sufficient to induce the formation of stress fibers. Further, degradation of focal adhesions has been shown to result in activation of phosphatase activity which can be detected by assays using FAK and paxillin as substrates. Two active sites of the phosphatase close to each other
The thiol group of two cysteine residues reacts with phenylarsine oxide. PT
The catalytic domain of P-PEST contains the sequence 231-CSAGC-235 (Reference 3; SEQ ID NO: 3), and the cysteine at position 231 is essential for catalytic activity. Because of this and the fact that paxillin is hyperphosphorylated in PEST knockout cell lines, PTP-PEST may be a phosphatase involved in the degradation of focal adhesions under the conditions used in these experiments. It was suggested that the role might be similar to the role inferred for cell migration described above.

A model of the role of PTP in the degradation of focal adhesions suggests that phosphatase overexpression may also inhibit cell migration. However, in this case, it is considered that the formation of focal adhesion in the traveling direction of the cell is impaired. In fact, other Ps that can dephosphorylate p130cas
Experiments with PTP1B, a TP (37), showed that overexpression of this phosphatase in rat fibroblasts reduced cell migration while increasing the time required for cells to spread on fibronectin ( Reference 51).
This was associated with malformation of focal adhesions. Interestingly, even in such cells, a number of large focal complexes scattered on the lower surface of the cells, such as those found in PTP-PEST (-/-) type cells, eventually formed. These experiments and those of the present invention suggested that moderate PTP activity on p130cas is required for normal focal adhesion formation and cell migration. This is consistent with the role of PTP in focal adhesion turnover.

It is thought that PTP-PEST also plays a role in the control of the cytoskeleton via PSTPIP, which is a cleavage groove binding protein. PSTPIP
Was first identified as a binding partner and substrate for PEST tyrosine phosphatase, PTP-HSCF (Reference 16). PTP-HSCF is
Dephosphorylate tyrosine residues of PSTPIP modified either by co-expression of v-Src tyrosine kinase or by the presence of pervanadate, a non-specific PTP inhibitor. One such site located within the SHTP domain of PSTPIP is known to control binding to the proline-rich region found in WASP (Wiskott-Aldrich syndrome protein) (Reference 52). Regulation of the actin cytoskeleton is also known.

It is not known whether PTP-PEST directly binds to PSTPIP,
A peptide derived from the C-terminus of ST competes with PTP-HSCF for binding to PSTPIP (Reference 16). Also, no experiments have been performed to show that PTP-PEST dephosphorylates PSTPIP, but the fact that PSTPIP is hyperphosphorylated in PEST (− / −) type cells (FIG. 7) is due to the fact that PTP-PEST is highly phosphorylated in PEST (− / −) type cells. PEST is PSTPI
It suggests that it functions to regulate the phosphorylation level of P. Since PTP-HSCF and PTP-PEP, another member of the PEST family, are not expressed in fibroblasts, PTP-PEST acts on PSTPIP in fibroblasts as well as PTP-HSCF in hematopoietic cells. Could be. P
The exact site of hyperphosphorylation in STPIP is not known, and the effects of this hyperphosphorylation are still under investigation. One of the effects is an increase in the relative amount of M-phase cells in unsynchronized cultured cells (FIG. 8),
It may be due to the action of STPIP. However, this effect appears secondary since the growth rate of (-/-) cells is comparable to that of (+/-) cells. Also, such an increase in tyrosine phosphorylation of PSTPIP is a result of an increase in the number of cells in the M phase, and may not be the cause. But PSTP
Since the high phosphorylation of IP is more pronounced than the increase in the number of cells in metaphase,
The opposite is more likely.

In the present invention, a specific PTP in controlling the cytoskeleton of fibroblasts, ie, P
Two roles of TP-PEST were reported. The first role is probably p1
Decomposition of focal adhesions, presumably due to the dephosphorylation ability of 30 cas. This event is necessary for cell migration on an extracellular matrix such as fibronectin. When COS cells were seeded on fibronectin,
PTP-PEST was shown to be localized around the cell membrane, indicating that PTP-PEST has a physiological role in cell migration. This is not a secondary effect of overexpression. Most of PTP-PEST is cytoplasmic (cytoplas
mic pool) (ref. 3) and the fact that it can be moved as needed
, P130cas or v-Src are overexpressed, giving cells a way to increase the rate of focal adhesion turnover in proportion to the stimulus (13, 14, 53). PTP-PEST also plays a role in regulating the phosphorylation level of PSTPIP, a protein that binds to the cytoskeleton (Reference 16). Such a role of PTP-PEST activity is attributed to PSTP of WASP.
Binding to IP, function of PSTPIP in mitotic sulcus formation, or PTP
It is not known whether STPIP is involved in an unknown function.

Example 3 Identification of Substance Presumed to be a Substrate of PTP In a knockout mouse in which PTP-PEST was knocked out by gene targeting, the specific physiological substrate of the enzyme, PTP-PEST, was highly phosphorylated. Based on the premise that it should be present, a substance presumed to be a substrate for PTP-PEST was identified using (− / −) type fibroblasts obtained from the knockout mouse. Analysis of the (− / −) type cells for phosphorylated tyrosine confirmed the presence of hyperphosphorylated p180, p130 and p97. The p130cas protein was obtained from cells treated with pervanadate by PT
Since it was isolated using a substrate trap mutant of P-PEST (Reference 6), the tyrosine phosphorylation state of p130cas was analyzed. From the precipitate obtained by subjecting p130cas to a direct immunoprecipitation reaction, this protein was identified as (-/
−) Type cells were found to be highly phosphorylated. Also, PTP-PEST
Strong interactions were observed in vivo with the Cas-like molecules Hef1 and the SH3 domain of Sin. By combining a genetic technique and a biochemical technique, substances putative as substrates for PTP-PEST are identified. Indeed, Cas-like molecules Hefl and Sin are able to bind to PTP-PEST via their SH3 domains.
The present inventors have for the first time found that they bind to a proline-rich motif of, and thus they may be substrates for PTP-PEST. Based on the results of these studies, a method combining PTP gene targeting and substrate trapping technique can be reasonably applied to the identification of PTP-PEST and other PTPases.

In the process of searching for a PTP-PEST target, a substrate trapping experiment (Reference 6)
Identified the adapter protein p130cas as a substrate for PTP-PEST. Two "substrate trap" type variants, PTP-PEST D1
Using 99A and PTP-PEST C231S, the unique substrate p130cas was isolated from extracts of Helana cells treated with pervanadate. In experiments using mice, the ligand-dependent activation of the EGF receptor was used to generate the pp130 kD of the "substrate trap" mutant PTP-PEST C231S.
This opens the way to induce binding to protein a (which has not yet been confirmed but is likely to be p130cas (Reference 3)). Tyrosine phosphorylated p13 by PTP-PEST D199A in extracts of cells treated with pervanadate
Ocas recognition was recently found to be strongly inhibited by mutational inactivation of the proline-rich region (proline residue at position 337), and we have found that p130
A substrate recognition mechanism of PTP-PEST through the SH3 domain of cas was proposed (Reference 4).

P130cas has one SH3 domain, two proline-rich regions, 15
Domain serving as a substrate for tyrosine kinase containing the YxxP site of
, Reference 56) and a binding site (YDYV motif) to the SH2 domain of Src kinase. The SH3 domain of p130cas has been shown to interact with FAK and FAK-related non-kinases (FRNK) and with proline-rich sequences found in PTP1B (37, 54).
Reference 55). In a recent study, the SH3 domain of p130cas was found to be essential for targeting p130cas to focal adhesion in non-transformed cells (Reference 9). The substrate domain of p130cas is v-
Crk (Reference 57), Crk (Reference 58), Crk-II (Reference 59), CRKL
(Reference 60), Nck (Reference 61) and some other unidentified SH2 domain-containing proteins (Reference 57) can interact with the SH2 domain in a phosphorylated tyrosine-dependent manner. p130cas converts cells into EGF (Reference 58), NGF (
Reference 59), treatment with PDGF (Reference 62) and activation of integrins (
It is known that tyrosine phosphorylation is carried out by literature 63). p130c
The YDYV motif and the proline-rich region on the C-terminal side of as are the binding sites for the SH2 and SH3 domains of the tyrosine kinases belonging to the Src family, so that other tyrosine residues in p130cas are also mediated by these tyrosine kinases. It is phosphorylated (Reference 64).

As two proteins having high homology to p130cas, Hef1 / Cas-
L (Reference 65, Reference 66) and Efs / Sin (Reference 67, Reference 68). H
Since ef1 and Sin have a similar module structure to p130cas, they are adapter proteins belonging to the p130cas family. SH3 domains of Hef1 and Sin are 74% each of SH130 domain of p130cas.
And 64% homology. The substrate domains of Hef1 and Sin are 13
And eight YxxP motifs. Many of these YxxP motifs differ from the YxxP motif of p130cas, suggesting that Hefl and Sin may interact with other SH3 domain containing proteins.
Although expression of Hef1 and Sin is restricted to specific cell populations, p130cas is expressed in all cells. Hef1 is mainly expressed in hematopoietic cells (Reference 66). S
In mRNA is observed in high concentrations in mouse embryos (Reference 67), but the expression of Sin is not well known.

The identification of physiological substrates for protein tyrosine phosphatases is an essential element for understanding the biological function of this family of enzymes. The main objectives of this study were PTP gene targeting technology and PTP-PEST (paradigm
) Is to demonstrate a novel experimental approach in PTP substrate identification by combining it with a substrate trapping experiment. By using this approach, p130cas was isolated in large quantities by the PTP-PEST trapping mutant, thus supporting this approach. Another objective of this study was to determine that proteins belonging to the p130cas family, namely Hef1 and Sin, via their SH3 domains, proline-rich regions of PTP-PEST and p1
It is for the first time to clarify an interaction similar to 30 cas.

Materials and Methods Generation of Gene-Targeted Cell Lines for PTP-PEST. PTP-PEST (+/-) type and (-/-) type embryonic fibroblast cell lines are PTP-PEST heterozygous types obtained by specifically disrupting the PTP-PEST locus by homologous recombination in ES cells. It was established from primary cultures of embryonic fibroblasts isolated from (+/-) and homozygous (-/-) mouse embryos, respectively (Reference 1). Briefly, female mice of the PTP-PEST (+/-) type were transformed into PTP-PEST (+/-
Embryos were isolated 8.5 days after mating with male mice of type-) and treated with trypsin for 30 minutes at 37 ° C. The cells in the dispersion were isolated and cultured. After becoming confluent, the cells are separated and subjected to dominant negative mutant p53 (C153S) gene (Reference 25) and E1A / E using Lipofectamine (GibcoBRL).
Cells were immortalized by stable transfection of the 1B virus gene (obtained from Dr. Branton of McGill University) with a vector conferring puromycin resistance. Pool cells that survive treatment with puromycin,
Genotype is determined by Southern and Northern blots using conventional methods. For Southern blots, genomic DNA was digested with BamHI and the resulting fragment was 0.8
% Agarose gel. After transfer to Hybond N +, PTP-PE
A KpnI / SacI fragment of ST cDNA labeled with a radioisotope was blotted as a probe. For Northern blotting, RNA was separated on a 1% formaldehyde agarose gel, transferred to Hybond N +, and a fragment obtained by PCR from the PTPase domain of PTP-PEST labeled with a radioisotope was probed. For blotting. The probe was stripped from the blot, confirming that the amount of RNA loaded using GAPDH as the probe was equal.

Preparation of Gene Construct The mouse p130cas cDNA integrated into pBluescript was obtained from Dr. Steven H.
Presented by anks (Vanderbilt School of Medicine, Tennessee). p13
0cas cDNA was prepared by standard recombinant DNA techniques using pJFTAG, pBlues
It was cloned between the EcoRI and SalI sites of cript II KS (Stratagene) and pCDNA3 (Invitrogen). pJFTAG incorporates six copies of the c-myc epitope between the HindIII and EcoRI sites of pCDNA3. The SH3 domain of p130cas was cloned between the EcoRI and ApaI sites of pJFTAG. GST and p130c
As a construction of a fusion protein with the SH3 domain of as or Sin, a region encoding the SH3 domain is amplified by a PCR method using an oligonucleotide containing a BamHI site and an EcoRI site, and the resulting product is pGEX 2TK (Pharma).
Cia) by cloning between the BamHI and EcoRI sites. A fusion protein of GST and the SH3 domain of Hef1 was obtained from Dr. Yuzhu.
Zhang cloned into pGEXILAmbdaT, Dr. Erica Golem
gift from Fox Chase Cancer Center, Philadelphia.

Substrate Trapping Experiment 100 ng of GST pre-bound to glutathione sepharose beads
-PTP-PEST (amino acid residues 1 to 453) or GST-PTP-P
EST C231S (amino acid residues 1 to 453) was incubated with 1 mg of a lysate as an extract of the above cells for 90 minutes according to a known method (Reference 4). After thoroughly washing the beads with HNTG buffer, the protein was eluted with SDS-sample buffer. After performing SDS-PAGE and transferring the protein to a PVDF membrane, anti-phosphotyrosine monoclonal antibody 4G10 or rabbit anti-p130casB + F polyclonal antibody (University
of Virginia, a gift from Dr. Amy H. Bouton).

Cell culture, transfection and co-immunoprecipitation 293-T cells, COS-1 cells, PTP-PEST (+/−) type cells and P
TP-PEST (-/-) type cells were maintained in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% fetal calf serum and penicillin / streptomycin. 293-T cells were co-transfected with a plasmid having PTP-PEST and a plasmid having p130cas into a total of 20 pg transients using the calcium phosphate precipitation method described in Reference 5. In addition, 40 μg of pACTAG PTP-PEST was added to COS-1 cells growing exponentially in a 15 cm tissue culture dish using the electroporation method described in Reference 3.
Was transfected. To perform co-immunoprecipitation, the HA-
293-T cells expressing PTP-PEST protein or Myc-p130cas protein were collected and washed with PBS. Cells were transformed with the HN described in Reference 5.
Dissolved in METG. The protein complex was conjugated to the anti-PTP-PEST antibody 1075.
(Reference 3) and protein G-agarose beads (Gibco-BRL) were used for immunoprecipitation for 90 minutes at 4 ° C. After washing the immune complexes three times with HNTG, the proteins were eluted from the beads by boiling in SDS sample buffer for 5 minutes. The protein was separated by 11% SDS-PAGE, transferred to a PVDF membrane (Immobilon-P, Millipore), and analyzed for the presence or absence of Myc-labeled p130cas protein using an anti-Myc epitope monoclonal antibody 9E10. COS-1 cells were treated with pervanadate to increase protein tyrosine phosphorylation levels. Briefly, 360 μl of freshly prepared pervanadate (10 mM sodium orthovanadate in 50 mM hydrogen peroxide).
1 was added to 15 ml of DMEM and the cells were incubated for 15 minutes before harvesting the cells.

Examination of the presence or absence of binding A GST fusion protein composed of various different parts of PTP-PEST is known (Reference 3). GST fusion protein was purified using Glutathione Sepharose (Pharmacia) according to the protocol attached thereto. To perform a binding experiment in a liquid, 100 ng of GST alone or 100 ng of GST-
The SH3 domain fusion proteins (all of which were previously bound to glutathione sepharose) were incubated with 1 mg of lysate of COS-1 cells expressing HA-PTP-PEST for 90 minutes at 4 ° C. After washing with HNTG three times,
The protein was eluted with sample buffer, separated by SDS-PAGE,
The presence or absence of HA-PTP-PEST was analyzed by immunoblotting using the A5 antibody. For far western assays,
Purified PTP-PEST fusion protein was separated by SDS-PAGE and transferred to a PVDF membrane. The proteins on the blot were incubated in binding buffer (1
0 mM Tris-HCl (pH 7.4), 1 mM EDTA, 150 mM NaC
l, 1 mM DTT, 0.2% BSA)
I let it. The blot was probed by incubating the blot with a probe consisting of a [ ³² P] -labeled GST-SH3 domain fusion protein in binding buffer at 4 ° C. for 8-16 hours (for labeling of the probe, its manufacturer Pharma
Labeled according to the manual provided by cia). Blots were extensively washed in binding buffer at room temperature and exposed to X-ray film (Kodak) for 20 minutes.

Results A cell line with PTP-PEST gene targeting (PTP-PEST- /
Preparation of-) In order to identify a novel substrate of PTP-PEST by examining the content of phosphorylated tyrosine in cells deficient in PTP-PEST, male and female mice of PTP-PEST heterozygous (+/-) Mouse embryonic fibroblasts were obtained from mouse embryos isolated 8.5 days after mating. The resulting primary cultured cells were transformed with the C153S mutant p53 (Reference 69) and E1A / as described in the "Materials and Methods" section above.
Immortalization was achieved by transformation with the gene for the E1B protein. The genotype of the established cell lines was determined by Southern blot. FIG. 9A shows that one of the established cell lines
kb wild-type allele and 7 kb target allele (PTP-PEST +
/-) And the other established cell lines have the target allele (PTP-PEST- /
-) Only. As a positive control, using wild-type DNA isolated from embryonic fibroblasts, only the wild-type allele was observed. PTP-PEST
Northern blot analysis was performed to confirm the absence of PTP-PEST mRNA in the (− / −) type cell line. In FIG. 9B, PTP-PEST
It is shown that 3.8 kb of PTP-PEST mRNA is present in the (+/-) type cell line, but not in the PTP-PEST (-/-) type cell line (upper drawing). reference). As a positive control, RNA isolated from embryonic fibroblasts
Was used, the presence of a 3.8 kb band was observed. As a control for electrophoresis, the probe was removed from the blot and subjected to blotting using GAPDH as a probe (see the lower drawing in FIG. 9B). Phenotypic characterization of PTP-PEST deficient cell lines is currently underway.

Analysis of Phosphorylated Tyrosine Status of PTP-PEST (− / −) Type Cells In order to identify substances that are presumed to be substrates for PTP-PEST, PTP-
The status of phosphorylated tyrosine in PEST (-/-) type cells was analyzed. By comparing immunoblotting of total cell lysate (TCL) of (+/−) type cells with an anti-phosphotyrosine antibody and that of (− / −) type cells, about 180 kDa and about 13 kDa were obtained.
It was found that hyperphosphorylated proteins of 0 kDa and about 97 kDa were present in the (-/-) type cells (see the first two lanes in Fig. 10A). The TCLs of the (+/-) type cells and the (-/-) type cells were respectively subjected to immunoprecipitation using an anti-phosphotyrosine antibody, and analyzed by Western blot with an anti-phosphotyrosine antibody.
The same hyperphosphorylated protein was detected (see the last two lanes in FIG. 10A).

To identify these highly phosphorylated proteins, the putative protein in
In vivo immunoprecipitation was performed. PTP-PEST (+/-) type cell line and (-/
P130cas was immunoprecipitated from both −) type cell lines and the phosphorylation status of p130cas was analyzed by anti-phosphotyrosine immunoblotting using the 4G10 antibody. In (-/-) type cells deficient in PTP-PEST, PTP
-P130cas compared to p130cas in PEST (+/-) type cells
Was in a hyperphosphorylated state (upper drawing in FIG. 10B). The presence of the same amount of p130cas protein in the immunoprecipitate and total cell lysate (TCL) is shown in FIG.
As shown in B (see lower drawing), the probe was stripped from the blot and confirmed by blotting again using a rabbit anti-p130cas polyclonal antibody as a probe. The inventors have previously shown that PTP-PEST has been tested for E via Grb2.
Since it was shown that a complex was formed with the GF receptor (Reference 3), the immunoprecipitate obtained from the (+/-) type cells using an anti-phosphotyrosine antibody and the The amount of EGF receptor in immunoprecipitates obtained using a phosphotyrosine antibody was compared. The results showed that the amounts of EGF receptors in these immunoprecipitates were equal to each other (data not shown). As a 97 kDa protein presumed to be a substrate of PTP-PEST, a p130cas-like protein He
For f1 and Sin, the amount of these proteins in immunoprecipitates obtained from (+/−) type cells using anti-phosphotyrosine antibody and in immunoprecipitates obtained from (− / −) type cells using anti-phosphotyrosine antibody Was analyzed. He in these immunoprecipitates
f1 and Sin were not detected (data not shown). Furthermore, the inventors have found that Hef1 and Sin are not expressed in the PTP-PEST (+/-) type fibroblast cell line and the (-/-) type fibroblast cell line used in this study.

Identification of Physiological Substrates by Combining PTP-PEST Gene Targeting and Substrate Trapping Experiments By inactivating the PTP-PEST gene (ie, PTP-PEST (− //
Tempertures that exhibit high specificity for PTP-PEST (in −) type cell lines) are assumed to be hyperphosphorylated and their substrates are determined using a “substrate trap” mutant of PTP-PEST (C231S mutant). It is believed that it can be isolated. An important property of the novel approach of the present invention is that cells are placed in a physiological (
(Growth factor) or artificial (pervanadate). Substrate trapping experiments were performed in PTP-PEST (+/-) and (-/-) cell lysates using the GST-PTP-PEST C231S mutant.
As shown in FIG. 11 (a), the amount of the phosphorylated pp130 protein isolated using the GST-PTP-PEST 231S mutant was small in the PTP-PEST (+/-) type cell lysate, and the amount of PTP- In the case of using GST-PTP-PEST WT, PTST (-/-) type cell lysate showed a large amount.
Phosphorylation substrates could not be isolated from either P-PEST (+/-) type or (-/-) type cell lysates. As a positive control, a similar substrate trapping experiment was performed on lysates of pervanadate-treated COS-1 cells, and PTP-P
The EST130 protein was used to isolate the pp130 protein. FIG. 11 (
(b) shows the results obtained by removing the probe from the above plot and analyzing the presence of the p130cas protein using an anti-p130cas B + F antibody as a probe. The p130cas protein was mainly detected in the lanes using the C231S mutant on lysates of PTP-PEST (− / −) type cells and pervanadate treated cells. Interestingly, the same amount of p130cas protein was obtained from PTP-PEST (-/-) type cell lysate and pervanadate-treated COS-1 cell lysate by affinity purification using C231S mutant of PTP-PEST. Nevertheless, the tyrosine phosphorylation level of p130cas trapped from the PTP-PEST deficient cell line was assessed by densitometry, despite the trapping from the PTP-PEST deficient cell line (FIG. 11 (c)). It was approximately 20 times higher than p130cas.

PTP-PEST Binds to SH130 Domain of Hefl and Sin, p130cas-like Proteins The SH3 domain of p130cas interacts with PTP-PEST in a substrate recognition mechanism (Reference 7). p130cas-like proteins Sin and He
Since these SH3 domains are present in the PTP-PE
The present inventors considered the possibility of interacting with ST. p130 and S of Sin
The H3 domain was cloned into pGEX-2TK, the SH3 domain of Hefl was cloned into the pGEXILAmbdaT vector, and G3 was used to perform a liquid binding assay using PTP-PEST.
It was expressed as an ST fusion protein. Correct expression of the fusion protein used in the binding assay is shown in the Coomassie blue stained gel shown in the right panel of FIG. As is clear from the left drawing of FIG.
HA-PTP-PEST specifically bound to the purified Hefl and SH3 domains of Sin, even after extensive washing of the complex. GST alone, used as a negative control, did not bind to HA-PTP-PEST, whereas the SH3 domain of p130cas, used as a positive control, bound to HA-PTP-PEST.

The first proline-rich region (amino acids 316 to 346) of PTP-PEST has a proline-rich motif in PTP-PEST mapped as a binding site to the SH3 domain of Sin and Hefl. Five were found (Reference 3), S
Since the H3 domain is known to bind to a proline-rich motif,
We investigated which motifs mediate the interaction with the SH3 domain of Hefl and Sin. Five different proline-rich regions of PTP-PEST (
A group of GST fusion proteins with Pro1-5) were created for the purpose of mapping the PTP-PEST region essential for interaction with the SH3 domain of Cas-like proteins Sin and Hefl in a far-western binding assay. Schematic representations of the fusion proteins used in the Far Western assay and Coomassie blue stained gels to show correct expression are shown in FIGS. 13A and 13B, respectively. PTP-PE for Far Western assay
The ST fusion protein was separated by 11% SDS-PAGE, transferred to a PVDF membrane and blocked, and then [ ³² p] -labeled Sin, Hefl or p13.
Blotting was performed using each SH3 domain of 0 cas as a probe. S
The SH3 domain of in and Hefl is the Pro1 region of PTP-PEST (amino acids 316 to 346 of mouse PEST (SEQ ID NO: 4) or human PEST).
317-347 (SEQ ID NO: 5) with high affinity and, as shown in FIGS. 13D and 13E, the complete C-terminus of PTP-PEST (second
It also bound to a fusion protein containing amino acids 76-775) and the N-terminus (amino acids 1-453). These three fusion proteins are PTP-P
Both include the EST Pro1 region. Binding of the SH3 domain with Pro2-5 and GST alone (negative control) of PTP-PEST did not reach detectable levels even after prolonged exposure. As a positive control, SH130 of p130cas
The same blot was blotted using the domain as a probe, and similar results were obtained (FIG. 13A). These results indicate that the SH3 domain of p130cas, Sin and Hefl belonging to the adapter protein family is PTP-PEST
Of the Pro1 region with high affinity.

PTP-PEST and p130cas bind in vivo Conventionally, the binding of PTP-PEST and p130cas is a GST fusion protein (
Only in vitro was studied using either the present example) or a protein produced by a baculovirus (References 6, 11). p130cas and PTP-PES
To evaluate in vivo binding to T, co-immunoprecipitation experiments were performed. HA-PTP-PEST WT or HA was added to 293T cells using the calcium phosphate method.
-PTP-PEST C231S (10 μg) and either Myc-p130cas or Myc-p130cas SH3 domain (10 μg) were co-transfected. A schematic diagram of the p130cas protein is shown in FIG. 14A. Cell lysates were immunoprecipitated with anti-PTP-PEST antibody 1075 to confirm the presence of Myc-p130cas protein (FIG. 14B). Myc-p130cas and Myc-p130cas SH3 domains were found in the HA-PTP-PEST immunoprecipitate (FIG. 14B). As a negative control, Myc-p130cas or Myc
-P130cas SH3 domain was co-transfected with HA-T cell PTP (TC-PTP). FIG. 14B shows that Myc in TC-PTP immunoprecipitates.
This shows that the presence of neither -p130cas nor Myc-p130cas SH3 domain was observed, clearly indicating that this interaction is specific for PTP-PEST. These results indicate that in vivo PT
P-PEST has been shown to bind to full-length p130cas, or specifically to the SH3 domain of p130cas.

Discussion Identifying the substrate for protein tyrosine phosphatase is one of the key steps in understanding the biological function of this group of enzymes. Recent mutagenesis experiments on invariant amino acids within the conserved catalytic domain of the PTP family in connection with the analysis of the crystal structure of PTP1B have yielded two interesting variants that can be used for substrate trapping. Was. One is a mutant in which cysteine of the VHCSAG (SEQ ID NO: 6) motif is replaced with serine (C215S in PTP1B);
The other is a mutant in which aspartic acid involved in intrinsic catalytic activity is substituted with alanine (D199A in PTP1B) (Reference 2). This data offers a novel approach to identify putative physiological substrates by combining substrate trapping experiments with gene targeting of one of the PTPs, PTP-PEST. . PTP-PEST deficient fibroblast cell line
It was made from a T knockout mouse (Reference 1) and used as a protein extract in substrate trapping experiments. When such a method is used, PTP-PE
The major protein isolated with the ST substrate trapping mutant was p130cas. The advantage of this new approach is that cells have not been subjected to artificial treatment to increase the phosphotyrosine content, so that in gene-targeted tissues or cell lines, Only physiological substrates with high specificity are hyperphosphorylated. As a result, the substrate trapping mutant of the selected PTP will result in the isolation of such substrates primarily in large quantities.

It is logically correct that upon removal of PTPase, the physiological substrate is hyperphosphorylated at basal levels or after certain stimuli. For example, the serine-threonine-tyrosine phosphatase of vaccinia virus (ie, dual-specifi
Deletion of the phosphatase VH1, a ty phosphatase, from virion capsules results in hyperphosphorylation of the two substrates in the capsule, p18 and p11, which impairs the virus life cycle (Reference 70). PTP-PEST
PTP-PEST, provided that the physiological substrate of
A fibroblast cell line, PTP-PEST (-/-
) Was prepared. The anti-phosphotyrosine profile of the total cell lysate of the PTP-PEST (-/-) type cell line was compared with that of the total cell lysate of the PTP-PEST (+/-) type cell line, resulting in 180, 130 and 97 kDa. Of highly phosphorylated proteins were identified in PTP-PEST (-/-) type cells. One of these proteins was identified as p130cas. PTP-PEST (+/-) type and (-/-) type were immunoprecipitated against p130cas and analyzed directly by blotting using an anti-phosphotyrosine antibody. As a result, p130cas was deleted in the absence of PTP-PEST. Was found to be hyperphosphorylated. This means
Although a large number of PTPs are expressed in mouse fibroblasts and at least one of them, PTP1B, interacts with the SH3 domain of p130cas (37), phosphorylation of p130cas at basal levels is mainly PTP-
It is suggested that it is regulated via PEST. Identification of hyperphosphorylated p97 and p180 is currently underway. We have previously described PTP-P
Since it has been published that EST forms a complex with the EGF receptor via Grb2 (Reference 4), it was examined whether the p180 protein is an EGF receptor.
The same amount of EGF receptor was found in immunoprecipitates of PTP-PEST (-/-) type cells and (+/-) type cells with anti-phosphotyrosine antibody. It was considered that Hefl and Sin might be substrate candidates for hyperphosphorylated p97. Neither of these proteins was expressed in the above fibroblast cell lines. This finding is not surprising due to the limited expression pattern of Hefl and Sin.

The data presented here is for p130ca belonging to the adapter protein family.
s, Sin and Hefl, via their SH3 domains, PTP-PEST
It also shows that it interacts with. ³³³ PPKPPR ³³⁸ (SEQ ID NO: 7), the first proline-rich region (Pro1) of PTP-PEST, is a PTP-PES
Four other proline-rich motifs (Pro2-4) found to be present on T
Was found to mediate the interaction with the SH3 domain of the adapter protein family without the contribution of These data support the recent findings of Garton et al. (4) and suggest that binding of p130cas to PTP-PEST via the SH3 domain may play an important role in substrate specificity. ing. These results suggest that Hefl and Sin may also be substrates for PTP-PEST, since they also bind to PTP-PEST via their SH3 domains. It would be interesting to analyze the phosphorylation status of Hefl and Sin in PTP-PEST (-/-) type cells, but unfortunately they are not expressed.

A proline-rich region (Pro1) of PTP-PEST³³³PPKPP
R³³⁸Belongs to consensus sequence class 2 (PxxPxR) (Reference 71)
. Other proteins that are known to interact with the SH3 domain of p130cas
Lorin-rich domains also belong to class 2 of the consensus sequence, and PTP 1
B is PRPPKR (Reference 37), and FAK and RAFTK are PPPKR.
PKPSR (References 55 and 72). The other three enzymes, namely PTP-PE
P, PTP-HSCF and PTP-BDP have homology with PTP-PEST.
I do. Presumed to be a proline-rich region present in the PTP family in Table 1.
I put things together. PTP-PEP is a protocol belonging to class 2 of the consensus sequence.
Has two PPLPER-rich domains (two identical sequences) (Reference 7).
3). The two proline-rich sequences of PTP-PEP are p130cas, Sin
Or it is not expected to interact with the SH3 domain of Hefl. The reason is P
The same proline-rich sequence found on TP-PEST (Pro4, ⁶⁷⁵ PPPPER⁶⁸⁰, SEQ ID NO: 8) do not interact with these SH3 domains
That's why. (FIG. 5 of the present application). Recognized on PTPHSCF and PTP-BDP
Analysis of the PxxP motif (Table 1) shows that all consensus sequences were class 2
(PxxPxR), and in this context, these two PTPs are p130
that it will not interact with the SH3 domain of cas, Sin and Hefl.
Show. In conclusion, among those belonging to PEST such as PTP group, PTP-P
EST is the only protein that has an SH3 domain via a proline-rich region
Could interact with the quality p130cas, Sin and Hefl. P
Binding of TP-PEST to the SH130 domain of p130cas is a substrate recognition mechanism.
It has an important role in the culture (Reference 11). PTP-PEST is ubiquitous
In fact, p130cas, Sin and Hefl each show different expression,
, PTP-PEST is an adapter protein family in different cell types
It is reasonable to consider it as a substance that potentially modulates the tyrosine phosphorylation level of a protein.

[0084] Wild-type or C231S PTP-PEST is constitutively bound to p130cas in in vivo co-immunoprecipitation experiments and SH130 of p130cas.
It has been demonstrated that the domain alone mediates this interaction. In this context, the binding of the SH3 domain to the proline-rich domain is a phenomenon commonly seen in this type of interaction, as described above, where p130cas is dephosphorylated by the catalytic function of PTP-PEST. However, the binding does not terminate (Reference 6). PTP-PES
When the Pro1 domain of T is removed, p130cas in vivo (Reference 4)
), The binding of Hefl and Sin to the SH3 domain is completely lost, clearly indicating that this domain is the major binding site between these proteins. Interestingly, in co-immunoprecipitation experiments using transformed 293-T cells, the binding of PTP-PEST to p130cas was due to the p130cas SH
It is not lost by deletion of the Pro1 domain of 3 domains and / or PTP-PEST. These results indicate that PTP-PEST and p130cas also show SH
This suggests that it may be linked by a mechanism independent of 3, which is currently under investigation.

In conclusion, the combination of PTP gene knockout and trapping experiments can be a powerful tool for identifying physiological substrates of PTP. Such an approach confirmed that p130cas was a physiological substrate for mouse PTP-PEST. This technique reduces the risk of incorrect results because it does not require artificial treatment of the cells, and can also be used to identify other PTP substrates. SH3 of Sin and Hefl, a molecule similar to p130cas
The domain was found to interact with high affinity to the proline-rich region of PTP-PEST (Pro1, amino acids 333-338 (SEQ ID NO: 7)).
. These data suggest that Hefl and Sin can be substrates for PTP-PEST. This hypothesis is supported by the requirement for a complete Pro1 region for effective dephosphorylation of p130cas by PTP-PEST. Coimmunoprecipitation experiments revealed that the binding between PTP-PEST and p130cas in vivo was stable and constitutive (Reference 6). This has not been revealed before.

Example 4 PTP-PEST binds to paxillin via the PRO2 region. Since paxillin is highly phosphorylated in PEST-/-cells, it was further investigated which domains of each molecule interact. Paxillin is PES
It was confirmed that it could bind to T but was not a substrate for this phosphatase. Paxillin is a member of the adapter protein family that also includes Hic5 (Reference 17) and Lepaxin (Reference 18). Paxillin present in focal adhesions binds to protein tyrosine kinases present in adhesion plaques such as p125FAK, Pyk2 and c-Src (Reference 19), and important cytoskeletal proteins such as tailin and vinculin. . Especially p1
Binding of 25FAK to paxillin has been shown to be essential for focal adhesion targeting of p125FAK (Reference 20). In addition, paxillin bound to integrins is p125
It has been shown to be phosphorylated by FAK and c-Src (References 21, 2).
2). As a result, a binding site with the SH2 domain of the Crk protein is formed (
Reference 23), Integrin activation is transmitted via C3G or SOS bound to CrK to activate signal transduction pathways. In addition to such positive tyrosine phosphorylation (+ tyrosine phosphorylation), in cells seeded on fibronectin, the serine and threonine residues of paxillin are also significantly phosphorylated (
Reference 24). The phosphorylation of serine and threonine of paxillin is involved in targeting to focal adhesion and adhesion of cells to fibronectin (Reference 25).

Structurally, paxillin and paxillin-like proteins consist of an N-terminal LD motif and four C-terminal LIM domains. It is known that the LD motif of paxillin (brief reference 26) is involved in the binding of paxillin to p125FAK and vinculin (reference 19). The existence of the LD motif is known in various proteins via a protein-protein interaction (Reference 26). L
The IM domain is composed of about 50 amino acids, and is known to mediate protein-protein binding (schematic references 27 and 28). LIM domains have highly conserved consensus sequence _{[CXXC (Xr 16-23) HXXC]} XX [CXXC (X 1 6-21) CXXD / D / H] ( Document 28), a protein with LIM domains often It also has other domains such as homeo domain, kinase domain, SH3 domain and LD domain. The most characteristic action mediated by the LIM domain is the binding of the tyrosine-rich motif (tyrosine-rich turn) of the insulin receptor to the Enigma LIM3 (Reference 29). Enigma L
IM2 also interacts with Ret receptor tyrosine kinase (29). These interactions indicate that each LIM domain may actually recognize and interact with a particular protein domain. Paxillin LIM domain,
In particular, LIM3 is essential for correct focal adhesion targeting. Although the focal adhesion targeting ligand of LIM3 has not been identified (Reference 19), it has recently been shown that LIM2 and LIM3 bind to proteins having serine kinase activity (Reference 25).

In a recent report, paxillin was converted to P by an uncharacterized mechanism.
It was shown to bind directly to the C-terminus of TP-PEST (Reference 30). Current studies have discovered that the non-classical proline-rich motif of PTP-PEST (Pro2) is essential for binding of PTP-PEST to paxillin both in vitro and in vivo . More specifically, a mutation that replaces proline 362 with alanine causes a complete loss of this binding. Complete form of LIM3 and L in paxillin
The presence of the IM4 domain is required for binding to PTP-PEST. Finally, a substrate trapping assay using a mutant of PTP-PEST having the C231S mutation demonstrates that paxillin is not a substrate for PTP-PEST. Together, these results demonstrate a novel binding between the LIM domain and the proline-rich motif. As a physiological function of this binding, we found that once PTP-PEST migrated to the adhesion plaque, paxillin, p13
We propose the dephosphorylation of proteins such as 0Cas, Shc and Grb2 and / or the modulation of the focal adhesion degradation and turnover by binding to them.

Materials and Methods Cell Lines, Transient Transfection and Hypervanadate Treatment NIH 3T3 Cell Line, NIH3T3 Cell Line Overexpressing Constitutively Active Mutant Src Y527F, and HEK 293T Cells Strains were maintained in DMEM supplemented with 10% fetal calf serum and penicillin / streptomycin (Gibco-BRL) in the usual manner. According to the calcium phosphate method described in Reference 10, HEK 293T cells were transformed with 5 μg of the PTP-PEST plasmid. NIH 3T3 cells were treated with pervanadate for 30 minutes as described in Ref.

Antibodies The monoclonal antibody against paxillin (P13520) and the monoclonal antibody against p130Cas (P27820) were from Transduction Laboratories. The polyclonal antibody specific to tripaxillin described in Reference 19 was used. Anti-GST tag antibody and anti-HA tag antibody 12CA5 were obtained from Santa Cruz B
Obtained from iotechnology. Anti-phosphotyrosine antibodies 4G10, PY20 and PY
99 is Upstate Biotechnologies, Transduction Laboratories and Santa Cruz
Biotechnology ones were used respectively. As the polyclonal antibody 1075 against PTP-PEST, the one described in Reference 3 was used.

Construction of Plasmid A plasmid obtained by incorporating PTP-PEST cDNA into the expression vector pACTAG (containing an HA epitope as a tag) described in Reference 3 was used. PTP-PEST Pro1 domain (333-PPKPPR-338; SEQ ID NO: 7) and Pro2 domain (355-PPKPPR-338;
PPEPHPVPPILTPSPPSAFP-374; SEQ ID NO: 9) was subjected to PCR (Polymerase chain reactio).
n). Briefly, based on the sequence upstream of the Pro1 domain, an antisense oligonucleotide (5) was designed to insert an XhoI site.
PTP-PEST cDNA was amplified using '-TCGCTCGAGAGAGTCTTGCTTCTC-3'; SEQ ID NO: 10) together with a sense oligonucleotide T7. PTP-PEST
This PCR product encoding the N-terminus of the DNA was purified using a gel (QIAEXII, Qiagen
), NotI and XhoI. The second PCR product encoding the C-terminus of PTP-PEST is a sense oligonucleotide (5′-CCACTCGAGACTCGAAGTTGCCTT-3 ′; SEQ ID NO: 11) designed to insert an XhoI site based on the sequence downstream of Pro1. And an antisense oligonucleotide having an XbaI site (5′-
It was prepared using CCTCTAGATCATGTCCATTCTGAA-3 ′; SEQ ID NO: 12). PTP-P
The PCR product encoding the EST C-terminus was purified using a gel, and XhoI and Xb
digested with aI. To reconstruct the full-length sequence of PTP-PEST lacking the Pro1 region, digested PCR encoding the N- and C-termini of PTP-PEST
The product was ligated into the NotI and XbaI sites of pACTAG. PTP-PEST
Was deleted in the same manner. Confirmation of the defect
This was performed by confirming the sequence of the mutated region by the dideoxy method using m). GSTP
TP-PEST 344-437 (Pro2) (GST 344-427, 344
-417, 344-407 and 344-397) were performed by PCR using a T7 primer and an oligonucleotide specific to PTP-PEST having an EcoRI site. The PCR product was digested with BamHI and EcoRI, and pGEXR
C BamHI and EcoRI sites. GST PTP-PEST 344
-385 was obtained by digesting the pGEX RC 344-437 vector with KpnI and EcoRI, treating it with Mungbean nuclease, and religating the plasmid.

The N of the tripaxillin fused to GST inserted into the pGEX 2T vector
Constructs encoding the terminus, C-terminus, LIM1, LIM2, LIM3 and LIM4 have been described previously (Reference 19). For the construction of Paxillin GST LIM 1-3,
A BamH I / Sac II fragment was obtained from the GST C-terminal pGEX 2T construct,
Subcloned into pBluescript II (KS). Next, pB
The BamHI / SacI fragment in bluescriptII (KS) was converted to pGEXP
It was subcloned into BamHI / SacI of C. A BamHI / Xmal fragment was excised from the construct consisting of the GST C-terminus and was excised from the Bam of pGEX RC.
Subcloning into the HI / XmaI site gave paxillin GST LIM1-2. In addition, a SmaI / EcoRI fragment was cut out from the GST C-terminal construct and subcloned into pGEX RC at the SmaI / EcoRI site to obtain paxillin GST LIM3-4. Paxillin GST C-terminal @LIM
3 was obtained by digesting the C-terminus of paxillin GST with XmaI and SacII, treating with Mungbean nuclease, and religating the plasmid.

The; (SEQ ID NO: 9 355-PPE P H P V PP ILT P S PP SAF P -374), the [0093] Eight proline residues present Pro2 domain of site-directed mutagenesis PTP-PEST (underlined) Mutated to an alanine residue by PCR. Briefly, the following mutagenic oligonucleotides were made: (5'-CAGCCACCAGAAGCTCACCCGGTGC-3 '), P358A (SEQ ID NO: 13), (5'-CCAGAACCTCACGCGGTGCCACCCATC-3'), P360A (SEQ ID NO: 14). , (5'-CCTCACCCGGTGGCACCCATCCTGAC-3 '), P362A (SEQ ID NO: 15), (5'-CACCCGGTGCCAGCCATCCTGACGC-3'), P363A (SEQ ID NO: 16), (5'-CCCATCCTGACGGCATCACCTCCTTC-3 '), P367A (SEQ ID NO: 17) ), (5'-CTGACGCCATCAGCTCCTTCAGCC-3 '), P369A (SEQ ID NO: 18), (5'-ACGCCATCACCTGCTTCAGCCTTCC-3'), P370A (SEQ ID NO: 19) and (5'-CCTTCAGCCTTCGCAACCGTTACCAC-3 '), P374A (SEQ ID NO: 20).

PGEX RC PTP-PEST Pro2 (base sequence encoding amino acids 344 to 437) was used as a template for PCR. Each oligonucleotide was used together with a pGEX-specific antisense oligonucleotide to generate Pro
Two regions were amplified. The resulting eight different PCR products were purified using a gel.
The second PCR product consisted of a pGEX-specific sense oligonucleotide and PTP-P
It was prepared using an antisense primer specific to EST (5′-CCATGTGCAGCACTGGCTTT-3 ′; SEQ ID NO: 21), and recovered by purification using a gel. To reconstitute the full length of the Pro2 region, each mutant PCR product was incubated with a second PCR product and Vent DNA polymerase (New England Bid).
PCR at 30 ° C. for 30 seconds, 50 ° C. for 30 seconds and 72 ° C. for 30 seconds using Olabs) to perform a strand overlap extension step.
Was performed seven times. 100 pools of pGEX sense and antisense primers were added to each of the obtained products, and PCR was performed for 30 cycles under the same conditions as described above. Each PCR product was purified using a gel, digested with BamHI and EcoRI, and ligated to the BamHI and EcoRI sites of pGEX RC. E. FIG. coli
After transformation into DH5α, mutants were screened by sequencing. The cysteine at position 523 of paxillin was replaced with serine using the same method. Other paxillin mutants have been previously reported (Reference 19).
.

Translation in vitro Wild-type paxillin, ΔLIM3, C467A, C
470A, C467 / 470A, ΔLIM4 and C523S, and pCDNA
3 was digested with XhoI. Using the obtained linear plasmid as a template, in vitro transcription was performed using T7 RNA polymerase (NEB).
Using 2 μl of the transcription reaction solution, ³⁵ S-methionine and rabbit reticulocyte lysate (P
romega) was performed in vitro . In vitro translated products were tested for binding using only 1 μg of GST PTP-PEST 344-397 or GST. In order to perform a binding test in vitro , NIH 3T3 cells expressing the GST fusion protein, NIH 3T3 Src Y527F cells, NIH 3T3 cells treated with pervanadate and HEK 293T cells were ligated with Complete protease inhibitor (Complete protease inhibitor). , Boehrin
Germannheim) and 1 mM Na ₃ VO ₄ in HNMETG buffer (50 mM
Tris-HCl pH 7.5, 150 mM NaCl, 1.5 mM MgCl ₂ ,
1 mM EDTA, 1; Triton X-100 and 10% glycerin). The cell lysate was centrifuged at 16,000 × g at 4 ° C. using a microcentrifuge to remove cell debris. Protein concentrations were determined by the Bradford method (Bio-Rad) using fetal calf serum as a standard. PTP-PE
Each GST fusion protein of ST, p130Cas and paxillin was expressed by inducing bacteria cultured to logarithmic growth phase with 1 mM IPTG for 2 hours, and the fusion protein was immobilized on glutathione sepharose beads according to the attached manual. . An equivalent amount of cell lysate, 200-250 μg, is
The impurities were removed with glutathione sepharose (Pharmacia) in which only ST was fixed, and PTP-PEST GS was placed in 1 ml of HNMETG at 4 ° C for 90 minutes.
Incubated with each of the T fusion proteins. 200 μg of 293T cell extract was incubated with each of p130Cas and paxillin GST fusion protein in 1 ml HNMETG for 90 minutes at 4 ° C. After binding the beads to the protein by incubation, the beads were washed with HNTG buffer (20 mM).
Tris-HCl pH 7.5, 150 mM NaCl, 0.1% Triton
(X-100, 10% glycerin) and boiled in SDS sample buffer to elute the bound proteins. Perform SDS-PAGE, PV
The protein was transferred to a DF membrane (Immobillion-P), and the bound protein was detected by Western blotting using an appropriate antibody.

Immunoprecipitation One layer of cultured cells was added to 1 ml of immunoprecipitation buffer (50 mM Tris-HCl p
H 7.5, 150 mM NaCl, 1% NP-40, 1 mM Na ₃ VO ₄ and
(Complete protease inhibitor) and lysed directly in a tissue culture dish on ice. Lysates were centrifuged at 16,000 xg in a microcentrifuge to remove cell debris. Impurities were removed in advance using 200 μl of the obtained equivalent cell extract (200-500 μg) using 20 μl of protein A agarose (Gibro-BRL). The decontaminated extract was incubated for 90 minutes at 4 ° C. with 2 μl of anti-paxillin monoclonal antibody or 1 μl of polyclonal antibody against tripaxillin. 20 μl of protein A agarose was added to the obtained immune complex, and the mixture was reacted at 4 ° C. for 90 minutes and collected. The immunoprecipitates were washed three times with immunoprecipitate buffer and boiled with SDS-PAGE sample buffer. Perform SDS-PAGE,
PTP transcribed and bound to PVDF membrane (Immobillion-P)
-PEST was detected by Western blotting using anti-HA antibody 12CA5. The probe was removed from the blot and blotted again with an anti-paxillin antibody to confirm that the amount of protein contained in the precipitate was the same.

Western Blotting Non-specific portions of the blotted PVDF membrane were TBS-T (20 mM Tris-H) containing 1% (w / v) fetal calf serum and 1% (v / v) goat serum.
Cl pH 7.6, 137 mM NaCl, 0.1% Tween-20).
Blocked by incubating for 0 minutes at room temperature. The primary antibody was diluted in blocking buffer and incubated with the blot for 60 minutes at room temperature. The anti-paxillin monoclonal antibody was used at a dilution of 1: 10,000. The monoclonal antibodies 12CA5 and 4G10 were used at a dilution of 1: 2,000. Anti-p130
The Cas monoclonal antibody, the PY-20 monoclonal antibody, and the PY-99 monoclonal antibody were used at a dilution of 1: 1,000. The blot was washed extensively with TBS-T before incubation with the secondary antibody. HRP as a secondary antibody
The anti-mouse antibody conjugated to was diluted 1: 10,000 with blocking buffer and incubated with the blot for 60 minutes at room temperature. After repeated washing with TBS-T, the presence of the bound primary antibody was determined by chemiluminescence (NEN Life
science).

Results Paxillin binds to PTP-PEST in various mouse tissues. Recently reported chicken embryo (CE) fibroblasts and Swiss 3T
In 3 cells (Reference 30), PTP-PEST and paxillin bind. PT
To confirm that P-PEST and paxillin were physically bound in vivo, co-immunoprecipitation experiments were performed on lysates of various mouse tissues. As can be seen from FIG. 15, paxillin co-precipitated with PTP-PEST in lysates of liver, brain, heart, lung, spleen, thymus. PTP-PE binding to paxillin
The amount of ST was greatest in lung, spleen, and liver. Since the expression level of PTP-PEST in the kidney is small (Reference 31), PTP-PEST obtained from the lysate of the kidney is used.
Paxillin was not detected in the immunoprecipitates. In addition, paxillin and PT
Paxillin was not detected when pre-immunoprecipitation using an anti-paxillin antibody was performed on a liver extract showing specific binding to P-PEST. When blotting is again performed on the blot using the PTP-PEST antibody as a probe,
Except for the kidney where no signal is detected, a considerable amount of PTP-P
EST found (data not shown). These results indicate that binding of PTP-PEST to paxillin, which has been reported for fibroblasts (Reference 30), occurs in vivo in most mouse tissues.

Paxillin binds in vitro with a fragment containing the proline-rich domain of PTP-PEST. To determine the domains of PTP-PEST involved in paxillin binding,
In addition to several deletion mutants of PTP-PEST expressed as a GST fusion protein, a C-terminal sequence having no catalytic activity of PTP-PEST was used (
(FIG. 16A). GST PTP-PEST combined with glutathione sepharose
Fusion proteins (276-775, 276-613, 276-567
No. 276-453, amino acid sequence consisting of 276-437, or G
ST only) was incubated with 500 μg of protein extracted from NIH3T3 cells. After incubation for binding, the beads were washed several times and proteins bound to the beads were separated by SDS-PAGE. PTP-
Paxillin bound to PEST was detected by Western blotting. When the binding assay was performed, paxillin was detected in all tests except when only GST was used, as is clear from FIG. 16B (upper drawing). The GST fusion protein used in the binding assay was stained with Coomassie blue (FIG. 16).
B, lower drawing), indicating that the fusion protein was correctly expressed. These results indicate that the paxillin binding site of PTP-PEST is among the amino acids 276 to 437. The only characterized protein binding domain among amino acids 276-437 of PTP-PEST is p130cas,
A proline-rich domain (Pro1, 333-PPKPPR-338, SEQ ID NO: 7) that serves as a binding site for the SH3 domains of Hefl, Sin and Grb2
(References 6, 9, and 10). This PTP-PEST fragment also contains Pro2, a segment of 20 amino acids of undefined function, rich in proline residues. Pr of PTP-PEST expressed as a GST fusion protein to define the minimum domain required to bind paxillin
o1, Pro2 344-437, Pro2 344-427, Pro2 344
-417, Pro2 344-407, Pro2 344-397, Pro2 3
Binding assays were performed using 44-385 (FIG. 16A). Paxillin, PT
P-PEST fusion protein Pro2 344-437, 344-427, 3
44-417, 344-407, and 344-397 (FIG. 16C, upper drawing). GST Pro2 344-385, GST Pro 1, GST Pr
Only o5 and GST did not bind to paxillin, as can be seen from FIG. 16C (upper drawing). These results indicate that mouse PTP-PEST 344-39.
It suggests that the presence of up to 7 amino acids is sufficient for binding to paxillin. This peptide has SEQ ID NOs: 22 and 23 (corresponding to mouse and human, respectively). Of course, this sequence, which is sufficient to bind paxillin, is not necessarily the minimum sequence required to bind paxillin.

The Pro2 domain is required for PTP-PEST to bind to paxillin in vitro and in vivo . The Pro2 domain of PTP-PEST exists between amino acids 355-374. Since the smallest GST Pro2 fusion protein used in FIG. 15C encodes amino acids 344-397, the PTP-
A PEST mutant was generated and used to pinpoint its role in paxillin binding. Binding assays were performed using two Paxillin GST fusion proteins, Paxillin N (encoding the LD motif) and Paxillin C (encoding the four LIM domains). In addition to GST p130cas SH3 and GST alone, these two proteins were purified and purified from the wild-type or ΔPro2 H
A Incubated with cell extract of HEK293T transfected with PTP-PEST. The bound PTP-PEST protein was isolated from anti-HA antibody 1
Detection was performed by Western blotting using 2CA5. As shown in FIG. 17A (upper view), GST paxillin C and p130cas SH3 were WT HA PT
Bound to P-PEST. GST Paxillin N and GST only, WT HA
-Did not bind to PTP-PEST. Loading 5 μg of extract of HEK 293T transformed with WT HA PTP-PEST showed that the construct used for transformation was correctly expressed. ΔPro2 HA-PTP-
When a binding assay was similarly performed using a cell extract transformed with PEST, GST paxillin C did not bind to ΔPro2 PTP-PEST (FIG. 1).
7A, upper drawing). Since the SH3 domain of p130cas required only the complete Pro1 domain, it could bind to the ΔPro2 mutant of PTP-PEST. ΔPro2PTP-PEST was expressed to the same extent as WT as seen in the TCL (5 μg protein) lane. To confirm the presence and correct expression of the GST fusion protein, the probe was stripped from the blot and the anti-GS
Blotting was performed using the T polyclonal antibody as a probe (FIG. 17A, lower drawing). From these results, it is clear that the Pro2 motif of PTP-PEST is involved in binding to paxillin. Furthermore, the LIM domain of paxillin is more involved in binding to PTP-PEST than the LD motif.

The importance of the Pro2 domain for PTP-PEST binding to paxillin was also confirmed by in vivo co-immunoprecipitation experiments. HEK293T
Cells were transfected with mock recombinants (empty pACTAG), HA WT PTP-PEST, H
A ΔPro1 PTP-PEST, HA ΔPro2 PTP-PEST. 48 hours after the transformation, the cells were lysed and the expression of the transformed construct was confirmed by Western blotting of TCL (5 μg) using anti-HA antibody 12CA5 (upper drawing of FIG. 17B). Next, endogenous paxillin was immunoprecipitated with the monoclonal antibody. The presence of HA PTP-PEST was determined by blotting against the HA epitope using the 12CA5 antibody. As can be seen from FIG. 17B (middle drawing), WT and ΔPro1 PTP-PEST were found in paxillin immunoprecipitates. Corresponding to the results shown in FIG.
No PTP-PEST lacking was found in paxillin immunoprecipitates (see middle drawing). By stripping the probe from the blot and performing blotting again using an anti-paxillin monoclonal antibody as a probe, it was confirmed that an equal amount of paxillin had precipitated from each sample (FIG. 17B, lower drawing).

Analysis of mutations in the Pro2 domain revealed that proline 362 is important for binding to paxillin. The Pro2 domain of PTP-PEST consists of 20 amino acids, half of which are proline residues. (FIG. 18A). There are three PxxP motifs in Pro2, all of which have a consensus sequence at the SH3 binding site (Class 1: RxxPxxP and Class 2: PxxPxR) and WW domain (PPxxY).
) (Reference 32). To better understand this novel paxillin binding domain, using the GST Pro2 344-437 construct (FIG. 16),
Mutants were created by replacing eight of the proline residues one by one with alanine. Eight prolines replaced with alanine and WT GST Pro2 were purified from cultured bacteria in which expression was induced and incubated with NIH3T3 cell lysate. FIG. 18B
As can be seen, paxillin bound to all of the fusion proteins used, except for the P362A mutant. As can be seen from the lower panel of FIG. 18B, the product was correctly expressed and the amount of GST fusion protein was the same. From these results, it is concluded that the proline residue at position 362 is indispensable for binding to paxillin, and is contained in the minimal sequence that binds to paxillin.

Binding to PTP-PEST requires paxillin LIM domains 3 and 4. Having determined the domains of PTP-PEST important for binding to paxillin, the next step was to map the paxillin segments required for binding to PTP-PEST. The GST fusion protein encoding the four LIM domains of paxillin showed binding activity to PTP-PEST (FIG. 16A). To determine which LIM domains are important for binding to PTP-PEST, a series of GST fusion proteins with the LIM domain of paxillin were generated. A schematic diagram of these constructs is shown in FIG. 19A. These fusion proteins were purified from cultures in which expression was induced and incubated with lysates of 293T cells expressing HA PTP-PEST. Bound PTP-PEST was detected by Western blotting using anti-HA antibody 12CA5. As can be seen from FIG. 19B, the carboxyl terminus of paxillin (paxillin C, LIM1-4) was bound to PTP-PEST. As with the individually expressed LIM domains, the mutants lacking the C-terminus (LIM1-3, LIM1-4ΔLIM3, LIM1-2) were PTP-PEST
And did not bind. Interestingly, only the LIM3-4 constructs had PTP-PE
It was a defective construct having binding activity to ST (FIG. 19B). The GST-only protein used as a negative control failed to precipitate PTP-PEST. Coomassie blue stained gels of the GST fusion proteins used in the binding assay were used to show that the products were correct as well as that each construct was expressed, respectively, at the bottom of FIG. 19B. Shown in the side drawings.

Only one of the LIM domains (LIM3 or LIM4) is PTP-PE
Although they may interact with ST, both domains may be required for conformation to exert biological activity. This is considered an important issue to consider, especially since the fusion protein was expressed in bacteria. To test this hypothesis, co-immunoprecipitation experiments were performed between PTP-PEST and full-length wild-type paxillin, ΔLIM3 mutant, or ΔLIM4 mutant. HEK 293T cells were co-transfected with HA PTP-PEST and either pcDNA3 alone (empty), pcDNA3 into which wild type, tripaxillin ΔLIM3, or tripaxillin ΔLIM4 was introduced. 5 μg equivalent of each sample was analyzed by Western blotting to verify PTP-PEST expression (FIG. 20B). Equivalent cell lysates were immunoprecipitated using an antibody specific for tripaxillin. The presence of PTP-PEST in the paxillin sediment was determined by Western blotting using the anti-HA antibody 12CA5. FIG. 20A
As can be seen from (a), PTP-PEST was detected in the sediment of wild-type paxillin, but was not detected in the sediment of the LIM3 deficient mutant or the LIM4 deficient mutant. In order to confirm that the same amount of paxillin had precipitated in each sample, the probe was stripped from the blot and blotted again using an anti-paxillin monoclonal antibody as a probe. As can be seen from FIG. 20A (b), the same amount of paxillin was detected from each sample other than the negative control sample (pcDNA3, FIG. 20A (b)).

For paxillin to bind to PTP-PEST, complete LIM3 and LIM4
Is necessary. To confirm that the LIM3 and LIM4 domains of paxillin are involved in binding to PTP-PEST, point mutations were introduced in these domains.
Each LIM domain consists of two zinc fingers stabilized by critical cysteine, histidine or aspartic acid residues (19). C467A
The mutant, the C470A mutant, disrupts the first and second zinc fingers of the LIM3 domain, respectively. C467 / 470A is a double mutant. C523S
Destroys the first zinc finger of LIM4. Wild-type paxillin and all mutants were translated in vitro in the presence of ³⁵ S-methionine. Thereafter, the obtained product was incubated with GST PTP-PEST Pro2 (amino acids 344 to 397, SEQ ID NO: 22). As can be seen from FIG. 21 (upper panel), only wild-type paxillin showed interaction with the PTP-PEST GST fusion protein. The GST-only protein used as a control showed no interaction with wild-type paxillin (upper panel, last lane). In the lower panel, 15% of each product used in the binding assay showed that the protein was correctly expressed. This figure is also a control for electrophoresis. These data clearly indicate that complete LIM3 and LIM4 are indispensable for paxillin to bind to PTP-PEST.

Paxillin is not a substrate for PTP-PEST. Since paxillin is a protein in which tyrosine is phosphorylated, PTP-PEST
Whether or not paxillin is a physiological substrate of PTP-PEST was examined by a substrate trapping experiment using a C231S mutant having a mutation of the catalytic domain (References 7, 8, and 10). The substrate trapping mutant used was GST 1-45
3 C231S (including Pro2) and GST 1-354 C231S (Pro
PTP-PEST fusion protein (lacking 2). Protein extracts were obtained from NIH3T3 cells as controls, NIH3T3 cells stably expressing Src Y527F, and NIH3T3 cells treated with vanadium peroxide. These samples were prepared using PTP-PE expressed as a GST fusion protein.
ST C-terminal (amino acids 276-775), PTP-PEST 1-453
WT, PTP-PEST 1-453 C231 S, or PTP-PEST
Incubated with 1-354 C231S. Tyrosine phosphorylated proteins, which precipitate with the PTP-PEST GST fusion protein, were visualized by immunoblotting using an anti-phosphotyrosine antibody. As can be seen from FIG. 22 (a), in the Src Y527F-expressing cells and the NIH3T3 cells treated with vanadium peroxide, the protein of 130 kDa shows no PTP-PE
It is the major tyrosine phosphorylated protein that binds to the C231S mutant of ST. This protein is known to be p130cas (References 8, 10) and serves as a control to confirm that the trapping experiment worked. This means
This is clear from the blotting using the anti-p130cas antibody shown in FIG. PTP-PEST 1-354 60 kDa detected in lane C231S
The band shows GST that showed non-specific cross-reactivity with anti-phosphotyrosine antibody PY99.
It is a fusion protein. In NIH3T3 cells and NIH3T3 Src 527F cells, the 70 kDa tyrosine phosphorylated protein was converted to PTP-PEST
C-terminal, 1-453WT, detected in C231S, but 1-354 C231S
Was not detected. This p70 protein was identified as paxillin when the blot was blotted again using an anti-paxillin monoclonal antibody as a probe (FIG. 22 (b)). Paxillin was not detected in the lane of PTP-PEST 1-354 C231S, the only construct lacking Pro2. This clearly indicates that the PTP-PEST catalytic domain does not directly interact with tyrosine phosphorylated paxillin. This result clearly indicates that paxillin is not a substrate for the PTP-PEST catalytic domain in the substrate trapping assay. Interestingly, PTP-PEST with Pro2 when 3T3 cells were treated with vanadium peroxide.
Very little paxillin binding to the fusion protein is detected (overirradiation of the blot; data not shown). This suggests that binding to PTP-PEST can be prevented by hyperphosphorylating the tyrosine residue of paxillin. Proper expression of each fusion protein used in the trapping assay was confirmed by Coomassie blue staining of PVDF shown in FIG. 22 (d).

Discussion In a recent report, Shen et al. (Reference 30) showed that a protein with protein tyrosine phosphatase activity (PTP) and FAK form a complex. One of these PTPs, PTP-PEST, was found in the FAK complex formed via direct binding to paxillin (30). These findings are significant and suggest a dynamic regulation of protein tyrosine phosphorylation in focal adhesions. In support of this hypothesis, fibroblasts deficient in FAK (Reference 33) or PTP-PEST (Reference 14) show a lack of migration ability and a very large number of focal adhesions. Can be This indicates that tyrosine kinases and phosphatases have an active function in associating and disassociating the adhesive structure of the cells. To gain a better understanding of the function of PTP-PEST in regulating focal adhesion turnover, we analyzed the binding of PTP-PEST to paxillin.

Co-precipitation of paxillin and PTP-PEST was performed on CE (bird embryo) and Swiss 3T3
It was reported in cultured fibroblasts (Reference 30). Both proteins co-precipitate from various tissues of normal mice, ie, liver, brain, heart, lung, spleen and thymus (FIG. 15).
Show that the binding of paxillin to PTP-PEST is physiologically relevant. As previously indicated, paxillin was not detected in PTP-PEST immunoprecipitates obtained from kidneys, since PTP-PEST is only slightly expressed in the kidneys (Reference 31).

In the present study, the inventors have identified domains involved in the binding of PTP-PEST and paxillin. According to recent data by the inventors, in vitro ,
This shows that the PTP-PEST segment consisting of amino acid residues 44 to 397 was sufficient to bind paxillin. In vivo , P
Pro2 (355-PPEPHPVPPILTPS, a proline-rich motif of TP-PEST
PPSAFP-374; SEQ ID NO: 9) is essential for binding to paxillin. In addition, the complete domains of paxillin LIM3 and LIM4 are required for binding to PTP-PEST. Furthermore, tyrosine phosphorylated paxillin was unable to form a complex with the substrate trapping mutant Pro2-deficient PTP-PEST (C231S). This indicates that paxillin is not a physiological substrate for PTP-PEST in a "substrate trapping" type assay. The catalytic domain of PTP-PEST is adjacent to the carboxyl terminus that is rich in protein binding motifs. In this regard, p130cas, Hef1, Sin / Efs (Reference 10,
34), the SH3 domain of Grb2, v-Src (Reference 6) and Csk (Reference 31)
It was shown to bind directly to the proline-rich sequence in TP-PEST. Furthermore, the coiled-coil domain of PSTPIP is also a PTP-P
Sh binds to the non-classical polyproline-rich domain of EST (35)
The PTB domain of c was shown to bind to the NPLH motif of PTP-PEST (Reference 5). Pro2 of PTP-PEST does not contain any consensus sequences that can act as ligands for either the SH3 domain or the WW domain. This suggests that Pro2 of PTP-PEST is a novel protein binding motif that can bind at least to the LIM domain of paxillin. Interestingly, Pro2 is conserved between human and mouse PTP-PEST, but is absent from other members of the PTP-PEST family of enzymes (PEP, PTP-HSCF and PTP20). Deficiency of Pro2 that prevents paxillin binding results in misfolded PTP-P
There is a concern of producing ESTs. However, p130cas is still SH3-P
This hypothesis is unlikely to hold because it can interact with PTP-PEST ΔPro2 via ro1 binding (FIG. 17A). Furthermore, site-directed mutagenesis analysis indicates that proline 362 is essential for binding to paxillin, while the other seven proline variants have little or no effect on paxillin binding. Was done. To gain a better understanding of Pro2, see 362
Experiments involving more point mutations of the amino acid residues around the second proline and NMR determination of the peptide binding to the partner paxillin are currently underway.

Paxillin is an adapter protein composed entirely of protein binding modules, including LD, LIM, SH2 binding sites and proline-rich domains. Our results indicate that only LIM3 and LIM4 are essential for PTP-PEST binding activity. When expressed as a GST fusion protein, both the LIM3 and LIM4 alone, by all means, suffer from PT, despite the fact that the construct containing the LIM3 and LIM4 domains supports binding.
It could not bind to P-PEST. LIM3 and LIM4 may not fold properly when expressed in bacteria, however, Brown et al.
n et al.) (ref. 25) shows that GST-LIM2 and GST-
This shows that LIM3 alone binds to serine / threonine kinase, suggesting that the GST fusion protein is a properly folded domain (Reference 2).
Five). The present inventors have also observed that paxillin lacking either LIM3 or LIM4 from its full length cannot bind to PTP-PEST in vivo . In addition, we have clearly shown that point mutations that disrupt either the LIM3 or LIM4 zinc fingers of paxillin prevent binding to PTP-PEST. At the same time, the above data shows that LIM3 and LIM4
This indicates that the domain is essential for binding to TP-PEST.

The ligands for the LIM domain are very diverse (27, 28). For example, some L
IM domains heterodimerize and others bind structurally distinct protein motifs (Reference 27). Among others, Enigma LIM3 has previously been shown to bind to a tyrosine-based motif (tyrosine tight turn) in the β-chain of the insulin receptor (29). Interestingly, the tyrosine tight turn (GPLGPLYA) of the insulin receptor contains a PxxP motif, and a mutation that replaces two prolines with alanine abolishes Enigma binding to LIM3.
(Reference 29). Furthermore, when Enigma LIM3 was used to screen a random peptide library with immobilized tyrosine (position 0), prolines at positions -1 and +2 were preferred. The consensus sequence for a preferred ligand of Enigma LIM3 was determined to be GPHydGPLHyd (Y / F) A (Hyd = hydrophobic residue). Our data show that Pro2 of PTP-PEST, ie, 355-PPEPHPV P PILTPSPPSAFP-374 (SEQ ID NO: 9), is the binding site for paxillin, and that proline 362 (underlined) is essential for binding. Is shown. The consensus sequence of the ligand for Enigma LIM3 is PTP-PE
It is not found in Pro2 of ST. Thus, paxillin LIM3 and LIM4
Binds to a novel polyproline motif, which is added to various types of LIM domain ligands. Paxillin LIM3 and LIM
Discovery of other ligands for P4 and comparison of PTP-PEST with Pro2
The sequence of the preferred ligand for the LIM domain will be detailed.

In parallel with this study, PTP-PEST has also been reported to bind to Hic-5, similar to paxillin (36). The C-terminal LIM domain of Hic-5 has 68% homology with the paxillin LIM domain. Lic of Hic-5
3 was shown to be most important for binding to PTP-PEST, but not yet sufficient. Surprisingly, LIM4 was not essential for Hic-5 binding to PTP-PEST. In addition, point mutations in the zinc finger of either LIM3 or LIM4 of Hic-5 indicate that PTP-PEST
Did not prevent binding. These results differ from those observed for the binding between PTP-PEST and paxillin, and the mechanism of binding to PTP-PEST even though the LIM domains of Hic-5 and paxillin have 68% homology. Implies that they are not identical.

PTP-PEST has been shown to be very selective for physiological substrates (8,10). Selectivity for p130cas can be explained by the fact that both recognition of the substrate by the domain of PTP and SH3-mediated binding occur before dephosphorylation. An important issue that needs to be resolved to understand the significance of the binding of paxillin to PTP-PEST was to clarify whether paxillin is a substrate for PTP-PEST. According to the data of the inventors, Pro
2 clearly shows that tyrosine-phosphorylated paxillin did not bind to the PTP-PEST C231S mutant lacking 2. This means that PTP
Indicate that paxillin is not directly recognized by the catalytic domain of Further, GST-PTP-PEST WT or GST-PTP-PEST C231S
Equal amounts of paxillin were found in the immunoprecipitates of E. coli and the fact that paxillin was not further tyrosine phosphorylated in the C231S sample, suggesting that there was no co-operation between the catalytic domain and Pro2. (Fig. 22 (c
)). In support of these findings, Garton and Tonks (Garton & T
onks) (16) showed that overexpression of PTP-PEST in Rat-1 fibroblasts prevented cell migration in a wound healing migration assay. In these cells, the phosphotyrosine levels of p130cas were greatly reduced, despite the unchanged tyrosine phosphorylation levels of paxillin and FAK (Reference 16). Thus, the above results also support that paxillin is not a target of PTP-PEST.

PTP1B (References 7, 37), Tcell-PTP (Reference 38) and SHP-1 (Reference 39)
Have been reported to have significant specificity for substrates. In contrast, since the PTP catalytic domain alone does not dephosphorylate PTP-PIP, the presence of the PTPIP-binding motif
It was shown to be essential for specific tyrosine dephosphorylation of TPIP (35). Since our conclusions are based solely on the substrate trapping approach, the possibility remains that paxillin is a weak substrate for PTP-PEST in vivo . The formation of some protein complexes may also favor the dephosphorylation of paxillin by PTP-PEST. In an in vitro dephosphorylation assay, GST-PTP-PEST weakly dephosphorylated paxillin peptide compared to p130cas peptide. The known promiscuous activity of the PTPases in vitro prevents the inventors from being based on substrate identification performed using the above assays.

If paxillin is not a substrate for PTP-PEST, paxillin and PTP-PE
What is the physiological significance of binding to ST? The first clue to answer this question came from a finding by Brown et al. (19) that showed that the localization of paxillin in cells was dependent on the binding of an unknown protein that binds to paxillin LIM3. A reasonable assumption is that the LIM3 ligand needs to co-localize with paxillin at the focal contact site. Since PTP-PEST is found primarily in the cytoplasm, it is unlikely to be a protein involved in the localization of paxillin in focal adhesions. The inventors have shown in a previous study that PTP-PEST is able to translocate to the periphery of the cell membrane after signaling by integrins (Reference 14). Then, the inventors translocated PTP-PEST to focal adhesion,
Reference 14) We propose a model that binds to paxillin (Fig. 9A). This couples p130cas localized in focal adhesion via SH3,
It inhibits downstream signaling through dephosphorylation of essential tyrosine residues of p130cas. Importantly, instead of binding to a ligand that is essential for targeting focal adhesion, paxillin's LIM3
That is, the binding of P-PEST to Pro2 is achieved (Reference 19). Also,
The SH3 domain of p130cas, possibly through binding to p125FAK,
It has been shown that p130cas is essential for properly targeting focal adhesion (40). In the same manner as paxillin LIM3, p130c
as SH3 domain replaces P125FAK with PTP-PEST Pro
It seems to bind to 1. At the same time, this cascade will result in the release of paxillin and p130cas from the focal adhesion complex, with dephosphorylation of tyrosine of p130cas as seen in FIG. 9B. PTP-P
Further migration of Csk via direct binding to ESTs (ref. 31) will result in phosphorylation of Src inhibition sites, thus inhibiting the formation of new focal adhesions. . The inventors' finding that PTP-PEST knockout cells accumulate a large number of focal adhesions indicates that the most important function of PTP-PEST is the disassociation of focal adhesions, that is, the turnover of focal adhesion complexes. It suggests an event that is to promote (Reference 14). Thus, we believe that tyrosine phosphatase activity on p130cas, as well as p130 required for focal adhesion targets.
The direct binding of cas and paxillin to the essential domains allows PTP-PE
ST proposes to promote focal adhesion turnover.

Example 5 Therapeutic Application of PTP-PEST Inhibitors Based on the above results, PTP-PE was applied to various domains of signaling proteins.
Any substance that inhibits ST binding can be expected to have the ability to interfere with cell migration and / or cell division. Such effects can be applied clinically in treating tumors and preventing pro-inflammatory cell migration, as well as preventing unwanted angiogenesis. These inhibitors are thought to be specific for PEST, for example, specific inhibitors, competitive binding peptides, monoclonal antibodies against substrate or enzyme binding sites, monoclonal antibodies against enzyme catalytically active sites, etc. included.

Such preferred agents have been designed for the following two peptides that compete with the endogenous enzyme for the natural substrate (ie p130cas) and bind to the substrate via its SH3 domain: 1: Thr Thr Gly Thr Met Val Ser Ser IIe Asp Ser Glu Lys Gln Asp
Ser Pro Pro Lys Pro Pro Arg Thr Arg Ser Cys Leu Val Glu Gly (SEQ ID NO: 4) or peptide 2: Gln Asp Ser Pro Pro Pr as an example of a shorter sequence
o Lys Pro Pro Arg Thr Arg (SEQ ID NO: 24).

In addition, peptides can be designed that interfere with the binding of PEST to a signaling protein involved in cell migration and / or cell proliferation, and the signaling protein need not be a substrate for PEST. Examples of such peptides include P
Peptides consisting of the ESTs at positions 344 to 397 (see SEQ ID NO: 22) and peptides shorter than that and binding to paxillin at their Pro 2 domain are mentioned. This binding peptide is expected to competitively inhibit endogenous enzymes from binding to their corresponding paxillin binding sites.

Anti-tumor activity can be obtained by interfering with multiple events involving cells, such as cell division, tumor cell migration, endothelial cell migration, and the like. As a result, the metastasis of the tumor to the whole organism becomes impossible, the tumor becomes indivisible, the supply of blood, oxygen and nutrients is cut off, and the metabolic waste cannot be removed.

As demonstrated with the Pro 2 peptide in the examples relating to the binding of PEST to paxillin, the peptides of the invention enter the target cells.

In that case, about 10 μmol of Pro 2-TAT-HA (data not shown)
The fusion peptide actually penetrates the cell and is introduced into the coronal focal adhesion. This was confirmed by immunofluorescence using a mouse anti-HA monoclonal antibody and a goat anti-mouse fluorescein-conjugated secondary antibody.

The binding between PTP-PEST and its binding partner has conventionally been 5 in vitro.
It is destroyed by the addition of 00 nM peptide (Reference 5). Considering that the effect of S. Dowdy's TAT-fusion protein system can be obtained by using only 15 nM (Reference 74), TAT-Pro 1 to Pro 5 show that the activity of PTP-PEST enzyme and its activity are cell migration. The inventors believe that they provide a more desirable means for inhibiting the invasion of cells into other tissues, their effects on angiogenesis and tumorigenesis.
This active peptide-based drug is administered so that its extracellular concentration is 10 nM to 1 mM. Preferably, the pharmaceutical composition is prepared so that the extracellular concentration is 0.1 μM to 100 μM, and the intracellular concentration is on the order of nanomolar.

[0123] The pharmaceutical composition further comprises a pharmaceutically acceptable carrier. Ideally, the inhibitory peptides described above are dosed in a suitable form and by a suitable route to deliver an appropriate amount to the target tissue. Such carriers include liposomes, immunoliposomes, nanoerythrosomes, and derivatives thereof (eg, PEG, immuno-nanoerythrosomes, etc.), nanoparticles (
nanoparticles) and derivatives thereof (eg, PEG particles, immuno-nanoparticles, etc.)
And the like. The above peptides can be modified to form derivatives thereof. This `` derivative '' refers to substituting an amino acid with a natural or unnatural amino acid to improve the pharmacokinetic and pharmacodynamic properties of the peptide, or adding a substituent to an atom of the natural amino acid, And any peptide variants obtained by such methods. Therefore, derivatives are compounds that satisfy pharmaceutical requirements such as metabolic degradation resistance, solubility, and potency.

As an alternative to the above peptides, antisense oligonucleotides to PEST can be made. The antisense molecule can be immobilized on the same carrier as the pharmaceutically acceptable carrier described above in connection with the peptide, and the antisense molecule can confer metabolic degradation resistance, improve stability, and / or It can be modified for purposes such as binding to a molecule. Peptides and antisense DNA molecules can also be administered via plasmid vectors or viral vectors. These vectors may or may not express the peptide. If the peptide is expressed from the vector, the peptide will act as a competitor for the endogenous PTP-PEST enzyme. On the other hand, antisense m
If an RNA molecule is produced, it may bind to an endogenous messenger RNA molecule encoding this endogenous enzyme. In the latter case, the expression of the endogenous enzyme is reduced or stopped. To ensure targeting of the tissue to be treated, the antisense molecule or peptide may be conjugated to a ligand capable of specifically and selectively binding to a receptor on the target cell.

Example 6 Based on binding studies, the binding of phosphatases to signaling proteins
Studies on the binding between p130cas and PEST, and the binding between paxillin and PEST, which design interfering peptides, provide valuable information about the properties of the peptide with potential clinical applications.

While all of the above agents act on PEST, it is also possible to create agents that act on the SH3 domain or other domains important for the activity of signaling proteins. These drugs can also be obtained based on the above-mentioned studies on binding.

As summarized in the table below, similar binding studies can be performed for different proline-rich regions of PEST and different signaling proteins, which may be useful in treating various disease states This can be understood by those skilled in the art.

Some signaling proteins are ubiquitous in any tissue and some are tissue-specific. Since the tissue targeting ability of the pharmaceutical composition can be improved by the choice of the carrier, a treatment specific to the lesion tissue is conceivable.

[0129]

[Table 1]

Although the invention has been described with reference to specific embodiments, it is clear that these embodiments can be modified without departing from the spirit and teaching of the invention. These modifications are within the scope of the invention as defined in the claims.

Reference 1. Neel, BG, & Tonks, NK (1997) Current Opinion in Cell Biology 9 ,
193-204. 2. Flint, AJ, Tiganis, T., Barford, D., & Tonks, NK (1997) Proceed
ings of the National Academy of Sciences of the United States of America
94 , 1680-5. 3.Charest, A., Wagner, J., Shen, SH, & Tremblay, ML (1995) Bioche
mical Journal 308 , 425-32. 4.Charest, A., Wagner, J., Kwan, M., & Tremblay, ML (1997) Oncogene
14 , 1643-51. 5. Charest, A., Wagner, J., Jacob, S., McGlade, CJ, & Tremblay, ML
(1996) Journal of Biological Chemistry 271 , 8424-9. 6. Garton, AJ, Flint, AJ, & Tonks, NK (1996) Molecular & Cellul
ar Biology 16 , 6408-18. 7. Cote, JF, Charest, A., Wagner, J., and Tremblay, ML (1998) Bioc
hemistry 37 , 13128-13137. 8. Black, DS, and JB Bliska. (1997) EMBO Journal 16 , 2730-44. Nakamoto, T., R. Sakai, H. Honda, S. Ogawa, H. Ueno, T . Suzuki, S. Aizaw
a, Y. Yazaki, and H. Hirai. (1997) Molecular & Cellular Biology 17 , 3884
-3897. 9. Hanks, SK, and TR Polte. (1997) Bioessays. 19 : 137-45. Garton, AJ, Burnham, MR, Bouton, AH, & Tonks, NK (1997) Onc.
ogene 15 , 877-885. 10. Burridge, K., and M. Chrzanowska-Wodnicka. (1996) Annual Review of C
ell & Developmental Biology 12 , 463-518. 11.Cary, LA, DC Han, TR Polte, SK Hanks, and JL Guan. (1
998) Journal of Cell Biology 140 , 211-21. 12. Fincham, VJ, and MC Frame. (1998) EMBO Journal 17 , 81-92. 13. Shen, Y., G. Schneider, JF Cloutier, A. Veillette , and MD Scha
ller. (1998) Journal of Biological Chemistry 273 , 6474-6481. 14. Spencer, S., Dowbenko, D., Cheng, J., Li, WL, Brush, J., Utzig, S
., Simanis, V., & Lasky, LA (1997) Journal of Cell Biology 138 , 845-8
60. 15. Shibanuma, M., Mochizuki, E., Maniwa, R., Mashirno, J., Nishiya, N.,
16. Imai, S., Takano, T., Oshimura, M., and Nose, K. (1997) Mol. Cell. B
iol. 17 , 1224-1235. 17. Lipsky, BP, Beals, CR, and Staunton, DE (1998) J. Biol. Che
m. 273 , 11709-11713. 18. Brown, MC, Perrotta, JA and Turner, CE (1996) J. Cell Biol.
135 , 1109-1123. 19.Tachibana, K., Sato, T., N., DA, and Morimoto, C. (1995) J. Ex.M
ed. 182 , 1089-1099. 20. Turner, CE, and Miller, JT (1994) J. Cell Sci. 107 , 1583-1591 .
21. Schaller, MD, and Parsons, JT (1995) Mol. Cell. Biol. 1 , 2635-
2645. 22. Bellis, SL, Miller, JT, and Turner, CE (1995) J. Biol. Chem.
270 , 17437-17441. 23. Bellis, SL, Perrotta. JA, Curtis, MS, and Turner, CE (19
97) Biochem. J. 325 , 375-381. 24. Brown, MC, Perrotta, JA, and Turner, CE (1998) Mol. Cell. B
iol. 9 , 1803-1816. 25. Brown, MC, Curtis, MS, and Turner, CE (1998) Nature Structu
ral Biology 5 , 677-678. 26.Dawid, IB, Breen, JJ, and Toyama, R. (1998) Trends in Genetics.
14 , 156-162.27.Sanchez-Garcia, I., and Rabbitts, TH (1994) Trends in Genetics 10
28. Wu, RY, Durick, K. Zhou, SY, Cantley, LC, Taylor, SS, an
d Gill, GN (1996) J. Biol. Chem. 271 , 15934-15941. 29.Shen, Y., Schneider, G., Cloutier, JF, Veillette, A., and Schaile.
r, MD (1998) J. Biol. Chem. 273 , 6474-6481. 30. Davidson, D., Cloutier, JF, Gregorieff, A., and Veillette, A. (19
97) J. Biol. Chem. 272 , 23455-23462. 31. Chen, HI, Einbond, A., Kwak, SJ, Linn, H., Koepf, E., Peterson
, S., Kelly, JW, and Sudol, M. (1997) J. Biol. Chem. 272 , 17070-17077 .
32. Ilic, D., Furuta, Y., Kanazawa, S., Takeda, N., Sobue, K., Nakatsuji
, N., Nomura, S., Fujimoto, J., Okada, M., Yamamoto, T., and et al. (199
5) Nature 377 , 539-544.33.Garton, AJ, Burnham, MR, Bouton, AH, and Tonks, NK (1997
) Oncogene 15 , 877-885. 34.Spencers-S., Dowbenko, D., Cheng, J., Li, WL, Brush, J., Utzig, S.
., Simanis, V., and Lasky, LA (1997) J. Cell Biol. 138 , 845-860.35.Nishiya, N., Iwabuchi, Y. Shibanuma, M., Cote, J.-F., Tremblay , M.,
and Nose, K. (1999) J. Biol. Chem. (In press) 36. Liu, F., Hill, DE, and Chernoff, J. (1996) J. Biol. Chem. 271 , 31 .
290-31301.37.Tiganis, T., Bennett, AM, Ravichandran, KS, and Tonks, NK (
1998) Mol. Cell. Biol. 18 , 1622-1634.38.Timms, JF, Carlberg, K., Gu, H., Chen, H., Kamatkar, S., Nadler,
MJ, Rohrschneider, LR, and Neel, BG (1998) Mol. Cell. Biol. 18 ,,
3838-3850. 39. Nakamoto, T., Sakai, R., Honda, H., Ogawa, S., Ueno, H., Suzuki, T.,
40. Aizawa, S., Yazaki, Y., and Hirai, H. (1997) Mol. Cell. Biol. 17 , 38
84-3897. 41. Casio and Zaret (1991) 42. Ang et al. (1993) 43. Altabha et al. (1993) 44. Managhan et al. (1993) 45. Sasaki and Hogan (1993) 46. Richardson et al. (1997) 47. Fankhauser et al. (1995) 48. Chrzanowska-Wodnicka et al. (1996) 49. Wilson et al. (1995) 50.Schwarzbauer (1997) 51. Lui et al. (1998) 52. Wu et al. (1998) voir nouveau tableau 53. Akanaka et al. (1995) 54. Sakai, R., Iwamatsu, A., Hirano, N., Ogawa, S., Tanaka, T., Mano, H.
, Yazaki, Y., & Hirai, H. (1994) EMBO Journal 13 , 3748-56. 55.Polte, TR, & Hanks, SK (1995) Proceedings of the National Acad
emy of Sciences of the United States of America 92 , 10678-82.56.Harte, MT, Hildebrand, JD, Burnham, MR, Bouton, AH, & Pa
rsonsn JT (1996) Journal of Biological Chemistry 271 , 13649-55.57.Burnham, MR, Harte, MT, Richardson, A., Parsons, JT, & Bout
on, AH (1996) Oncogene 12 , 2467-72.58.Khwaja, A., Hallberg, B., Warne, PH, & Downward, J. (1996) Oncoge
ne 12 , 2491-8.59.Ribon, V., & Saltiel, AR (1996) Journal of Biological Chemistry 2
71 , 7375-80.60.Salgia, R., Pisick, E., Sattler, M., Li, JL Uemura, N., Wong, W.
K. Burky, SA, Hirai, H., Chen, LB, & Griffin, JD (1996) Journal
of Biological Chemistry 271 , 25198-203. 61. Schlaepfer, DD, Broome, MA, & Hunter, T. (1997) Molecular & Ce
llular Biology 17 , 1702-13.62.Casamassima, A., & Rozengurt, E. (1997) Journal of Biological Chemis
try 272 , 9363-70.63.Vuori, K., Hirai, H., Aizawa, S., & Ruoslahti, E. (1996) Molecular &
Cellular Biology 16 ,, 2606-13.64.Sakai, R., Nakamoto, T., Ozawa, K., Aizawa, S., & Hirai, H. (1997) O
ncogene 14 , 1419-26.65. Law, SF, Estojak, J., Wang, B., Mysliwiec, T., Kruh, G., & Golemi
s, EA (1996) Molecular & Cellular Biology 16 , 3327-37.66.Minegishi, M., Tachibana, K., Sato, T., Iwata, S., Nojima, Y., & Mor
imoto, C. (1996) Journal of Experimental Medicine 184 , 1365-75.67.Alexandropoulos, K., & Baltimore, D. (1996) Genes & Development 10 ,
1341-55. 68.Ishino, M., Ohba, T., Sasaki, H., & Sasaki, T. (1995) Oncogene 11 , 2
331-8. 69. Slingerland, JM, Jenkins, JR, & Benchimol, S. (1993) EMBO Jour
nal 12, 102937. 70. Liu, K. , Lemon, B., & Traktman, P. (1995) Journal of Virology 69, 78
23-34. 71. Mayer, BJ, & Eck, MJ (1995) Current Biology 5 , 364-7. 72. Astier, A., Avraham, H., Manie, SN, Groopman, J., Canty, T. , Avra
ham, S., & Freedman, AS (1997) Journal of Biological Chemistry 272 , 2
28-32. 73. Matthews, RJ, Bowne, DB, Flores, E., & Thomas.ML (1992) Mo
lecular & Cellular Biology 12 , 2396-405.74.Nagahara et al. (1998) Nature Medicine 12 , 1449-1452.75.Angers-Lousteau et al. (1999) .J . Cell. Biol. 144 , 1019-1031 . .

[Sequence list]

 SEQUENCE LISTING <110> McGill University Tremblay, Michel L. Cote, Jean-Francois Angers-Lousteau, Alexandre Charest, Alain <120> AGENTS INTERFERING WITH THE BINDING OF PROTEIN TYROSINE PHOSPHATASE PEST TO DOMAINS OF SIGNALLING PROTEINS AS INHIBITORS OF CELL OR OF FOCAL ADHESION <130> 100-5015 <150> CA 2,238,654 <151> 1998-05-21 <150> US 60 / 111,993 <151> 1998-12-11 <150> PCT / CA99 / 00461 <151> 1999 -05-21 <160> 23 <170> PatentIn Ver. 2.1 <210> 1 <211> 775 <212> PRT <213> Mus sp. <400> 1 Met Glu Gln Val Glu Ile Leu Arg Arg Phe Ile Gln Arg Val Gln Ala 1 5 10 15 Met Lys Ser Pro Asp His Asn Gly Glu Asp Asn Phe Ala Arg Asp Phe 20 25 30 Met Arg Leu Arg Arg Leu Ser Thr Lys Tyr Arg Thr Glu Lys Ile Tyr 35 40 45 Pro Thr Ala Thr Gly Glu Lys Glu Glu Asn Val Lys Lys Asn Arg Tyr 50 55 60 Lys Asp Ile Leu Pro Phe Asp His Ser Arg Val Lys Leu Thr Leu Lys 65 70 75 80 Thr Pro Ser Gln Asp Ser Asp Tyr Ile Asn Ala Asn Phe Ile Lys Gly 85 90 95 Val Tyr Gly Pro Lys Ala Tyr Val Ala Thr Gln Gly Pro Phe Arg Asn 100 105 110 Thr Val Ile Asp Phe Trp Arg Met Ile Trp Glu Tyr Asn Val Val Ile 115 120 125 Ile Val Met Ala Cys Arg Glu Phe Glu Met Gly Arg Lys Lys Cys Glu 130 135 140 Arg Tyr Trp Pro Leu Tyr Gly Glu Asp Pro Ile Thr Phe Ala Pro Phe 145 150 155 160 Lys Ile Ser Cys Glu Asn Glu Gln Ala Arg Thr Asp Tyr Phe Ile Arg 165 170 175 Thr Leu Leu Leu Glu Phe Gln Asn Glu Ser Arg Arg Leu Tyr Gln Phe 180 185 190 His Tyr Val Asn Trp Pro Asp His Asp Val Pro Ser Ser Phe Asp Ser 195 200 205 Ile Leu Asp Met Ile Ser Leu Met Arg Lys Tyr Gln Glu His Glu Asp 210 215 220 Val Pro Ile Cys Ile His Cys Ser Ala Gly Cys Gly Arg Thr Gly Ala 225 230 235 240 Ile Cys Ala Ile Asp Tyr Thr Trp Asn Leu Leu Lys Ala Gly Lys Ile 245 250 255 Pro Glu Glu Phe Asn Val Phe Asn Leu Ile Gln Glu Met Arg Thr Gln 260 265 270 270 Arg His Ser Ala Val Gln Thr Lys Glu Gln Tyr Glu Leu Val His Arg 275 280 285 Ala Ile Ala Gln Leu Phe Glu Lys Gln Leu Gln Leu Tyr Glu Ile His 290 295 300 Gly Ala Gln Lys Ile Ala Asp Gly Asn Glu Ile Thr Thr Gly Thr Met 305 310 315 320 Val Ser Ser Ile Asp Ser Glu Lys Gln Asp Ser Pro Pro Pro Lys Pro 325 330 335 Pro Arg Thr Arg Ser Cys Leu Val Glu Gly Asp Ala Lys Glu Glu Glu Ile 340 345 350 Leu Gln Pro Pro Glu Pro His Pro Val Pro Pro Ile Leu Thr Pro Ser 355 360 365 Pro Pro Ser Ala Phe Pro Thr Val Thr Thr Val Trp Gln Asp Ser Asp 370 375 380 Arg Tyr His Pro Lys Pro Val Leu His Met Ala Ser Pro Glu Gln His 385 390 395 400 Pro Ala Asp Leu Asn Arg Ser Tyr Asp Lys Ser Ala Asp Gln Trp Gly 405 410 415 Lys Ser Glu Ser Ala Ile Glu His Ile Asp Lys Lys Leu Glu Arg Asn 420 425 430 Leu Ser Phe Glu Ile Lys Lys Val Pro Leu Gln Glu Gly Pro Lys Ser 435 440 445 Phe Asp Gly Asn Thr Leu Leu Asn Arg Gly His Ala Ile Lys Ile Lys 450 455 460 Ser Ala Ser Ser Ser Val Val Asp Arg Thr Ser Lys Pro Gln Glu Leu 465 470 475 480 480 Ser Ala Gly Ala Leu Lys Val Asp Asp Val Ser Gln Asn Ser Cys Ala 485 490 495 Asp Cys Ser Ala Ala His Ser His Arg Ala Ala Glu Ser Ser Glu Glu 500 505 510 Ser Gln Ser Asn Ser His Thr Pro Pro Arg Pro Asp Cys Leu Pr o Leu 515 520 525 Asp Lys Lys Gly His Val Thr Trp Ser Leu His Gly Pro Glu Asn Ala 530 535 540 Thr Pro Val Pro Asp Ser Pro Asp Gly Lys Ser Pro Asp Asn His Ser 545 550 555 560 Gln Thr Leu Lys Thr Val Ser Ser Thr Pro Asn Ser Thr Ala Glu Glu 565 570 575 Glu Ala His Asp Leu Thr Glu His His Asn Ser Ser Pro Leu Leu Lys 580 585 590 Ala Pro Leu Ser Phe Thr Asn Pro Leu His Ser Asp Asp Trp His Ser 595 600 605 Asp Gly Gly Ser Ser Asp Gly Ala Val Thr Arg Asn Lys Thr Ser Ile 610 615 620 Ser Thr Ala Ser Ala Thr Val Ser Pro Ala Ser Ser Ala Glu Ser Ala 625 630 635 640 Cys His Arg Arg Val Leu Pro Met Ser Ile Ala Arg Gln Glu Val Ala 645 650 655 Gly Thr Pro His Ser Gly Ala Glu Lys Asp Ala Asp Val Ser Glu Glu 660 665 670 Ser Pro Pro Pro Leu Pro Glu Arg Thr Pro Glu Ser Phe Val Leu Ala 675 680 685 Asp Met Pro Val Arg Pro Glu Trp His Glu Leu Pro Asn Gln Glu Trp 690 695 700 Ser Glu Gln Arg Glu Ser Glu Gly Leu Thr Thr Ser Gly Asn Glu Lys 705 710 715 720 His Asp Ala Gly Gly Ile His Thr Glu Ala Ser Ala Asp Ser P ro Pro 725 730 735 Ala Phe Ser Asp Lys Lys Asp Gln Ile Thr Lys Ser Pro Ala Glu Val 740 745 750 Thr Asp Ile Gly Phe Gly Asn Arg Cys Gly Lys Pro Lys Gly Pro Arg 755 760 765 Glu Pro Pro Ser Glu Trp Thr 770 775 <210> 2 <211> 780 <212> PRT <213> Homo sapiens <400> 2 Met Glu Gln Val Glu Ile Leu Arg Lys Phe Ile Gln Arg Val Gln Ala 1 5 10 15 Met Lys Ser Pro Asp His Asn Gly Glu Asp Asn Phe Ala Arg Asp Phe 20 25 30 Met Arg Leu Arg Arg Leu Ser Thr Lys Tyr Arg Thr Glu Lys Ile Tyr 35 40 45 Pro Thr Ala Thr Gly Glu Glu Lys Glu Glu Asn Val Lys Lys Asn Arg Tyr 50 55 60 Lys Asp Ile Leu Pro Phe Asp His Ser Arg Val Lys Leu Thr Leu Lys 65 70 75 80 Thr Pro Ser Gln Asp Ser Asp Tyr Ile Asn Ala Asn Phe Ile Lys Gly 85 90 95 Val Tyr Gly Pro Lys Ala Tyr Val Ala Thr Gln Gly Pro Leu Ala Asn 100 105 110 Thr Val Ile Asp Phe Trp Arg Met Ile Trp Glu Tyr Asn Val Val Ile 115 120 125 Ile Val Met Ala Cys Arg Glu Phe Glu Met Gly Arg Lys Lys Cys Glu 130 135 140 Arg Tyr Trp Pro Leu Tyr Gly Glu Asp Pro Ile Thr Phe Ala Pro Phe 145 150 155 160 Lys Ile Ser Cys Glu Asp Glu Gln Ala Arg Thr Asp Tyr Phe Ile Arg 165 170 175 Thr Leu Leu Leu Glu Phe Gln Asn Glu Ser Arg Arg Leu Tyr Gln Phe 180 185 190 His Tyr Val Asn Trp Pro Asp His Asp Val Pro Ser Ser Phe Asp Ser 195 200 205 Ile Leu Asp Met Ile Ser Leu Met Arg Lys Tyr Gln Glu His Glu Asp 210 215 220 Val Pro Ile Cys Ile His Cys Ser Ala Gly Cys Gly Arg Thr Gly Ala 225 230 235 240 Ile Cys Ala Ile Asp Tyr Thr Trp Asn Leu Leu Lys Ala Gly Lys Ile 245 250 255 Pro Glu Glu Phe Asn Val Phe Asn Leu Ile Gln Glu Met Arg Thr Gln 260 265 270 Arg His Ser Ala Val Gln Thr Lys Glu Gln Tyr Glu Leu Val His Arg 275 280 285 Ala Ile Ala Gln Leu Phe Glu Lys Gln Leu Gln Leu Tyr Glu Ile His 290 295 300 Gly Ala Gln Lys Ile Ala Asp Gly Val Asn Glu Ile Asn Thr Glu Asn 305 310 315 320 Met Ile Ser Ser Ile Glu Pro Glu Lys Gln Asp Ser Pro Pro Lys 325 330 335 Pro Pro Arg Thr Arg Ser Cys Leu Val Glu Gly Asp Ala Lys Glu Glu 340 345 350 Ile Leu Gln Pro Pro Glu Pro His Pro Val Pro Pro Ile Leu Thr Pro 355 360 365 Ser Pro Pro Ser Ala Phe Pro Thr Val Thr Thr Val Trp Gln Asp Asn 370 375 380 Asp Arg Tyr His Pro Lys Pro Val Leu His Met Val Ser Ser Glu Gln 385 390 395 400 His Ser Ala Asp Leu Asn Arg Asn Tyr Ser Lys Ser Thr Glu Leu Pro 405 410 415 Gly Lys Asn Glu Ser Thr Ile Glu Gln Ile Asp Lys Lys Leu Glu Arg 420 425 430 Asn Leu Ser Phe Glu Ile Lys Lys Val Pro Leu Gln Glu Gly Pro Lys 435 440 445 445 Ser Phe Asp Gly Asn Thr Leu Leu Asn Arg Gly His Ala Ile Lys Ile 450 455 460 Lys Ser Ala Ser Pro Cys Ile Ala Asp Lys Ile Ser Lys Pro Gln Glu 465 470 475 475 480 Leu Ser Ser Asp Leu Asn Val Gly Asp Thr Ser Gln Asn Ser Cys Val 485 490 495 Asp Cys Ser Val Thr Gln Ser Asn Lys Val Ser Val Thr Pro Pro Glu 500 505 510 Glu Ser Gln Asn Ser Asp Thr Pro Pro Arg Pro Asp Arg Leu Pro Leu 515 520 525 Asp Glu Lys Gly His Val Thr Trp Ser Phe His Gly Pro Glu Asn Ala 530 535 540 Ile Pro Ile Pro Asp Leu Ser Glu Gly Asn Ser Ser Asp Ile Asn Tyr 545 550 555 560 Gln Thr Arg Lys Thr Val Ser Leu Thr Pro Ser Pro Thr Thr Gln Val 56 5 570 575 Glu Thr Pro Asp Leu Val Asp His Asp Asn Thr Ser Pro Leu Phe Arg 580 585 590 Thr Pro Leu Ser Phe Thr Asn Pro Leu His Ser Asp Asp Ser Asp Ser 595 600 605 Asp Glu Arg Asn Ser Asp Gly Ala Val Thr Gln Asn Lys Thr Asn Ile 610 615 620 Ser Thr Ala Ser Ala Thr Val Ser Ala Ala Thr Ser Thr Glu Ser Ile 625 630 635 640 Ser Thr Arg Lys Val Leu Pro Met Ser Ile Ala Arg His Asn Ile Ala 645 650 655 Gly Thr Thr His Ser Gly Ala Glu Lys Asp Val Asp Val Ser Glu Asp 660 665 670 Ser Pro Pro Pro Leu Pro Glu Arg Thr Pro Glu Ser Phe Val Leu Ala 675 680 685 Ser Glu His Asn Thr Pro Val Arg Ser Glu Trp Ser Glu Leu Gln Ser 690 695 700 Gln Glu Arg Ser Glu Gln Lys Lys Ser Glu Gly Leu Ile Thr Ser Glu 705 710 715 720 Asn Glu Lys Cys Asp His Pro Ala Gly Gly Ily His Tyr Glu Met Cys 725 730 735 Ile Glu Cys Pro Pro Thr Phe Ser Asp Lys Arg Glu Gln Ile Ser Glu 740 745 750 Asn Pro Thr Glu Ala Thr Asp Ile Gly Phe Gly Asn Arg Cys Gly Lys 755 760 765 Pro Lys Gly Pro Arg Asp Pro Pro Ser Glu Trp Thr 770 775 780 <210 > 3 < 211> 5 <212> PRT <213> Homo sapiens <400> 3 Cys Ser Ala Gly Cys 1 5 <210> 4 <211> 31 <212> PRT <213> Mus sp. <400> 4 Thr Thr Gly Thr Met Val Ser Ser Ile Asp Ser Glu Lys Gln Asp Ser 1 5 10 15 Pro Pro Pro Lys Pro Pro Arg Thr Arg Ser Cys Leu Val Glu Gly 20 25 30 <210> 5 <211> 31 <212> PRT <213> Homo sapiens <400> 5 Asn Thr Glu Asn Met Ile Ser Ser Ile Glu Pro Glu Lys Gln Asp Ser 1 5 10 15 Pro Pro Pro Lys Pro Pro Arg Thr Arg Ser Cys Leu Val Glu Gly 20 25 30 <210> 6 <211> 6 <212> PRT <213> Homo sapiens <400> 6 Val His Cys Ser Ala Gly 1 5 <210> 7 <211> 6 <212> PRT <213> Homo sapiens <400> 7 Pro Pro Lys Pro Pro Arg 1 5 <210> 8 <211> 6 <212> PRT <213> Homo sapiens <400> 8 Pro Pro Leu Pro Glu Arg 1 5 <210> 9 <211> 20 <212> PRT <213> Homo sapiens <400> 9 Pro Pro Glu Pro His Pro Val Pro Pro Ile Leu Thr Pro Ser Pro Pro 1 5 10 15 Ser Ala Phe Pro 20 <210> 10 <211> 24 <212> DNA <213> Artificial Sequence <220> <223> Description of Artificial Sequence: Oligonucleotide-Anti-sense <400> 10 tcgctcgaga g agtcttgct tctc 24 <210> 11 <211> 24 <212> DNA <213> Artificial Sequence <220> <223> Description of Artificial Sequence: Oligonucleotide <400> 11 ccactcgaga ctcgaagttg cctt 24 <210> 12 <211> 24 <212 > DNA <213> Artificial Sequence <220> <223> Description of Artificial Sequence: Oligonucleotide-Anti-sense <400> 12 cctctagatc atgtccattc tgaa 24 <210> 13 <211> 25 <212> DNA <213> Artificial Sequence <220 > <223> Description of Artificial Sequence: Oligonucleotide <400> 13 cagccaccag aagctcaccc ggtgc 25 <210> 14 <211> 27 <212> DNA <213> Artificial Sequence <220> <223> Description of Artificial Sequence: Oligonucleotide <400> 14 ccagaacctc acgcggtgcc acccatc 27 <210> 15 <211> 26 <212> DNA <213> Artificial Sequence <220> <223> Description of Artificial Sequence: Oligonucleotide <400> 15 cctcacccgg tggcacccat cctgac 26 <210> 16 <211> 25 <212> DNA <213> Artificial Sequence <220> <223> Description of Artificial Sequence: Oligonucleotide <400> 16 cacccggtgc cagccatcct gacgc 25 <210> 17 <211> 26 <2 12> DNA <213> Artificial Sequence <220> <223> Description of Artificial Sequence: Oligonucleotide <400> 17 cccatcctga cggcatcacc tccttc 26 <210> 18 <211> 24 <212> DNA <213> Artificial Sequence <220> <223 > Description of Artificial Sequence: Oligonucleotide <400> 18 ctgacgccat cagctccttc agcc 24 <210> 19 <211> 25 <212> DNA <213> Artificial Sequence <220> <223> Description of Artificial Sequence: Oligonucleotide <400> 19 acgccatcac ctgcttcagc cttcc 25 <210> 20 <211> 26 <212> DNA <213> Artificial Sequence <220> <223> Description of Artificial Sequence: Oligonucleotide <400> 20 ccttcagcct tcgcaaccgt taccac 26 <210> 21 <211> 20 <212> DNA <213> Artificial Sequence <220> <223> Description of Artificial Sequence: Oligonucleotide-Anti-sense <400> 21 ccatgtgcag cactggcttt 20 <210> 22 <211> 54 <212> PRT <213> Mus sp. <400> 22 Val Glu Gly Asp Ala Lys Glu Glu Ile Leu Gln Pro Pro Glu Pro His 1 5 10 15 Pro Val Pro Pro Ile Leu Thr Pro Ser Pro Pro Ser Ala Phe Pro Thr 20 25 30 Val Thr Thr Val Trp Gln Asp Ser Asp Arg Tyr His Pro Lys Pro Val 35 40 45 Leu His Met Ala Ser Pro 50 <210> 23 <211> 12 <212> PRT <213> Homo sapiens <400> 23 Gln Asp Ser Pro Pro Pro Lys Pro Pro Arg Thr Arg 1 5 10

[Brief description of the drawings]

FIG. 1A, FIG. 1B and FIG. 1C relate to the production of PTP-PEST null mice. FIG. 1A is a targeting vector map of the vector used to generate PTP-PEST null mice; FIG. 1B is a Southern blot of ES cells showing a 2 kb deletion of the PTP-PEST allele; ,
FIG. 1C shows Western blots of embryos showing different genotypes (+ / +, +/− and − / −).

FIG. 2A (1-4) and FIG. 2B (1-4) show the early development of malformations observed in embryos of PTP-PEST null mice.

FIG. 3 shows Western blots of − / − type and wild type embryos using anti-phosphotyrosine antibody.

FIG. 4 shows that PTP-PEST gene targeting suppresses fibroblast motility on extracellular matrix fibronectin. The monolayer of each cell line was injured and maintained at 37 ° C. for 72 hours before fixation. The ability of the cells to migrate to the damaged area was monitored by phase contrast microscopy of unstained cells and photographed (magnification: 100 times). The appearance of each lesion indicates typical results obtained after 5 independent experiments. As compared with the (+/-) type of FIG.
In the (-) type, it can be seen that the cells are moving.

FIG. 5 shows PTP-PEST heterozygous cells seeded on fibronectin (
5 (a), 5 (b), 5 (e) and 5 (f)) and homozygous cells (FIG.
(C), FIG. 5 (d), FIG. 5 (g) and FIG. 5 (h)) show fluorescent antibody staining images. After allowing the cells to stand for 25 minutes before fixation (FIGS. 5 (a) -5 (d)), actin filaments stained with the rhodamine-phalloidin complex (FIGS. 5 (a) and 5 (
c)) and no focal difference (FIGS. 5 (b) and 5 (d)) stained with an anti-vinculin antibody and visualized with a HITC-conjugated secondary antibody. However, after the cells were left for 3 hours before fixation, the +/- type cells became round (FIG. 5 (e)), and the focal adhesions like dots were slightly formed around the cell membrane. (FIG. 5F). On the other hand, the − / − type cells form many stress fibers in an expanded form (FIG. 5 (g)), and large focal adhesions are dispersed on the lower surface (FIG. 5 (h)). ). The magnification is 1,
000 times.

FIG. 6 shows PTP-PEST (+/−) type cells and (− / −) type cells,
Figure 4 shows constitutive tyrosine phosphorylation of cortactin, FAK and paxillin. p13
0cas is a substrate of PTP-PEST (Reference 6), and PTP-PEST (
It has been shown that hyperphosphorylation occurs in-/-) type cells, and was used here as a control. Only paxillin was found to be highly phosphorylated in PEST (-/-) type cells. This form of phosphorylation corresponded to the upper band of the double band of paxillin obtained by blotting with an anti-paxillin antibody (right drawing).
. Tyrosine phosphorylation of the focal adhesion component, vinculin, was not detectable in any of the cell lines, presumably because it did not stimulate the cells.

FIG. 7 shows unsynchronized PTP-PEST (+/−) type cells and (− / −)
Figure 4 shows constitutive phosphorylation of PSTPIP in type cells. PSTPIP was immunoprecipitated under non-denaturing conditions, and HRP-conjugated anti-phosphotyrosine antibody was used as a probe. PSTPIP is hyperphosphorylated in (-/-) type cells and appears to be further complexed with other, yet unidentified, tyrosine phosphorylated proteins (right lane).

FIG. 8 shows PTP- non-synchronized cultured on uncoated tissue culture slide glass.
FIG. 2 shows staining of PEST (− / −) type cells using a rhodamine-phalloidin complex. Arrows indicate the dividing furrows of each of the two pairs of cells in M phase, seen on the same field. Magnification: 400 times.

FIG. 9 shows PTP-PEST (+/−) type cell lines and PTP-PEST (− / −) cells established from primary culture cells of embryos isolated from PTP-PEST knockout mice.
1 shows Southern blot and Northern blot analysis of DNA and RNA isolated from-) type cell lines. FIG. 9A shows that genomic DNA was digested with BamHI and
-Results of analysis using a KpnI / SacI fragment of PEST cDNA as a probe. The 12 kb band corresponds to the wild type allele and the 7 kb band corresponds to the target allele. Wild-type DNA was used as a positive control. FIG. 9B shows that a 3.8 kb transcript was PTP-PEST (
(+/-) type cell line, but not detected in 5 of the PTP-PEST (-/-) type cell lines. RNA isolated from wild-type fibroblasts was used as a positive control (upper panel). Each blot was detected using GAPDH as a probe to show the same amount of RNA loaded (lower drawing).

FIG. 10 shows PTP-PEST (+/−) type cell lines and PTP-PEST (− / −).
The behavior of phosphotyrosine in-) type cell lines is shown. FIG. 10A shows PTP-PE
Cell lysate 15 of ST (+/-) type cells and PTP-PEST (-/-) type cells
pg was analyzed by immunoblotting against phosphotyrosine (first and second).
Lane) Results. PTP-PEST (+/-) type cell line and PTP-PES
Immunoprecipitates of the T (-/-) type cell line against phosphotyrosine were also analyzed by anti-phosphotyrosine immunoblotting. For PTP-PEST (-/-) type cells, 1
Highly phosphorylated proteins of 80, 130 and 97 kDa were detected. FIG.
OB is p130 in the PTP-PEST deficient cell line (PTP-PEST − / −).
Indicates that cas is hyperphosphorylated. Upper drawing: PTP-PEST +/-
P130cas was immunoprecipitated from cell lines and PTP-PEST − / − cell lines, and the degree of phosphorylation was analyzed by anti-phosphotyrosine immunoblotting using a 4G10 monoclonal antibody. Lower drawing: After the above monoclonal antibody was peeled off, blotting was again performed using the anti-p130cas B + F antibody as a probe, and the same amount of p was added to the total cell lysate (hereinafter often abbreviated as “TCL”) and immunoprecipitate.
130 cas was confirmed to be present.

FIG. 11 shows that p130cas is a physiological and specific substrate of PTP-PEST by a substrate trap experiment on a lysate of PTP-PEST-deficient cell (− / −) cells. The substrate trap experiment was performed using PTP-PEST (+/-)
1 mg of each cell lysate of cell type, PTP-PEST (− / −) type cell, and COS-1 cell treated with pervanadate was added to 100 ng of GST-PTP-PES.
T WT (aa.1-453) or GST-PTP-PEST C231S (1
-453). Analysis of the bound protein was performed by Western blotting against phosphotyrosine using a 4G10 monoclonal antibody (upper drawing), or after removing the antibody, anti-p130cas
The blotting was performed by B + F antibody (middle drawing). The gel in which the GST fusion protein used for substrate trapping was stained with Coomassie blue is shown in the lower drawing. This figure shows that the fusion protein was correctly expressed and serves as a control for electrophoresis. Two arrows are written beside the band corresponding to pp130 (phosphorylated p130 protein) to highlight the diffuse band.

FIG. 12 shows that PTP-PEST was transformed with p130cas, Hef1 and S in vitro.
2 shows binding to the SH3 domain of in. Cell lysate (1 mg) of COS-1 cells transformed with HA-PTP-PEST was pre-bound to glutathione sepharose, p130cas linked to 100 ng GST, Hef1
Alternatively, it was incubated with the SH3 domain of Sin or GST alone. After washing the Sepharose thoroughly, the protein was eluted and separated by SDS-PAGE,
HA-PT by Western blotting using 12CA5 monoclonal antibody
The presence of P-PEST was analyzed (left drawing). The right drawing is a gel stained with Coomassie blue for the GST fusion protein used in the in vitro binding experiment, and shows that the fusion protein is correctly expressed.

FIG. 13 shows that the proline-rich region of PTP-PEST, Pro1, is p1
It shows that it controls the interaction between the SH3 domain of the adapter molecule family consisting of 30cas, Sin and Hef1 and PTP-PEST. FIG. 13A is a schematic diagram of a PTP-PEST GST fusion protein with different proline-rich regions (Pro1-5) of PTP-PEST used in the Far Western binding assay. FIG. 13B shows that 100 ng of P used in the Far Western binding assay.
TP-PEST GST fusion gel stained with Coomassie blue, also showing proteins transferred to PVDF membrane. PTP-P
Three blots containing the EST fusion protein were each treated with p130cas (
13C), incubated with the GST SH3 domain radiolabeled with [ ³² P] of Sin (FIG. 13D) and Hef1 (FIG. 13E), followed by extensive washing and exposure to X-ray film for 20 minutes.

FIG. 14 shows that PTP-PEST and p130cas bind in vivo . Myc tag-introduced p130cas vector or Myc SH3 domain
The p130cas vector was transformed into HA-PTP-PEST, WT PEST or C2.
293T cells were co-transfected with 31S. For p130cas protein with various Myc tags introduced, 107 of PTP-PEST protein
Five immunoprecipitates were analyzed. FIG. 14A is a schematic diagram of a p130cas protein into which a Myc tag has been introduced, used in a co-immunoprecipitation experiment. FIG. 14B shows that following co-immunoprecipitation, the resulting complex was washed extensively and the bound protein was removed from SDS-PA.
GE development, transfer protein to PVDF membrane, anti-Myc 9E10
By Western blotting using a monoclonal antibody, various Myc
The presence of the tag-introduced p130cas protein was confirmed. As a negative control, T cell PTP into which the HA tag was introduced (HA TC-PTP) was Mycp
The immunoprecipitate co-transfected with the 130cas construct and obtained with the anti-HA antibody 12CA5 was used. 130 kD for full-length Myc p130cas protein
a slightly upstream of a, the MycSH3 domain of p130cas is 30 kDa
Moved to FIG. 14C shows the results of Western blotting of TCL using the 9E10 antibody and the 12CA5 antibody, and various Myc p130ca visualized.
s protein, HA-PTP-PEST and HA-TC-PTP.

FIG. 15 shows that PTP-PEST binds to paxillin in various mouse tissues. PTP-PEST was immunoprecipitated from 1 mg of lysate of liver, brain, heart, kidney, lung, spleen and thymus, and the presence of bound paxillin f was confirmed by Western blotting using an antibody against paxillin. Paxillin
Liver, brain, heart, lung, spleen and thymus lysates were present in PTP-PEST immunoprecipitates, but not in kidney lysates. In addition, when immunoprecipitation and Western blotting were performed on the liver lysate in the same manner as described above after preliminary immunoprecipitation, paxillin was not detected.

FIG. 16 shows that Pro2 of PTP-PEST is required for paxillin binding in NIH 3T3 cells. FIG. 16A shows the PTP used in the binding assay.
FIG. 2 is a schematic diagram of a PEST GST fusion protein. The results of incubating the PTP-PEST C-terminal deletion mutant or proline-rich motif with the lysate of NIH 3T3 cells and detecting bound paxillin by Western blotting are shown in the upper drawings of FIGS. 16B and 16C, respectively. 16B and 16C are gels in which the fusion protein used in the binding assay is stained with Coomassie blue.

FIG. 17 shows that PTP-PEST Pro2 is essential for paxillin binding in vitro and in vivo . FIG. 17A: HA-WT or HAΔPr
HEK293T cells were transiently transformed with o2 PTP-PEST. 200 μg of cell lysates of the transformants were incubated with GST paxillin N, GST paxillin C, GST SH3p130cas or GST alone.
Bound PTP-PEST was detected by western blotting using the 12CA5 antibody (upper panel). 5 μg of TCL was also blotted with the 12CA5 antibody to confirm that the expression levels of HA-WT and HAΔPro2 PTP-PEST were equivalent to PTP-PEST. The presence of the GST fusion protein and its correct expression were confirmed by re-blotting using a polyclonal antibody against GST. FIG. 17B: Pseudo-recombinant (empty pACTAG)
, HA-WT plasmid, HAΔPro1 plasmid or HAΔPro2 PT
HEK 293T cells were transiently transformed with any of the P-PEST plasmids. Correct expression of each construct was determined by HA to 5 μg TCL.
It was confirmed by performing immunoblotting using the antibody 12CA5. Paxillin was immunoprecipitated with a monoclonal antibody, and the presence of PTP-PEST was determined using anti-HA.
It was confirmed by immunoblotting using the antibody 12CA5 (middle part drawing). The presence of almost the same amount of paxillin in all immunoprecipitates was confirmed by re-blotting using a monoclonal antibody against paxillin (lower drawing).

FIG. 18 is, P362A mutations in the Pro2 of PTP-PEST is, in vitro
Fig. 5 shows that the binding of paxillin is lost. FIG. 18A shows the GST PT
It is a schematic diagram of the mutant which substituted proline of the P-PEST Pro2 construct (amino acids 344-437) with alanine. The region shown in the figure is the paxillin binding site in FIG. FIG. 18B shows that GST Pro 2 and eight mutants in which proline was substituted for alanine were purified, and each product was incubated with 200 μg of a lysate of NIH 3T3 cells. Bound paxillin was detected by Western blotting using a monoclonal antibody against paxillin. (C) In order to confirm the expression of the fusion protein used in the binding experiment and that the expression product was correct, an equivalent amount of the sample from each experiment was separated by SDS-PAGE and stained with Coomassie blue.

FIG. 19 shows that LAX domains 3 and 4 of paxillin are PTP-P in vitro .
Indicates that it is required for binding to EST. FIG. 19A is a schematic diagram showing the region of use of a fusion protein between the LIM domain of paxillin and GST, which was used to identify the PTP-PEST binding site. FIG. 19B: HEK29 expressing purified Paxillin GST fusion protein and transiently expressing HA-PTP-PEST
Incubated with cell lysates of 3T cells. PTP-P bound to paxillin
EST was detected by immunoblotting using anti-HA antibody 12CA5 (upper drawing). The gel developed by SDS-PAGE was stained with Coomassie blue to confirm the expression of each fusion protein and that the expression product was correct.

FIG. 20 is a LIM domain 3 of paxillin 4, in the in vivo PTP-
Required for binding to PEST. HEK 293T cells were transiently co-transfected with HA-PTP-PEST and pcDNA3, wild-type paxillin (WT), ΔLIM3 or ΔLIM4 tripaxillin. FIG.
Upper panel of A: 48 hours after transformation, paxillin was immunoprecipitated from a lysate of each cell using a tripaxillin-specific polyclonal antibody, and co-precipitated PTP-P.
EST was analyzed by immunoblotting using anti-HA antibody 12CA5. FIG.
Bottom drawing of OA: Each blot was blotted again using a monoclonal antibody against paxillin, confirming that the same amount of immunoprecipitate as PTP-PEST was obtained. Paxillin was not detected in the sample using the empty vector (pcDNA3). FIG. 20B: HA-P of each transformant
The expression level of TP-PEST was detected by immunoblotting 5 μl of TCL using the anti-HA antibody 12CA5.

FIG. 21 shows that binding of paxillin to PTP-PEST requires complete LIM3 and LTP.
Indicates that IM4 is required. pcDNA3, wild-type paxillin (WT) and paxillin mutants (ΔLIM3, C467A, C470A, C467 / 470A
, ΔLIM4 and C523S) were translated in vitro in the presence of ³⁵ S-methionine. These translation products were converted to GST-PTP-PEST Pro2 (344
-397). After repeatedly washing the GST matrix, the bound proteins were subsequently separated by 10% SDS-PAGE and the bound proteins were visualized by exposing the film for 8 hours (upper drawing).
GST alone was incubated with wild-type paxillin to serve as a negative control. In the lower figure, 15% of the amount of each paxillin product used in the binding assay was used to show that each transgene was correctly expressed (exposure time: 8 hours).

FIG. 22 shows that paxillin is not a substrate for PTP-PEST. NIH
Lysates of 3T3 cells, NIH3T3 cells expressing Src Y527F, and NIH3T3 cells treated with pervanadate were each incubated with four different PTP-PEST GST fusion proteins: C-terminal (
aa. 276-775), 1-453WT, 1-453 C231S, and 1-
354 C231S. The bound protein was developed by SDS-PAGE, and the tyrosine phosphorylated protein was detected by immunoblotting using an anti-phosphotyrosine antibody (a). Each blot was blotted again using the paxillin monoclonal antibody (b) or anti-p130cas monoclonal antibody (c) as a probe. (D) The expression of each fusion protein and that the expression product was correct was confirmed by staining the blot on PVDF with Coomassie blue.

FIG. 23 shows a functional model of PTP-PEST in focal adhesion. FIG. 23A: Simplified model of proteins involved in focal adhesion. PTP-PEST moves to focal adhesion by a mechanism not currently known (14). So paxillin LI
Binding to M3-4 and then to the SH3 domain of p130cas,
Dephosphorylate tyrosine residues (pY) of p130cas required for binding to domain-containing proteins (eg, Crk). By dephosphorylating p130cas, PTP-PEST is thought to inhibit downstream signaling starting from Src, SOS and C3G. Csk bound to PTP-PEST is also Src
Src can be inhibited by phosphorylating the negative regulatory site of. FIG. 23B: Paxillin binds to p130cas as a result of PTP-PEST moving to focal adhesions. Paxillin LIM (Reference 19) and p130
The SH3 domain of cas (40) is required to move these proteins into focal adhesions. We believe that proteins bound to PTP-PEST migrate to the cytoplasm as a result. Focal adhesion formation is further inhibited by inactivation of Src activity by Csk bound to PTP-PEST. These events imply that PTP-PEST likes the collapse of focal adhesions and allows cell migration before the focal adhesions are reformed.

FIG. 24 shows the complete sequence of mouse and human PEST (SEQ ID NOS: 1 and 2, respectively).
). Proline-rich domains that are highly conserved among species are boxed.

──────────────────────────────────────────────────続 き Continued on the front page (51) Int.Cl. ⁷ Identification symbol FI Theme coat ゛ (Reference) A61P 43/00 C07K 14/00 C07K 14/00 14/17 14/17 C12N 9/16 B C12N 9/16 C12Q 1/02 15/09 ZNA 1/42 C12Q 1/02 A61K 37/02 1/42 C12N 15/00 ZNAA (81) Designated countries EP (AT, BE, CH, CY, DE, DK, ES, FI, FR, GB, GR, IE, IT, LU, MC, NL, PT, SE), OA (BF, BJ, CF, CG, CI, CM, GA, GN, GW, ML, MR, NE, SN, TD, TG), AP (GH, GM, KE, LS, MW, SD, SL, SZ, UG, ZW), EA (AM, AZ, BY, KG, KZ, MD, RU, TJ, TM) , E, AL, AM, AT, AU, AZ, BA, BB, BG, BR, BY, CA, CH, CN, CU, CZ, DE, DK, EE, ES, FI, GB, GD, GE, GH , GM, HR, HU, ID, IL, IN, IS, JP, KE, KG, KP, KR, KZ, LC, LK, LR, LS, LT, LU, LV, MD, MG, MK, MN, MW, MX, NO, NZ, PL, PT, RO, RU, SD, SE, SG, SI, SK, SL, TJ, TM, TR, TT, UA, UG, US, UZ, VN, YU, ZA , ZW (72) Inventor Angel-Roustou, Alexandre H2A3G8, Canada, Montreal, Quebec, 2M Avenue 7217 (72) Inventor Charle, Alan United States, Massachusetts 02150, Chelsea, AD Lull's Way 19 F-term (reference) 4B024 AA01 BA10 CA04 CA07 DA02 DA05 EA02 EA04 FA02 GA11 HA01 4B050 CC03 CC05 DD07 LL01 4B063 QA01 QA08 QQ01 QQ08 QQ26 QR06 QR33 QR48 QR57 QR59 QR62 QR77 QR80 QS05 BAQA48 QS25 AQA4A NA14 ZB112 ZB262 4H045 AA20 AA30 BA14 BA15 CA40 DA89 EA20 FA10 FA72 FA74 [Continued from summary] The present invention relates to a method for detecting an enzyme substrate. The use of gene-targeted knockout generation methods results in fewer falsely detected results than other techniques for artificially increasing nonspecific hyperphosphorylation levels (eg, using vanadate compounds).

Claims (23)

[Claims] 1. A compound which interferes with the binding of protein tyrosine phosphatase PEST (PTP-PEST) to a signaling molecule involved in cell migration, cell adhesion or cell division. 2. The compound according to claim 1, wherein said signaling molecule is p130cas. 3. The compound according to claim 1, wherein said signaling molecule is paxillin. 4. The compound according to claim 1, wherein the compound is a peptide containing amino acid residues 333 to 338 of SEQ ID NO: 1 or amino acid residues 334 to 339 of SEQ ID NO: 2. 3. The compound according to 2. 5. The compound according to claim 1, wherein the compound has the amino acid sequence of SEQ ID NO: 4 or 5, or p130ca
The compound according to claim 4, characterized in that the compound comprises a derivative or a partial sequence of the amino acid sequence capable of binding to s. 6. The compound according to claim 1, wherein the compound is a peptide containing the amino acid residues 355 to 363 of SEQ ID NO: 1 or the amino acid residues 356 to 364 of SEQ ID NO: 2. 3. The compound according to 3. 7. The compound according to claim 6, wherein the compound comprises the amino acid sequence of SEQ ID NO: 22 or 23, or a derivative or a partial sequence of the amino acid sequence capable of binding to paxillin. 8. A medicament comprising an amount of the compound according to any one of claims 1 to 7 and a pharmaceutically acceptable carrier exhibiting anti-cell proliferation, anti-cell adhesion or anti-cell migration. Composition. 9. The pharmaceutical composition according to claim 8, wherein the compound is administered so that the extracellular concentration of the compound is 10 nM to 1 mM. 10. An administration drug for the treatment of a disease caused by any phenomenon selected from the group consisting of cell proliferation, cell migration, inflammation and angiogenesis.
The use of the compound according to any one of claims 1 to 7. 11. The use according to claim 10, wherein the disease is cancer. 12. A method for detecting a true enzyme substrate for a wild-type enzyme in a cell which expresses the enzyme in a normal state, wherein the cell is modified so that the expression of the enzyme is suppressed, and the silenced cell is modified. Obtaining a strain, thereby causing a detectable change in said intrinsic enzyme substrate by said modification relative to the enzyme substrate of a cell expressing the wild-type enzyme; Providing a construct comprising a variant of the enzyme that exhibits no activity but is capable of binding to the enzyme substrate; contacting the cell contents of the silenced cell line with the construct to form the enzyme substrate and the enzyme Obtaining a complex with the mutant; and detecting the enzyme substrate. 13. The method according to claim 12, wherein the silenced cell line is a null mutant cell line of the enzyme. 14. The method according to claim 12, wherein the enzyme is a protein phosphatase. 15. The method according to claim 14, wherein the protein phosphatase is a protein tyrosine phosphatase. 16. The method according to claim 15, wherein said protein tyrosine phosphatase is PTP-PEST. 17. The enzyme mutant is a mutant PTP-PEST, which is obtained by replacing the 231st cysteine residue of SEQ ID NO: 1 or 2 with a serine residue. The method according to claim 16, characterized in that: 18. The method according to claim 14, wherein the change caused in the enzyme substrate by the modification is hyperphosphorylation of the enzyme substrate. 19. The method according to claim 18, wherein the detecting step is performed using a ligand specific to a highly phosphorylated enzyme substrate. 20. The method according to claim 19, wherein said ligand is an antibody. 21. The method according to claim 16, wherein the presence of the hyperphosphorylated substrate p130cas is detected as a positive control indicating the presence of the intrinsic enzyme substrate.
0. The method according to any of 0. 22. The method according to claim 12, wherein the construct containing the enzyme variant is immobilized on a solid support. 23. The method according to claim 12, wherein the enzyme substrate is separated from a complex of the enzyme substrate and the enzyme variant before the step of detecting the enzyme substrate.
The method according to any of the above.