MIM (Missing-in-Metastasis), also known as MTSS1 (metastasis suppressor 1), is a scaffold protein that is down-regulated in multiple metastatic cancer cell lines compared with non-metastatic counterparts. MIM regulates cytoskeletal dynamics and actin polymerization, and has been implicated in the control of cell motility and invasion. MIM has also been shown to bind to a receptor PTP (protein tyrosine phosphatase), PTPδ, an interaction that may provide a link between tyrosine-phosphorylation-dependent signalling and metastasis. We used shRNA-mediated gene silencing to investigate the consequences of loss of MIM on the migration and invasion of the MCF10A mammary epithelial cell model of breast cancer. We observed that suppression of MIM by RNAi enhanced migration and invasion of MCF10A cells, effects that were associated with increased levels of PTPδ. Furthermore, analysis of human clinical data indicated that PTPδ was elevated in breast cancer samples when compared with normal tissue. We demonstrated that the SRC protein tyrosine kinase is a direct substrate of PTPδ and, upon suppression of MIM, we observed changes in the phosphorylation status of SRC; in particular, the inhibitory site (Tyr527) was hypophosphorylated, whereas the activating autophosphorylation site (Tyr416) was hyperphosphorylated. Thus the absence of MIM led to PTPδ-mediated activation of SRC. Finally, the SRC inhibitor SU6656 counteracted the effects of MIM suppression on cell motility and invasion. The present study illustrates that both SRC and PTPδ have the potential to be therapeutic targets for metastatic tumours associated with loss of MIM.
According to the American Cancer Society (Breast Cancer Facts and Figures 2013–2014), approximately 233000 women will be diagnosed with invasive breast cancer in the U.S.A. in 2014, which is predicted to lead to 40000 deaths. It is expected that breast cancer will account for 29% of all new cancers diagnosed in women. Metastasis, the spread of cancer from the primary tumour to remote sites, is the major cause of fatality, resulting in approximately 90% of deaths . The pathophysiology of metastasis encompasses release of cells from the primary tumour site, transport of the cells in the blood vessels or the lymphatic system, attachment to the remote site, and invasion and colonization of secondary site . It is therefore of utmost importance to gain an understanding the mechanisms that facilitate the invasive transition in breast cancer and the complex processes that underlie changes in cell migration and cell invasion. Although considerable attention has focused on the oncogenes and tumour suppressors that trigger the establishment of primary tumours, it is now apparent that there is also a class of metastasis suppressor genes, the products of which exert a regulatory influence at each step in the development of a metastatic state without affecting growth of the primary tumour . MIM (Missing in Metastasis), also referred to as MTSS1 (metastasis suppressor 1), is one such metastasis suppressor, with potential roles in the control of cellular migration and invasion. MIM is expressed in a diverse range of tissues and is down-regulated in several cancers, including breast [4–8]. Thus it has been proposed that loss of MIM function may promote the metastatic potential of cancer cells.
MIM is a multidomain protein, the structure of which suggests a scaffolding function . The presence of a WH2 [WASP (Wiskott–Aldrich syndrome protein) homology 2] domain and IMD [IRSp53 (insulin receptor substrate protein of 53 kDa) and MIM homology domain] implies a functional link with the actin cytoskeleton and recruitment of MIM to specific cytoskeletal networks. Such interactions implicate MIM in cytoskeletal changes that underlie regulation of metastasis. The central segment of MIM is rich in proline, serine and threonine residues, and plays an important regulatory role in MIM function. Importantly, this central segment of MIM binds to several key cellular proteins, including association with the RPTP (receptor protein tyrosine phosphatase) family protein PTPδ (protein tyrosine phosphatase δ) . Interaction studies using the yeast two-hybrid approach and analysis of MIM interaction with recombinant PTPδ illustrate that MIM binds to the cytoplasmic domain of PTPδ . Since RPTPs are known to regulate tyrosine-phosphorylation-dependent signalling, this MIM–PTPδ interaction may provide a functional link that influences the signal transduction events that underlie the establishment of an invasive state.
The RPTPs are transmembrane proteins that have the potential to regulate signalling through ligand-controlled protein tyrosine dephosphorylation [11,12]. The LAR (leucocyte common antigen)-like RPTPs are a subfamily of these receptor-like proteins that have been extensively characterized in neurons, and have been directly implicated in axon growth regulation in invertebrates . The LAR-like PTPs (LAR, PTPσ and PTPδ) contain two intracellular phosphatase domains and an extracellular segment consisting of Ig-like and fibronectin type III-like domains, commonly found in cell adhesion molecules . This structural architecture implies that LAR-like phosphatases may function in integrating tyrosine-phosphorylation-dependent cellular signalling with cell–cell and cell–matrix interactions. In fact, LAR-like RPTPs have been implicated in biological processes that rely upon regulated cell–cell and cell–matrix contact, such as control of neuronal pathfinding [14,15]. LAR localizes to the ends of focal adhesions, implicating LAR-like RPTPs in their disassembly . Signalling events downstream of LAR-like RPTPs that regulate axonal guidance involve Rho-GTPase-dependent cytoskeletal remodelling . It has also been demonstrated, using LAR−/− mice, that LAR plays a significant role in the signalling events responsible for mammary gland development and function . Overall, this suggests that RPTPs, in particular those of the LAR subfamily, might be important regulators of signalling in breast cancer.
The LAR-like enzyme PTPδ is also expressed in the central nervous system, where it is concentrated in growth cones of elongating processes . PTPδ is a homophilic cell adhesion molecule . These homophilic interactions serve to promote neuronal adhesion and neurite outgrowth [20,21]. PTPδ has been suggested to be a tumour suppressor due to its inactivation in a number of human cancers, including head and neck, melanoma, lung cancer and neuroblastoma [22–24]. Multiple mutations have been identified in tumours that may compromise not only activity, but also the function of the extracellular segment . Chromosome 9p, which harbours PTPRD (gene encoding PTPδ), is also a frequent target of microdeletion in primary tumours and is subject to chromosome shedding in 6% of tumours studied [25,26]. In contrast with genomic studies, PTPδ loss of function in mice is associated with impaired learning, but has not been reported to increase tumour incidence . Furthermore, reconstitution studies failed to demonstrate a growth-suppressive function for PTPδ . Therefore, consistent with the complex role of other PTPs in cancer [29,30], it appears that the function of PTPδ may be context-dependent. In the present study, we have expanded the potential roles of PTPδ in cancer by testing the hypothesis that it functions in the regulation of tyrosine-phosphorylation-dependent signalling events that underlie cell motility and cell invasion in MIM-negative cells. We present evidence that suppression of MIM led to increased expression of PTPδ, which enhanced invasion of breast epithelial cells through activation of the protein tyrosine kinase SRC. These data define a mechanism by which MIM may exert activity as a metastasis suppressor through regulating tyrosine-phosphorylation-dependent signalling in breast epithelial cells.
Anti-PTPδ antibody was from Novus Biologicals. Stained tissue sections in the Human Protein Atlas (http://www.proteinatlas.org) were generated using the same antibody. Antibodies against SRC-pTyr527, SRC-pTyr416 and total SRC protein, cortactin-pTyr421 and total cortactin, as well as antibodies against MIM, were from Cell Signaling Technology.
MCF-10A cells were obtained from A.T.C.C. (Manassas, VA, U.S.A.) and cultured in Dulbecco's modified Eagle medium (DMEM)/Ham's F12 (Invitrogen) supplemented with 5% (v/v) donor horse serum, 20 ng/ml EGF (epidermal growth factor), 10 μg/ml insulin, 100 ng/ml hydrocortisone, 100 ng/ml cholera toxin, 100 unit/ml penicillin and 100 μg/ml streptomycin. Growth-factor-reduced Matrigel™ was purchased from BD Biosciences.
Generation of cells expressing shRNA targeting MIM and PTPδ
For stable suppression of MIM in MCF10A cells, we expressed a pMLP retroviral vector (in a pMSCV backbone) using the targeting sequences 5′-TCTTCTGCAGCTTCAGCGT-3′ and 5′-TCTTTTTGATCTCATGCCG-3′ incorporated into the sequence of human miR-30. The infected cells were selected using puromycin (1–2 μg). For double selection, PTPRD shRNA, using the targeting sequence 5′-TGCATACATCTTAGACTCT-3′, was subcloned in pMSCV hygro and selected using hygromycin (100 μg/ml). pcDF1-PTPRD (plasmid 25642) was ordered from Addgene. Infections were carried out as described previously . The GST–PTPδ fusion construct in pGEX vector was a gift from Dr Timothy Chan. Inactive (C1553S) and substrate-trapping (D1521A) mutations were engineered into pcDF1-PTPRD and pGEX-PTPRD constructs using site-directed mutagenesis (QuikChange® II XL kit from Stratagene) as directed by the manufacturer. The coding sequences were verified by DNA sequencing.
Cell migration and invasion assays
Cell motility was measured using Cell Culture Inserts (8.0-μm pore size) for six-well plates (BD Falcon). To visualize cell invasion, we used eight-well chamber slides (BD Biosciences) pre-coated with 70 μl of 1:1 mixture of Matrigel™ and collagen I (BD Biosciences). On day 1, 4000 cells were grown per well in the presence of 5 ng/ml EGF . Cell morphology was photographed on days 8 and 10. The phase-contrast images were taken using a Zeiss Axiovert 200M microscope using AxioVison 4.4 software. To quantify cell invasion, we used BD BioCoat Matrigel™ Invasion Chambers, 8.0-μm pore size. MCF10A cells (106) were grown in the insert. After 48 h, the cells retained inside the insert were removed, and those that migrated to the other side of the insert were fixed and stained with DAPI and counted.
Quantitative real-time PCR (qPCR)
Total RNA from 70–80% cells was isolated using TRIzol® (Invitrogen), DNase I-treated and reverse-transcribed using reverse transcriptase and random hexamers. qPCR was performed for PTPRD according to the manufacturer's recommendations (Applied Biosystems). The mean cycle threshold (CT) value was used to calculate the gene expression. PCR products were normalized to β-actin levels. Primers used for PTPRD were 5′-AGAGAG-AAATGTCACCAATA-3′ and 5′-AATTCCCTTAGGATATAC-TG-3′ and for actin were 5′-TCCCTGGAGAAGAGCT-ACG-3′, 5′-GTAGTTTCGTGCATGCCACA-3′.
MCF10A cells expressing the appropriate shRNA were serum-starved overnight, followed by treatment with cycloheximide (100 μg/ml). Cell lysates were collected at the indicated times and protein concentrations were determined by the Bradford assay. Equal amounts of lysates were loaded and proteins were resolved by SDS/PAGE and detected by immunoblotting.
Immunoprecipitation and immunoblotting
Cells were grown to 90% confluence in 10-cm-diameter plates, washed with ice-cold PBS and extracted using 800 μl of lysis buffer consisting of 50 mM Tris/HCl (pH 7.5), 150 mM NaCl, 5 mM EDTA, 1% (w/v) Triton X-100, 0.1% sodium deoxycholate, 20 mM 2-glycerophosphate, 1 mM sodium vanadate, 1 mM sodium fluoride and protease inhibitor cocktail. All steps were carried out on ice or at 4°C. Cells were lysed for 1 h, centrifuged at 12000 g for 10 min and protein concentrations were determined using the Bradford assay. Lysates (total protein 1 mg) were pre-cleared for 60 min with Protein A/G–Sepharose. The supernatants were first incubated for 60 min with appropriate antibodies and 10 μl of Protein A/G–Sepharose was then added for another 60 min. The immune complexes were pelleted at 3000 g for 5 min and washed three times with ice-cold lysis buffer. The beads were resuspended in 20 μl of 5× Laemmli sample buffer and heated at 95°C for 1 min. Proteins were resolved by SDS/PAGE and detected by immunoblotting.
Serum-starved MCF10A cells expressing wild-type or substrate-trapping DA mutant of PTPδ were pre-treated with 50 μM pervanadate for 30 min. Cells were rinsed with ice-cold PBS and lysed in 50 mM Tris/HCl (pH 7.5), 150 mM NaCl, 5 mM EDTA, 1% (w/v) Triton X-100, 0.1% sodium deoxycholate and protease inhibitor tablet (EDTA-free, Roche). DTT was added to a final concentration of 10 mM and cells were lysed for 30 min on a rotating wheel at 4°C. Cell debris was centrifuged at 12000 g for 10 min and protein concentrations were determined by the method of Bradford. An equal amount of lysate (100 μg) was diluted in ice-cold lysis buffer and pre-cleared for 30 min with anti-FLAG–agarose beads. To test for PTPδ substrate-trapping capacity, lysates expressing either wild-type or PTPδ-trapping mutant were immunoprecipitated with anti-PTPδ antibody. The PTPδ–substrate complexes were pelleted at 3000 g for 5 min, washed three times with lysis buffer and resuspended in Laemmli sample buffer, and binding of SRC was detected by immunoblotting.
In vitro phosphatase assay
GST-tagged wild-type PTPδ and CS mutant were purified on glutathione–agarose . The reduced enzyme was then incubated with immunoprecipitated SRC at 30°C for 30 min. The reaction was terminated with Laemmli sample buffer, proteins were resolved by SDS/PAGE and substrate dephosphorylation was visualized by immunoblotting.
Loss of MIM induced migration of mammary epithelial cells
In breast carcinoma, the ability of tumour cells to migrate from the primary site and establish a secondary tumour is an early determinant of the transition from a benign to an invasive state. Although it has been reported that loss of MIM transcript correlates with the invasive state of breast cancer cell lines , the role of MIM during cell migration and cell invasion has not been elucidated. To investigate its role in the migration and invasion of mammary epithelial cells, we suppressed MIM using RNAi and generated stable MCF10A populations expressing MIM-directed shRNA. Depletion of MIM by two distinct shRNAs was validated by immunoblotting (Figure 1A), and the effects were investigated using a Transwell migration assay. We observed that suppression of MIM resulted in up to a 4-fold increase in migration, compared with those cells stably expressing a control shRNA (Figure 1B). In addition, MCF10A cells normally form clusters when grown as subconfluent cultures ; however, we observed that suppression of MIM led to a scattering phenotype with a preponderance of single cells that do not form clusters (Figure 1C). In contrast, we observed that suppression of MIM had no apparent effect on cell proliferation (Figure 1D).
Suppression of MIM induced an increase in mammary epithelial cell motility
Loss of MIM induced invasion of mammary epithelial cells
Given the intrinsic complexity of the invasion process, 2D culture systems are limited in their utility; however, 3D culture in modified Matrigel™/collagen matrices has permitted modelling of cell invasion . Consistent with the migration assay, we observed that attenuation of MIM expression in MCF10A cells led to disorganized acinar structures and formation of cellular protrusions indicative of an invasive state (Figure 2A). In addition, we performed quantitative assays in Transwell filters coated with Matrigel™ and observed that cells in which MIM was depleted displayed an ~5-fold increase in invasion compared with the parental cells (Figure 2B). These results are consistent with a role for MIM in regulating cell migration and invasion, both important features of metastasis.
Suppression of MIM promotes mammary epithelial cell invasion
Loss of MIM induced an increase in mRNA and protein levels of PTPδ
As MIM and PTPδ have been shown to be binding partners , this interaction may underlie a functional link between MIM and tyrosine-phosphorylation-dependent signalling events associated with cell migration and invasion. We observed that the level of PTPδ mRNA (Figure 3A) and protein (Figure 3B) was elevated in cells in which MIM was suppressed compared with the control cells; however, the stability of PTPδ protein was apparently unaffected, as assessed by immunoblotting to measure PTPδ protein in cycloheximide-treated compared with control cells (Figures 3C and 3D).
Suppression of MIM increased the mRNA and protein levels of PTPδ
PTPδ expression in human breast cancer
To investigate whether there were alterations in the level of expression of PTPδ in human breast cancer, we analysed microarray gene expression data from Finak et al.  for the level of PTPRD mRNA in 53 breast tumour stroma samples, compared with six normal breast stroma samples (data available at http://www.oncomine.org). This revealed that PTPRD mRNA levels were elevated in the tumour samples when compared with the normal samples (Figure 4A, 2.4-fold change, P = 7.94×10−15). We also investigated PTPδ protein expression using immunohistochemistry data from 11 patient samples that were available in the Human Protein Atlas . The normal breast tissue showed minimal staining for PTPδ (Figure 4B), whereas a patient with ductal carcinoma showed more intense staining for PTPδ. We observed high PTPδ staining in more than 50% of patient samples when compared with the normal tissues (Figure 4C). In addition, we attempted to determine whether there was a correlation between MIM expression and any clinical variables. However, this proved to be intractable owing to the genomic location of MIM, close to the c-MYC gene (hg18), which is part of a genomic region commonly amplified in breast cancer samples .
Expression of PTPδ in human tumour samples
The invasion phenotype induced by loss of MIM was mediated by altered levels of PTPδ
To investigate the importance of PTPδ for the enhanced cell motility and invasion caused by MIM suppression, we generated cell lines co-expressing shRNA targeting both MIM and PTPδ. We observed no apparent effects on cell invasion of suppressing PTPδ alone, but the co-expression of both MIM and PTPδ-directed shRNAs resulted in abrogation of the invasion phenotype caused by suppression of MIM alone (Figure 5A). This illustrates that PTPδ is required for the MIM-dependent signalling events that underlie cell invasion.
The formation of invasive structures that accompanied suppression of MIM was abrogated by co-suppression of PTPδ
We investigated further this effect of PTPδ on MIM signalling using quantitative invasion assays performed on Transwell invasion chambers (Figure 5B). Suppression of MIM resulted in an ~5-fold increase in cell invasion that was partially blocked upon co-suppression of MIM and PTPδ. To determine whether the enzymatic activity of PTPδ was required for its effects on cell invasion, we generated stable cells in which wild-type PTPδ and the catalytically inactive CS mutant were overexpressed and observed that wild-type PTPδ, but not the inactive CS mutant, resulted in an increase in cell invasion (Figures 5C and 5D). This indicated that elevated levels of PTPδ were sufficient to enhance invasion, highlighting further the importance of PTPδ in mediating MIM functions.
The invasion phenotype induced by loss of MIM was mediated by activation of SRC
MIM regulates Sonic Hedgehog signalling and promotes ciliogenesis by antagonizing SRC-dependent phosphorylation of cortactin . SRC contains an N-terminal myristoyl group, an SH3 (SRC homology 3) domain, an SH2 domain, a protein tyrosine kinase domain and a C-terminal regulatory tail . Under basal conditions, 90% of SRC is phosphorylated at Tyr527 in the C-terminal tail, which undergoes intramolecular association with the SH2 domain, blocking substrate binding and thereby rendering SRC inactive . Dephosphorylation of Tyr527 represents an important mechanism for SRC activation, being accompanied by autophosphorylation at Tyr416 in the activation loop, which promotes kinase activity. When we examined the phosphorylation status of SRC in MIM-depleted cells, we observed that phosphorylation at Tyr527 was markedly decreased (Figure 6A), whereas autophosphorylation on Tyr416 was enhanced (Figure 6B). These effects were accompanied by the expected changes in levels of MIM and PTPδ (Figure 6C). Furthermore, suppression of MIM was accompanied by enhanced phosphorylation of Tyr421 in cortactin, a direct substrate of SRC (Figure 6D). These data are consistent with an activating role for PTPδ in the regulation of SRC.
Suppression of MIM led to activation of the PTK SRC
MIM functioned through SRC family kinases to regulate cell motility and invasion
To test whether SRC activity was essential for the enhanced cell invasion that accompanies MIM suppression, we tested the effect of the inhibitor SU6656  on formation of invasive structures in 3D culture. We observed that this SRC inhibitor attenuated cell invasion that was induced by suppression of MIM (Figure 7A). Furthermore, inhibition of SRC with SU6656 was sufficient to antagonize the effect of MIM suppression measured in quantitative cell invasion assays (Figure 7B). These observations highlight the importance of SRC as a mediator of the effects of MIM loss on cell invasion.
Invasion induced by suppression of MIM was blocked by SU6656, an inhibitor of SRC
SRC was a direct substrate of PTPδ
To test whether there was an enzyme–substrate relationship, we investigated whether PTPδ directly dephosphorylated SRC in vitro. We immunoprecipitated SRC from lysates of pervanadate-treated MCF10A cells that expressed shRNA directed against either PTPδ (to promote phosphorylation at Tyr527) or MIM (to promote phosphorylation at Tyr416). The immunoprecipitated SRC was then incubated with wild-type active PTPδ or the catalytically dead CS mutant, and the extent of dephosphorylation was measured using specific antibodies. Immunoblots revealed that PTPδ specifically dephosphorylated SRC at Tyr527 (Figure 8A).
Identification of SRC as a direct substrate of PTPδ
Model to explain the relationship between expression of MIM and PTPδ-induced activation of SRC
Earlier work from our laboratory revealed that mutation of Asp181 to alanine abolishes the catalytic activity of PTP1B, but maintains its binding affinity for substrates, thus producing a substrate-trapping mutant form of the enzyme . Using this strategy, we generated a substrate-trapping mutant form of PTPδ. Wild-type PTPδ and the substrate-trapping (DA) mutant were expressed in MCF10A cells, immunoprecipitated from cell lysates and probed with an anti-SRC antibody. A stable interaction between the trapping mutant and SRC was observed, but not between wild-type PTPδ and SRC (Figure 8B). To examine whether the interaction of SRC with the PTPδ substrate-trapping mutant occurred through the PTP active site, we included sodium orthovanadate. Vanadate is a competitive inhibitor and transition state analogue of phosphate that binds at the PTP active site and competes with substrate for binding to the enzyme . We observed that the interaction of SRC with PTPδ was inhibited by vanadate (Figure 8C). In summary, these data illustrate that the effects of suppressing MIM on cell migration and invasion are mediated by the direct dephosphorylation and activation of SRC by PTPδ.
Our data demonstrate that the increased levels of PTPδ in MIM-depleted cells provide a direct link to tyrosine-phosphorylation-dependent signalling and the switch to a more invasive state. On this basis, we propose a model that illustrates a mechanism by which MIM and PTPδ may regulate breast epithelial cell motility and invasion (Figure 9). In the absence of MIM, the levels of PTPδ protein were increased, which in turn promoted dephosphorylation of Tyr527, the inhibitory site in the C-terminal tail of SRC. This was associated with autophosphorylation at Tyr416 and activation of SRC. These data suggest that, under normal conditions, MIM functions to suppress the levels of PTPδ and thereby to suppress SRC activation, motility and invasion of breast epithelial cells.
MIM as a metastasis suppressor
Metastasis is a major cause of morbidity and mortality in developed countries. Mortality associated with breast cancer has declined considerably due to early diagnosis and improvements in current therapies. Nevertheless, disease recurrence and the transition to metastasis remain a challenge. MIM is widely expressed in human and mouse tissues, suggesting a potential to exert effects broadly in multiple cancer contexts. It was originally defined by the fact that its transcript was missing from metastatic cells, but not from the non-metastatic counterparts, with methylation of CpG islands thought to play an important role in silencing. Consequently, a biomarker such as MIM may prove to be a valuable tool for detection of metastasis. Furthermore, loss of MIM, and the activation of PTPδ/SRC-dependent signalling, may define a signature that would permit specific targeting of that population of metastatic cells. However, expression analyses have revealed situations in which MIM may be down-regulated more generally in tumour cells, rather than exclusively in metastasis . Nevertheless, the absence of MIM could still provide a defining signature that would facilitate specific targeting of such tumour cells.
PTPδ as a SRC phosphatase
SRC is a critical regulator of signalling pathways that affect cell migration, adhesion and invasion. We have demonstrated that the enhanced cell migration and cell invasion that is a consequence of suppression of MIM resulted from increased levels of PTPδ, which catalysed the dephosphorylation and activation of SRC. As such, PTPδ joins those members of the PTP family with the ability to regulate the activation status of SRC. PTP-mediated regulation of SRC is complex, due to the ability of these enzymes to control the phosphorylation status of both activating and inhibitory sites. Initially, expression of another receptor PTP, PTPα, was shown to transform rat embryo fibroblasts through dephosphorylation and activation of SRC . The prototypic receptor PTP CD45 was also shown to act positively to promote antigen receptor signalling through dephosphorylation and activation of SRC family PTKs (protein tyrosine kinases) [11,12]. Non-transmembrane PTPs, including PTP1B , have also been reported to function as activators of SRC by dephosphorylating the C-terminal residue Tyr527. In contrast, other phosphatases, such as PTPN23 (PTP non-receptor type 23)  and PTP-BAS , have been shown to dephosphorylate SRC at its autophosphorylation site, thus antagonizing SRC activity. Therefore the members of the PTP family offer a mechanism for fine control over the signalling function of SRC. Overexpression of SRC is observed in many cancers, including breast cancer ; however, overexpression or hyperactivation of SRC in transgenic mouse models is not sufficient to induce a higher grade of breast tumour and metastasis . In contrast, transgenic overexpression of SRC in a p21−/− background dramatically induces tumour growth and metastasis, suggesting that the combination with a second hit may augment the transforming activity of SRC . Considering the positive effect of PTPδ on SRC activity, loss of MIM may promote tumour growth and metastasis particularly in breast tumours in which SRC is overexpressed.
PTPδ as a tumour suppressor or oncogene
When one considers the prevalence of PTKs as oncoproteins, it was anticipated that PTPs, through their catalysis of the complementary dephosphorylation step, would serve as tumour suppressors. Now we know that several PTPs exert a tumour-suppressor function. In fact, detailed sequence analyses of members of the PTP family have revealed a wide variety of mutations in PTP genes in various cancers . The PTPRD gene, encoding PTPδ, is one such example [22,23]. In this case, in addition to mutations that would be predicted to inactivate catalytic function, the clustering of mutations in the portion of the gene encoding the extracellular segment of the protein highlights the potential importance of interactions with ligands in the regulation of activity. Nevertheless, it is also now apparent that PTPs have the capacity to function positively to promote signalling. Furthermore, aberrant up-regulation of PTP genes has been detected in multiple cancers, indicative of an oncogenic function [11,12]. An excellent example is the SH2 domain-containing PTP SHP2, which is encoded by the PTPN11 gene. SHP2, which normally facilitates RAS activation, exists in a low-activity state under basal conditions and is activated following interaction of its SH2 domains with phosphotyrosine residues on proteins that target the PTP to signalling complexes. Activating somatic mutations in PTPN11/SHP2 allow the enzyme to adopt the active conformation in the absence of the normal stimulus and have been associated with hyperactivation of MAPK (mitogen-activated protein kinase) and other signalling pathways and increased risk of sporadic childhood malignancies, such as juvenile myelomonocytic leukaemia and acute myeloid leukaemia . Another example is the PTPN1 gene, which is located at chromosome 20q13, a region that is frequently amplified in breast cancer and associated with poor prognosis. It has also been reported that PTP1B is overexpressed in breast tumours together with the oncoprotein tyrosine kinase HER2 (human epidermal growth factor receptor 2). Mice expressing activated alleles of HER2 in mammary glands develop multiple mammary tumours and frequent metastases to the lung; however, when such mice were crossed with Ptpn1-knockout mice, tumour development was delayed and the incidence of lung metastases was decreased [52,53]. Conversely, targeted overexpression of PTP1B alone was sufficient to drive mammary tumorigenesis in mice , illustrating that it can play a positive role in promoting signalling events associated with breast tumorigenesis. A similar tumour-promoting role was suggested in prostate cancer . Nevertheless, SHP2 functions as a tumour suppressor in cartilage  and PTP1B has also been shown to exhibit tumour suppressive effects, for example in lymphomagenesis in a p53-deficient background . Consequently, there is a precedent for members of the PTP family to function as a tumour suppressor, a tumour promoter or both, depending on the context.
To date, somatic mutations in the PTPRD gene have been reported in diverse tumours [22,24], and germline mutations of PTPRD have been found in metastatic Ewing sarcoma . Also, PTPδ has been reported to exhibit tumour-suppressor activity through inhibition of STAT3 (signal transducer and activator of transcription 3) activation . Nevertheless, there have been conflicting reports of both tumour-suppressor and tumour-promoter functions of PTPδ in neuroblastoma [24,28]. The present study is consistent with a tumour-promoting function of PTPδ through dephosphorylation and activation of SRC. Data from the Human Protein Atlas, which indicated elevated expression of PTPδ in tumour samples from patients, compared with normal breast tissue, are also consistent with such a positive role in the regulation of signalling in breast cancer cells. This raises the exciting possibility that such positively acting PTPs as PTPδ may prove to be important therapeutic targets for new ways to intervene in cancer.
Therapeutic implications and conclusions
SRC has been established as a critical regulator of multiple signalling pathways involved in cell proliferation, survival, angiogenesis and metastasis . Elevated levels and activity of SRC protein have been reported in several cancers, including breast cancer, and the extent of overexpression and hyperactivation correlates with metastatic potential, particularly in colon and breast cancer . In addition, SRC activation has been associated with increased signalling through growth factor receptors, GPCRs (G-protein-coupled receptors) or hormone receptors and influences multiple stages of tumour growth and progression . Consequently, multiple SRC inhibitors are currently being assessed as therapeutics . Although these approaches show potential, all PTK-directed therapies have encountered the problems of limited response and acquired resistance , and combinatorial approaches are being considered to try to overcome this. Combinatorial therapies involving SRC inhibitors and chemotherapy are in trials in various settings of metastatic cancer ; however, combination strategies that facilitate a more targeted intervention in particular signalling pathways may be of greatest benefit. In addition, it will be extremely important to identify those patient populations that would benefit most from SRC-directed therapies. Overall, the present study suggests that patients with tumour or metastatic lesions defined by loss of MIM may be one such population. Furthermore, our work suggests that targeting such metastases with combinations of inhibitors of SRC and PTPδ may be more effective in abrogating the signals that underlie aberrant cell invasion than targeting either the PTK or the PTP alone. Although several members of the PTP family have been validated as therapeutic targets, they remain underexploited in large part due to the challenge of developing active site-directed inhibitors with drug development potential. Nevertheless, recent studies have highlighted the potential importance of dimerization in the inactivation of receptor PTPs [64,65]. Consequently it may be possible to design agents that target the extracellular segment of RPTPs and antagonize RPTP function indirectly via regulation of dimerization. This raises the possibility of producing therapeutic agents that act via the extracellular segment of PTPδ, which may offer a new approach, together with inhibitors of SRC, to targeting specifically the MIM-depleted population of metastatic cells.
Fauzia Chaudhary performed the experiments, Fauzia Chaudhary, Robert Lucito and Nicholas Tonks designed the experiments, analysed and interpreted the data and wrote the paper. Nicholas Tonks directed the study.
We thank Dr Timothy Chan (Memorial Sloan Kettering Cancer Center, NY, U.S.A.) for providing us with a GST–PTPRD construct and Dr Todd Waldman (Georgetown University, Washington, DC, U.S.A.) for the pcdF1-PTPRD construct. We thank Dr Gaofeng Fan, Dr Michael Feigin and Dr Senthil Muthuswamy for helpful discussions. Also, we thank James Duffy from the CSHL Graphics Department for assistance with the Figures.
This research was supported by the National Institutes of Health (NIH) [grant number CA53840 (to N.K.T.)] and the Cold Spring Harbor Laboratory Cancer Centre Support Grant [grant number CA45508]. N.K.T. is also grateful for support from the following foundations: Joni Gladowsky Breast Cancer Foundation, The Don Monti Memorial Research Foundation, Hansen Memorial Foundation, West Islip Breast Cancer Coalition for Long Island, Glen Cove CARES, Find a Cure Today (FACT), Constance Silveri, Robertson Research Fund and the Masthead Cove Yacht Club Carol Marcincuk Fund.
epidermal growth factor
human epidermal growth factor receptor 2
leucocyte common antigen
Missing in Metastasis
protein tyrosine kinase
protein tyrosine phosphatase
quantitative real-time PCR
receptor protein tyrosine phosphatase
SH2 domain-containing PTP
Present address: Hofstra North Shore-LIJ School of Medicine, Hempstead, NY 11549, U.S.A.