Skip to main content
European Commission logo print header

Understanding and fighting metastasis via dissection of the Core Invasive Machinery

Final Report Summary - METAFIGHT (Understanding and fighting metastasis via dissection of the Core Invasive Machinery)

Cancer diseases account for 7.6 million deaths a year (World Health Organization (WHO) number for 2005). Metastases, the most fearsome aspect of cancer, are responsible for approximately 90 % of all cancer deaths. The study of the spread of cells from a primary cancer and the formation of new tumours in distant organs is the basis of modern cancer research. Understanding all steps behind this process and the dissection of its core invasive machinery is the prerequisite for a successful fight against cancer.

The aim of METAFIGHT was been to improve tremendously the understanding of the molecular processes behind metastasis. Leading research groups have collaborated over three years to find new candidate targets for blocking metastasis outgrowth of cancer cells. The project has been supported by the Seventh Framework Programme (FP7) of the European Commission (EC) with funds amounting to EUR 3 million and involved 10 organisations from European countries. METAFIGHT has concentrated on molecular functional studies to understand the dissemination and outgrowth of metastasis. The research groups involved have collaborated intensively to study all steps involved in the complex interactions between a cancer cell and its surroundings, including the process that leads to the formation of a new invasive tumour tissue. By focusing on the molecular mechanisms of metastasis, the METAFIGHT consortium has focused on the core invasive machinery contained within integrin-mediated attachment structures as a major player in regulating cell adhesion, migration and invasion in the early onset of tumour metastasis and in homing of metastatic cells to target organs. The goal was to develop new potential therapeutic molecules together with industrial partners.

Project context and objectives

Project context

Metastases, the spread of cells from a primary cancer and the formation of new tumours in distant organs, are the most fearsome aspect of cancer. Metastasis consists of a series of sequential steps, including the shedding of cells from a primary tumour into the circulation, arrest in a new organ, extravasation into the surrounding tissue and initiation and maintenance of growth. All steps involve complex interactions between the cancer cell and its surroundings. Tumour-cell migration, invasion and metastatic dissemination all depend on:

i) changes in cell adhesion, and on
ii) gain in migratory phenotype, which are strictly correlated events.

Cell adhesion, migration and invasion ultimately derive from a network of molecular interactions which originate from adhesive receptors such as integrins and involve for example, integrins and matrix components, integrins and adaptor proteins, adaptor proteins and actin, adaptor proteins and kinases, etc.

These interaction networks increase in complexity in a hierarchical fashion to produce recognisable systems of intermediate complexity (such as those controlling actin polymerisation, cytoskeletal tension, and adhesion complex turnover), which themselves are components of even more complex biological phenomena including tumour metastasis. An early event in invasion and metastasis is the conversion from the stationary to the migratory phenotype, implying that integrin-dependent adhesions are functionally down-regulated and strong cell-matrix adhesion are resolved, whereas transient and weak adhesions are a prerequisite for migration. Although it is essential for cell adhesion mechanisms to be disrupted for tumour cells to become mobile and breach the basement membrane, the reattachment of malignant cells to metastatic sites requires an increase in cellular adhesive capacity.

The METAFIGHT consortium has developed a comprehensive approach by focusing on the core invasive machinery contained within integrin-mediated ECM attachment structure. This includes a large and discretely localised intracellular signalling network, consisting of the focal adhesion kinase (FAK) as well as their proximal effectors such as small GTPases and adaptor proteins. This core platform has been shown to be activated by numerous oncogenic pathways, resulting in reduction of effectiveness of cytostatic-induced cell death and of radiotherapy. It remains unclear how this platform is finely tuned and coordinated as a whole to control adhesion, migration and invasion. This platform represents a potential therapeutic target by acting on either upstream activators or downstream regulators. In this platform, the core invasive machinery components have been taken into account by the consortium for their role in metastasis outgrowth.

The consortium has defined a series of strategic objectives for the METAFIGHT project.


The METAFIGHT project has aimed at making a step change in metastasis understanding by focusing on the core invasive machinery emanating from an adhesive / signalling platform including the beta1 integrin and the FAK.

The specific objectives are as follows.

1)To break down and identify harmful aspects of the core invasive machinery signalling in metastasis outgrowth by a systems biology analysis of the early events generated by this platform in disseminating tumour cells. This included analysing how changes in molecular architecture in metastatic cells occur, how this was perceived and connected into intracellular signals and how these signals have changed the migratory and invasive cell phenotype. Dissection of signalling has been carried out in highly metastatic cells where signalling events have been quantified and characterised in vitro in a standardised and comparable manner. Particular emphasis has been put on in vivo analysis of mouse models of spontaneous tumour and metastasis formation with genetic manipulations leading to over-expression or knocking-down of specific pathways involved in the core invasive machinery.

2) To evaluate the fate of single metastatic cells within the mouse by taking advantages of imaging non-invasive strategies. The METAFIGHT consortium has heavily exploited fluorescence imaging system, bioluminescence, enhanced tomography and ultrasound scan that have been recently developed for use in small animals. Imaging studies for metastasis have included single-cell imaging studies that have been shed new light on the biology of metastasis and the interactions of metastatic cells with their microenvironment early in the process of metastasis. The option to detect metastatic cell clusters and gross metastases within the whole animal by using imaging has facilitated the inclusion of metastasis endpoints in preclinical development of new drugs.

3) To uncover candidate targets that have been exploited to enhance the understanding of signalling and to develop new treatment strategies in a joint translational research activity. Different screening techniques have been applied, such as an RNAi-based systematic screen for kinases involved in controlling the migratory machinery, and a systematic proteomics-based approach to identify molecular mechanisms involved in cell motility and invasion. Quantisation of changes in protein expression and phosphorylation in metastatic cells have been achieved using the novel techniques of stable-isotope labelling by amino acids in cell culture (SILAC) or in mice (SILAM) combined with mass-spectrometric and bioinformatics technologies. Small molecules, peptides and proteins have been screened for their potential to alter migration and invasion and the related signalling pathways. This has been followed by development of gene targeted treatments using conventional drug development and RNAi technology.

4) To develop new strategies to inhibit detrimental and to promote beneficial signals for blocking metastasis outgrowth exploiting the knowledge on the key core invasive machinery. This strategy has included translational research on existing molecules, proof of principle studies with molecules already developed by the partners of the consortium and the development of new candidate molecules identified during the course of the integrated project. Candidate targets (genes or proteins) have been tested for their effects on invasion and metastasis in a series of model platforms of highly metastatic cancer cells via engineered orthotopic xenografts. Also, human cancer tissue banks have been screened for the expression of the best candidate targets.

5) To increase the knowledge of experts and the lay population regarding metastasis by collaborating with European societies and national programs.

Project results:

Description of the main results / foregrounds

The METAFIGHT project was based on functional collaborative studies to understand early mechanisms of dissemination and outgrowth of metastases of primary human cancers. A comprehensive approach to understand the onset of metastasis with in vivo and in vitro approaches has been used, through blocking and interfering with the pivotal molecules that in the core invasive machinery drive the cells to migrate and recognise the target organ microenvironment. The consortium developed new potential therapeutic molecules together with industrial partners. This included the development of small molecule based drugs and new strategies for the delivery of proteins.

The strategy covered:

- identification of relevant molecular targets combining a detailed systems biology analysis of the core invasive machinery to in vivo modelling of this biological complexity using animal models;
- evaluation of candidate targets in human breast cancer tumour samples;
- development of new potential therapeutic molecules with IND/SME partners.

The project has been divided in seven work packages (WPs) designed to ensure a high level of integration while allowing flexibility regarding the adaptation of the research activities to the global needs of the project:

- WP1: Quantitative biology-based identification of core invasive machinery components that drive dissemination of tumour cells;
- WP2: Establishment of the role of the core invasive machinery platform in metastasis onset in in vivo tumour models, using the Kindlin-1fl/fl//Ros-Cre-ERT2 derived mouse mammary epithelial cells (MMEC) transplantation model as a method to study the role of Kindlin-1 in mammary gland development;
- WP3: Establishment of the role of the core invasive machinery platform in metastasis by mammary transplantation;
- WP4: Functional and quantitative proteomics of the core invasive machinery platform;
- WP5: Testing of candidates for therapeutic interventions. Pattern of expression of candidate genes in human tumour specimens, their correlation with clinical parameters of patients and potential for prediction of metastatic potential;
- WP6: Dissemination, innovation and knowledge activities;
- WP7: Management activities.

WPs 1 to 4 have been the major knowledge building work packages. Knowledge created within these WPs has been integrated in the in vivo models of WPs 2 and 3 as much as possible for further knowledge generation. The novel protein-binding partners and post-translational modifications identified in WP4 have been functionally tested for their specific influence on the core invasive machinery using the framework of analyses developed in WP1. Candidate metastasis target genes identified in WP1 have been used for development of therapy in WP5, and candidate metastasis inhibitors developed in WP5 have been tested using the in vitro and in vivo models and tools developed in WPs 1, 2 and 3. WPs 6 and 7 have been integrated with all other WPs for dissemination of knowledge and project management.

WP1. Quantitative biology-based identification of core invasive machinery components that drive dissemination of tumour cells

From the start of the project, researchers have made significant progress towards the goals and aims. New methodology has been developed for the automated extraction from microscopy images of multi-parametric data first from fixed settings and for dynamic parameters from live cells. This methodology will now be used for the efficient characterisation of cell adhesion and migration, including of the core invasive machinery. Planned RNAi based screening efforts have made significant progress, with screenings being completed including knockdown of each kinase, phosphatase, and adhesion complex component. The development of open-source software specialised for rendering and tracking within confocal live cell three-dimensional (3D) images (BioimageXD) has progressed as planned. The BioimageXD image analysis programme has been successfully developed to get quantitative analyses about integrin clustering, lateral movement and internalisation. These properties in the programme will be used to study the role of signalling proteins in integrin-mediated cancer cell invasion and metastasis, as well as in the testing of putative inhibitor molecules.

The most significant progress is the development of new software tools and the successful siRNA-based screenings. This is in accordance with the plan and these new tools and hits will now be used for elucidation of biological functionality and mechanisms of the core invasive machinery in cell adhesion, migration and invasion.

WP2. Establishment of the role of the core invasive machinery platform in metastasis onset in in vivo tumour models


Researchers aimed at characterising the specific role of a number of proteins implicated in the core invasive machinery platform (GTP binding proteins of the Galpha family, E-cadherin, the Vav family of Rho/Rac GEFs, the AnxA family of phospholipid-binding proteins, the tyrosine kinase FAK, the serin/threonine kinase Pak4 and the cytoskeletal-related proteins p140Cap and Paxillin) in the growth and metastasis of specific tumour types (primarily breast and cancer breast metastasis to the lung, melanoma and melanoma metastasis to the lung). They designed a dual experimental approach based on the use of well-established cancer cell lines and transgenic / knockout / knockin animal models. They also monitored the metastatic process both under conventional analytic methods and through the implementation of non-invasive techniques (i.e. luciferase-based bioluminescence, intravital microscopy). New animal models to be used as tools have been developed to monitor, in real-time and in vivo, the growth of primary breast tumours and their metastatic dissemination.

The work done has enabled researchers to achieve most of the goals contemplated. They could:

(i) demonstrate the implication of the aforementioned proteins as either positive or negative regulators of the metastatic process;
(ii) identify post-translational modifications that modulate the activity of some of the proteins during the metastatic process;
(iii) unveil new signal transduction elements and gene signatures that mediate the effector functions of proteins in the metastatic process.

The main achievements are listed below:

1) Technical achievements:
All non-invasive techniques to visualise tumour growth and metastasis in vivo using syngenic and xenograft transplantation experiments have been developed, optimised and implemented during the project. The techniques have been used successfully to characterise the implication of components of the core invasive machinery platform in the growth and metastasis of breast cancer cells and melanomas. Research also generated new animal knockout (KO) (FAK;p53, double Vav2;Vav3, p140Cap KO mice) and transgenic mice (Pak4, p140Cap) to address the role of specific members of the core invasive machinery platform in breast cancer. Next to that, they have been successful at developing the MMTV-Her2 luciferase reporter mice to visualise the growth and metastasis of primary tumours in vivo.

2) Scientific achievements:
It is demonstrated that elements of the core invasive machinery platform play positive (AnxA1, Anx2, Vav2, Vav3, FAK, Pak4, Paxillin) and negative (constitutively active G13 mutant proteins, p140Cap) in metastasis. Researchers have also identified post-translational (phosphorylation) modifications, signalling elements, and gene signatures induced by those proteins that help the metastatic dissemination of tumour cells. The use of genetically modified mice has demonstrated the role of these invasive machinery platform components in the primary growth and metastasis of breast tumours.

Significant results

(a) Cell line models

Anx1 and Anx2, two members of the Anx family of phospholipids binding proteins, play critical roles in breast cancer metastasis. Researchers have found that the pro-metastatic role of AnxA1 is mediated by a crosstalk with the TGFbeta route. In the case of Anx2, they have demonstrated that its phosphorylation is critical for its pro-metastatic roles. Microarray experiments were conducted to identify the gene signature associated with the loss of function of these proteins. The cytoskeletal protein Paxillin is an important actor in breast tumour metastasis. It was demonstrated that its phosphorylation by the serine/threonine kinase JNK controls this process. Researchers have demonstrated that the role of Paxillin in metastasis is linked to the regulation of the expression of the EGF receptor at the plasma membrane. The ectopic expression of this receptor rescues the metastatic defects of Paxillin-deficient breast cancer cell lines.

Vav2 and Vav3, two gross domestic product (GDP) / GTP exchange factors (activators) for the Rho/Rac GTPase family, play synergistic roles in breast cancer metastasis at different levels, including growth of the primary tumour, dissemination to lymph nodes, intravasation, extravasation, and growth of tumour cells in the lung. Researchers have identified the main signal transduction pathways and gene signatures associated to this process. Some of the Vav2;Vav3-dependent genes have been investigated. These data indicated that these targets play overlapping but not identical roles in the metastatic process, thus giving a number of therapeutic avenues to block the metastatic process. The cytoskeleton- and integrin-related p140Cap protein plays negative roles in breast cancer metastasis. Cortactin was identified as one of the downstream elements of this route. The active versions of Galpha13, a small G protein coupled to membrane receptors, block melanoma metastasis. This activity has been linked to the activation of p190RhoGAP and the concomitant down-modulation of RhoA activity in melanoma cells. E-cadherin, a transmembrane protein involved in cell-to-cell contacts and epithelial-to-mesenchymal transitions, plays negative roles in melanoma metastasis. p120Ctn, a protein that binds physically to both E-cadherin and the F-actin cytoskeleton, has been identified as one of the players in this process.

(b) Animal models

- The use of the FAKFlox/Flox;p53Flox/Flox mice demonstrated that the loss of the tyrosine kinase FAK affects the genetic makeup of breast tumours. Gene signatures associated to FAK-deficient tumours have been identified using microarray experiments.
- Demonstration in vivo that Vav2 and Vav3, but not Vav1, plays a role in carcinogen-induced breast cancer in vivo. Additional work using genetically modified Vav family knockout mice has shown that the inhibition of these proteins may represent a valuable tool to inhibit the primary grown and metastatic dissemination of hepatocarcinoma, lung cancer, skin cancer and oral carcinomas.
- The analysis of transgenic mice overexpressing p140Cap confirmed that this adaptor protein plays an inhibitory role in the metastatic process.
- Successful generation of the MMTV-luciferase;Her2 reporter mice. These mice have been subsequently used to follow in vivo tumour dynamics when the routes of the serine / threonine kinase Pak4 and p140Cap are also genetically altered. These reporter mice are a major hallmark in this project, since they represent an excellent experimental tool for the investigation of the oncogenic and metastatic potential of other proteins putatively involved in tumourigenesis and metastasis.


Below are included deviations that have resulted in delays or changes in the original experimental plan. It is worth noting that these delays did not hamper the achievement of the expected goals.

1) Delay in the generation of MMTV-luciferase mice that impinged on subsequent delays in setting up the crosses with the other animal models. This problem was solved and, once available, the animals have been used to validate the system and, subsequently, to approach the experimental goals.
2) Due to the expression of Vav2 and Vav3 in Her2- breast tumours, modifications were made in the experimental plan initially contemplated in the project. Researchers decided to use Vav2-/-;Vav3-/-, Vav1-/-;Vav2-/-;Vav3-/-, and the respective control animals to analyse the role of these proteins in breast cancer tumourigenesis and metastasis. It was decided to use carcinogenesis-based methods to induce breast cancer to carry out the project. This alternative approach allows to fulfil the objectives before the end of the project since the researchers did not have to carry out additional crosses or background homogenisation procedures.
3) Given that the initially planned FAKFlox/Flox;NeuT mice have been used by another group in a publication, researchers decided to modify the protocol contemplated in these studies. They generated the FAKFlox/Flox;p53Flox/Flox animal model that has been characterised successfully in the project.
4) Due to the workload associated to the proteome characterisations in WP3, the work contemplated with kindlin1 knockout mice in this WP has not been carried out.

As a result of the work made, the basic scientific questions set up at the beginning of this project have been answered successfully during this period.

Concluding remarks

These data have allowed to:

(i) establish all those proteins as potential pharmaceutical targets to combat primary tumour growth and metastasis; (ii) identify a number of direct and indirect therapeutical avenues to inhibit the functions of those proteins; (iii) unveil new signalling elements and gene signatures that play critical roles in tumour growth and metastasis; (iv) identify gene signatures that will be valuable as diagnostic / prognostic tools for cancer patients;
(v) develop cell line and animal models to characterise the effect of new anti-cancer and anti-metastatic therapies in real-time and in vivo.

WP3. Establishment of the role of the core invasive machinery platform in metastasis onset by mammary transplantation


Researchers aimed at establishing models to study spontaneous breast cancer development and metastasis using genetic models and characterising of the role of a number of proteins implicated in the core invasive machinery platform in tumour development and metastasis. For efficient gene modulation, researchers developed a model involving the transplantation of primary MMECs in the cleared mammary fat pad of immunodeficient donor mice. The aim was to use MMECs derived from MMTV-NeuT mice that spontaneously form mammary gland tumours that can also metastase. By crossing these mice with MMTV-Luc mice, researchers would be able to follow metastasis formation using bioluminescence imaging. Application of lentiviral-based knock down approaches in these MMTV-NeuT;MMTV-Luc MMEC prior to transplantation should allow the analysis of the role of candidate metastasis genes in mammary gland tumour formation and metastasis in spontaneous tumour models.

The work done has enabled researchers to achieve some of the goals contemplated initially in this WP. They could demonstrate that the MMEC methodology is feasible to study the role of core invasive machinery components in mammary gland development. Due to considerable delay in the crossing of MMTV-NeuT and MMTV-Luc mice and undetectable luciferase activity levels with bioluminescence imaging in primary mammary gland tumours of these mice, the attention was shifted to tumour cell lines derived from MMTV-NeuT mice and MMTV-NeuT;MMTV-p140Cap mice.

Researchers implicated p130Cas and p140Cap as either positive or negative regulators, respectively, of the metastatic process.

1) Technical achievements:
All methods to perform the transplantation of MMECs in the cleared mammary fat pad of immunodeficient mice have been established. This involves the isolation of MMECs of donor mice followed by the conditional deletion of genes of interest using 4OH-tamoxifen for MMECs derived from either Rosa-Cre-ERT2;FAKfl/fl or Rosa-Cre-ERT2;Kindlin1fl/fl mice. The injection of cells in the cleared mammary fat pad of immuno-deficient acceptor mice followed by the reconstitution of the mammary gland and analysis of the gland reconstitution using different staining methods for both basal- and luminal-epithelial cell markers in combination with automated image analysis procedures. Researchers established the formation of mammary gland tumours out of MMTV-NeuT MMECs that were injected in the mammary fat pad.
2) Scientific achievements:
A direct role for the central core invasive machinery-signalling component FAK was established in the development of the mammary gland. Researchers also established a role for p130Cas in the signalling by which NeuT drives metastasis formation. This supports a role for p130Cas as an oncogene. In contrast, p140Cap acts as an onco-suppressor molecule since p140Cap over-expression inhibits the metastatic capacity of MMTV-NeuT-derived mammary gland tumour cells to the lung.


The role and mechanism of FAK-mediated signalling in mammary gland development and differentiation was studied. Researchers used a conditional FAK knockout mammary epithelial cell (MEC) transplantation model. Conditional Cre-mediated FAK deletion in primary cultured MECs isolated from Rosa-Cre-ERT2/FAKflox/flox donor mice caused loss of FAK in all mammary cells. Transplantation of FAK knockout MECs in a cleared mammary fat pad of immune deficient recipient mice resulted in development of new but dilated virgin ducts with a disrupted myo- and luminal epithelial cell multilayer, and aberrant ductal morphogenesis during pregnancy. In the absence of FAK, MECs spread poorly, showed enhanced ROCK-mediated cytoskeletal contractility and failed to respond to receptor-mediated cytoskeletal remodelling. FAK deficiency fully inhibited branching morphogenesis of mammary gland organoids in a ROCK-dependent manner. Altogether these data suggest a model in which FAK coordinates contractile forces in MECs to maintain the bilayered cellular organisation of myo- and luminal epithelial cells in ducts, thus allowing proper mammary gland development and function. These data have been published. The data demonstrate that the Rosa-Cre-ERT2 mouse models can be sued together with mammary gland reconstitution and obtain qualitative and quantitative information from these studies regarding mammary gland development.

The ErbB2 oncogene is often over-expressed in breast tumours and associated with poor clinical outcome. p130Cas represents a nodal scaffold protein regulating cell survival, migration and proliferation in normal and pathological cells. P130Cas is a promising anti-metastasis gene. The functional role of p130Cas in ErbB2-dependent breast tumourigenesis was assessed by its silencing in primary breast cancer cells derived from mouse mammary tumours over-expressing ErbB2 (N202-1A cells). It is demonstrated that p130Cas is necessary for ErbB2-dependent foci formation, anchorage- independent growth and in vivo growth of orthotopic N202-1A tumours. Intra-nipple injection of p130Cas-stabilised siRNAs in the mammary gland of Balbc-NeuT mice decreases the growth of spontaneous tumours. In human mammary cells MCF10A.B2 the concomitant activation of ErbB2 and p130Cas over-expression sustain and strengthen signalling, leading to Rac1 activation and MMP9 secretion, thus providing invasive properties. Consistently, p130Cas drives N202-1A cells in vivo lung metastases colonisation. These results demonstrate that p130Cas is an essential transducer in ErbB2 transformation and highlight its potential use as a novel therapeutic target in ErbB2 positive human breast cancers.

To study the role of selected onco-suppressor protein in MMEC NeuT cells, researchers decided to cross the NeuT mice that over-express the oncogene mutated at position 664 in the transmembrane domain and provide one of the most aggressive models of mammary carcinogenesis. P140Cap is a promising anti-metastasis candidate gene. The data reported here show that primary MECs derived from mammary gland tumours of p140Cap/NeuT mice display decreased in vitro migration and in vivo lung colonisation ability when compared to cells derived from NeuT mice. This is a proof of concept that over-expression of p140Cap behaves like an onco-suppressor in HER2+ cancer cells.


Originally the goal was to focus on FAK, integrin-beta1, kindlin1, p130Cas and Vav2 in the context of MMEC transplantation and oncogenic transformation. Given the early publication on FAK and integrin-beta1 in the context of mammary gland and breast cancer development and tumour progression, it was decided not to focus on these target genes. Instead only FAK was used as target gene to establish our models of MMEC transplantation. For FAK, researchers did not use the NeuT model to further investigate the role of FAK in mammary gland tumour formation mediated by this oncogene, instead they limited to the role of FAK in p53-null induced breast cancer in the context of WP2.

A major technical setback was related to the MMTV-NeuT; MMTV-Luciferase model that was central to this WP. This was related to the re-derivation and shipment of animals from UNITO to UL, animal experimental protocols that had to be accepted by the local animal welfare committee, and personnel matters due to maternity leave of two highly skilled technicians working on this project (a total of 10 months). After generation of the MMTV-NeuT;MMTV-Luciferase mice the luciferase activity was positive in the NeuT-derived mammary gland tumours in MMTV-NeuT//MMTV-Luciferase female mice, but not for all the tumours. IVIS-200 analysis of luciferase expression shows that it was possible to detect luminescent signals of tumours at week 12, before they become palpable. Subsequent analysis shows that some (2/10) palpable tumours do not express luciferase. Therefore, at this stage this was an inefficient model to follow metastasis formation to distant organs based on bioluminescence imaging for obtaining significant results for the final deadline of METAFIGHT project. Lentiviral-based knock down in MMEC derived from NeuT//Luc mice, would not allow the quantitative analysis of metastasis formation. Although researchers continue to characterise this model, they decided to fully focus their attention for validating candidate genes in the in vivo situation using the orthotopic metastasis models applied in WP2. They have used the mammary gland tumour cell lines derived from MMTV-NeuT mice to study the role of p130Cas and p140Cap in experimental metastasis models.


The work performed in WP3 indicates that the MMEC transplantation model is a suitable model to investigate the role of individual core invasive machinery components in the development of the mammary gland. Researchers established that the MMTV-NeuT MMEC can be used for transplantation and mammary gland tumour formation. They have established he role of FAK in mammary gland development and determined the role of both p130Cas and p140Cap as an oncogene or onco-suppressor, respectively, in NeuT-driven mammary gland tumour development and progression.

WP4. Functional and quantitative proteomics of the core invasive machinery platform


One of the major obstacles in modern cancer therapy is the treatment of metastasis. During cancer development, single cancer cells achieve the ability to detach from their parental tissue and invade into so far unaffected regions of the body, where their uncontrolled growth leads to disturbed organ function and finally death. Therefore analysis of the biological mechanisms behind these processes is of significant importance. Integrins constitute a group of transmembrane receptors which control cell migration, cell-cell interaction and differentiation as well as developmental processes. As their role in cancer and metastasis formation is of unquestioned importance we sought to investigate the roles of Integrin-1 and associated interaction partners under normal conditions and in cancer.

As a main tool, researchers established SILAM and SILAC, techniques making use of stable isotope labelling of either life mice, or cell cultured cells. This allowed to make comparisons between different mouse or cell lines using a mass spectrometer, obtaining high quality and significance results. They succeeded in completely labelling mice with a diet containing the 13C6-substituted version of lysine. Mice were labelled over more than 10 generations with the heavy diet. The diet neither affected development, nor growth, nor behaviour. SILAC analysis from various organs lacking expression of b1 integrin, b-Parvin, or the integrin tail-binding proteins Kindlin-1 and Kindlin-3 confirmed their absence and disclosed a structural defect of the red blood cell membrane skeleton in Kindlin-3-deficient erythrocytes. In the course of the project, researchers generated mouse strain with a constitutive and conditional null mutation of the KIND1 gene. KIND1 is a FERM domain-containing adaptor protein that binds integrins and is in FAs. Mutation in the KIND1 gene of man cause Kindler syndrome (KS), characterised by transient skin blistering followed by abnormal pigmentation, skin atrophy and skin cancer, is caused by mutations in the KIND1 gene. The constitutive KIND1 deletion in mice leads to a severe and perinatal lethal colitis ulcerosa (UC) type defect. The Kindlin-1 deficient mice represent the first animal model for UC caused by a single non-embryonic lethal gene mutation. The SILAM analysis of the skin surprisingly revealed a contribution of the immuno-system to the phenotype. Currently, researchers analyse the influence of Kindlin-1 deletion to cancer development.

In human and mouse skin, in addition to Kindlin-1, a closely related paralogue, Kindlin-2, is expressed. In order to define functions specific for each Kindlin, researchers isolated keratinocytes derived from Kindlin-1 fl/fl background. After cloning and characterisation of these cells, they established sub-clonal cell lines stably expressing Kindlin-1-GFP or Kindlin-2-GFP fusion proteins or GFP alone. After continuous culture in SILAC-labelled cell culture medium, GFP-IPs were performed under mild conditions. Immunoprecipitates of different SILAC labelling were mixed and underwent mass-spec-analysis to identify potential complex partners of both kindlins. Data analysis revealed a list of potential interaction partners. Researchers could identify Integrin-linked kinase (ILK) and Talin as known complex partners, as well as many other so far unknown potential complex partners, indicating a potential influence of kindlins to metastasis through their interaction partners. The crosstalk of integrin-mediated signalling with growth factor receptors is of high significance for many processes, including metastasis. ILK mutant cell lines display a defect in this crosstalk. Researchers used SILAC-based proteomics to determine the phospho-proteome of ILK floxed ad ILK-null cells before and after EGF treatment, and analysed the differentially phosphorylated proteins. They chose one candidate protein that showed diminished tyrosine phosphorylation upon EGF treatment in ILK-null cells. This protein is novel and has not been studied so far and is named Babak. GFP-tagging and transfection into normal fibroblasts revealed FA localisation and an increase in F-actin stress fibre formation. Researchers have made a targeting construct and will analyse the protein in vivo and in mutant cells derived from Babak mutant mice.

The basic mechanisms of cell motility was analysed through the comparison of the phospho-proteome of cells with different migratory capabilities. MCF7 cells were studied and compared to their derivatives, which express the full-length EGFR (MCF7-EGFR), or those that express a truncated form of the EGFR (MCF7-deltaEGFR). Researchers also compared MCF7 and MCF7-EGFR cells in their basal state to the cells after a short stimulation with EGF. In order to compare these five states, the SILAC approach was employed. Using this approach, we quantified > 6000 phosphorylation sites. Among these phosphorylation sites, phosphorylation on proteins involved in EGFR signalling were found, as well as cell adhesion and actin cytoskeleton dynamics, which are known to be involved in cell motility. Bioinformatic analysis and biological experiments were done to determine the correlation between the cellular behaviour and the phosphorylation patterns. Researchers isolated tumourigenic and non-tumourigenic skin samples from FVB/N mice and SILAM skin. They succeeded to enrich these samples for phosphopeptide by establishing a new protocol combining FASP and TiO2 columns. This approach allowed us to identify a huge number of proteins and phosphosites during phospho-proteomic analysis. They adapted this method for the investigation of signalling and proteomic changes in carcinoma development and were able to identify a large number of differentially phosphorylated and expressed proteins and chosen candidates were validated with IF in mouse tissue. Marked changes especially in the malignant tumours could be observed in adhesion proteins related to actin.

WP5. Testing of candidates for therapeutic interventions

The objectives were:

A. pattern of expression of candidate genes in human tumour specimens, their correlation with clinical parameters of patients, and potential for prediction of metastatic potential;
B. identification of inhibitory molecules targeting melanoma and breast cancer cell invasion;
C. production of a panel of AAV virions expressing functional suppressor of metastasis to set up delivery tools for target organs.


Objective (A)

1. First-generation tissue macroarrays were generated from 151 breast cancer samples following established criteria. Samples include early stage and metastatic specimen and samples from patients treated with chemotherapy and Herceptin. All samples were characterised for expression of several markers of breast cancer progression.
2. CIC found that Vav2 and Vav3 proteins were found to be overexpressed in breast cancer cells.
3. UNITO analysed the expression of p140Cap in 55 human breast cancer specimens and on normal breast tissue. In normal human breast, p140Cap is expressed selectively in luminal cells of alveoli, but not in ductal epithelial or myoepithelial cells. Analysis of breast cancer tissues showed that p140Cap is expressed in almost a third of tumours. Using several parameters and markers, researchers found that highly aggressive tumours lose p140Cap expression. None of the EGFR positive tumours express p140Cap, suggesting mutually exclusive correlation between EGFR and p140Cap expression. The data show an inverse correlation between p140Cap expression and the state of malignancy.
4. RIAM is a talin-binding protein, and therefore a good candidate to function in the Core invasive machinery in tumour cells. CSIC has identified RIAM as playing important roles in melanoma cell growth and metastasis. RIAM antibodies detected the expression of RIAM in human melanoma metastatic tissue (regional lymph nodes), and in surrounding stroma. Anti-RIAM antibodies were sent last July to Centro de Investigacion del Cancer tumour bank for screening of melanoma samples for correlation with clinical parameters and for prediction of metastatic potential. Following criteria to choose selected samples based on stage of metastasis and chemotherapy treatment the analysis is almost completed.
5. Blk is a Src kinase protein that CSIC found it tightly controls melanoma metastasis by regulating the levels of active RhoA. CSIC has generated anti-Blk antibodies and conducted initial immunohistochemistry on metastatic melanoma tissue. Blk is clearly expressed on melanoma and on infiltrating leukocytes. The antibody will be used for screenings at the CIC's tumour bank.

Objective (B)

1. KI tested the PF-228 FAK inhibitor from Actar in migration of breast carcinoma cells under two-dimensional (2D) conditions and found no effect on migration onto different extracellular matrix proteins. UL also tested this inhibitor and found no inhibition of FAK phosphorylation.
2. CSIC has tested 9440 marine crude extracts provided by Pharmamar S.A. in adhesion and invasion assays. After chemical fractionation and purification, one compound, PM080528, structurally related to the actin modulator hectochlorin, was isolated and tested in different in vitro assays to determine its biological activity profile. The compound modulates the cell cytoskeleton leading to a rapid and drastic reorganisation of the actin network. Another pure compound structurally related to hectochlorin, PM100099, also substantially inhibited adhesion and invasion. To assure the supply of the PM080528 compound, Pharmamar has defined a chemical synthetic route and has begun the synthetic process. The company has synthetised more than 100 substructures (building blocks and chemical intermediates) that will lead to the pure compound PM080528, as well as several other chemical analogues thereof. Their potency and cytotoxicity in in vivo models of metastasis will be next assessed.
3. Actar and KI are focusing on 10 different compounds which have been tested in a breast cancer in vivo model. One compound achieved a 30 % inhibition of tumour growth without toxic effects. The active compound belongs to a scaffold representing several in vitro active compounds with favourable ADMET properties. Actar is evaluating the tumours from the xenograft study.
4. Following development by University of Turku of a software analysis tool (BioImage XD) to test the effect of selected inhibitory molecules on several adhesion parameters such as integrin clustering and internalisation, University of Turku tested two Pharmamar pure compounds as potential inhibitors of integrin clustering and internalisation. The tested compounds showed no significant effects.
5. CIC tested commercially available inhibitors to test their effect on Vav-dependent responses, focusing on the ZF domain. The inhibitors did not block Vav-dependent cell transformation so far. Due to the negative data, in silico studies have been performed, focusing in the catalytic core of Vav proteins. These studies identified a set of potential drugs that will be tested in the future.

Objective (C)

UNITO set up an AAV vector model coding for luciferase and the GFP expression that was inoculated into WT and Neut Balb-c mice. The mammary gland and adjacent tissues were analysed using the IVIS200 Xenogen apparatus. Two constructs were used that allowed expression of luciferase in the MECs but also outside the glands in the muscle underneath the mammary glands. Likely, the CMV drives expression of luciferase that thus become detectable in muscle cells. FACS analysis testing GFP expression confirmed that 20-30 % of MECs express GFP. Therefore, intranipple injection of AAV particles allows expression of exogenous proteins into MECs.

Ethical issues

All research in the METAFIGHT project has been carried out in strict accordance with national and international legislations. The research project does not raise unusual ethical problems. The studies have been in accordance with the ethical principles that have their origin in the Declaration of Helsinki, adopted by the 18th World Medical Assembly Helsinki, Finland, June 1964, subsequent amendments and Note of Clarification on paragraph 29 of the WMA Declaration of Helsinki. Studies have been designed to comply also with the ICH Topic E6 Guideline for Good Clinical Practice, as implemented in Europe (CPMP guideline CPMP/ICH/135/95) and incorporated into respective national laws accordingly. An ethical review board in the management structure of this collaborative project was implemented, in accordance with the recommendations of the WHO and the ICH Topic E6 Guideline for Good Clinical Practice.

The use of human tissue bank was approved and monitored by local ethical review boards of the involved institutions. Special care was taken to ensure safe storage of personal data for a tissue bank, For the use of the human tumour bank, METAFIGHT has strictly followed the ethical rules established in Salamanca, Torino and Stockholm respectively, in particular in terms of anonymity and full confidentiality. Experiments relating to animal care have been subjected to prior evaluation and approval by local and national ethical committees, in accordance with the regulations of the member countries in which the work is being conducted. To further address the ethical issues, the METAFIGHT consortium have appointed two ethical advisers being specialists in the ethical aspects of animal experimentation from CSI and UL.

Potential impact

Due to its innovative research strategy and the exceptional composition of its consortium members, this project had a high impact:

1. scientific excellence and competitiveness;
2. identification and exploitation of new molecules involved in metastasis formation;
3. the creation of a consortium of excellence that has improved the quality of the state of the art and has attracted young scientists to participate in this field.

The activities in this project have shed further light on the early basic mechanisms underlying metastasis onset and have given proof of concept in translation into new therapeutic molecules. The concept of this collaborative project was based on the hypothesis that metastatic cells use basic adhesive machinery for escaping from primary tumours, surviving into the blood and homing in final target organs. The molecular mechanisms underlying these events are still not completely understood. The METAFIGHT approach has consisted in a new novel global strategy to identify the candidate molecules that intervene during the process of metastasis formation. The project started from the comprehensive analysis of the components of the core adhesive machinery regulating cell invasion and metastasis formation by in vitro and in vivo approaches and lead to identification of novel therapeutic molecules that specifically prevented the onset and the establishment of the metastatic phenotype. In the context of therapeutic interventions, knocking-down or interfering with the core invasive machinery resulted in acting with more upstream interventions. This may more selectively prevent and address the harmful metastatic process. This approach would be particularly relevant due to the well-recognised importance of adhesive signalling to confer higher resistance to ionising radiation and cytotoxic drug in metastasising cells.

The METAFIGHT consortium has generated internationally recognised scientific excellence and reinforcement of competitiveness. This has resulted from the strategy to bring together most of the experienced and well-known European research groups with the greatest expertise in adhesive signalling analysis and in genetic analysis of the adhesive platform. The groups have extended their established knowledge to proteomic and imaging techniques and in screening approaches under in vitro as well as in vivo conditions. The interaction of all groups resulted in WP1 in global quantitative analysis of the invasive platform in live cells in 2D or 3D structures. Xenograft and animal models presented in WP2 and WP3 and the proteomic approach in WP4 reflect the interaction of the groups of the consortium in understanding the role of the core invasive machinery in in vivo models. These three WPs have allowed identification of new therapeutic targets to potentially prevent and treat metastasis in WP5. The strategy of WP1-4 has established a platform for highly efficient, high throughput screening for new molecules as potential metastasis inhibitor drugs and in animal models relevant to metastasis. Moreover, in WP5 the models set in vitro and in vivo on the Core invasive machinery has been exploited by the pharmaceutical company Pharmamar because they represent unique models for following metastasis onset.

The close interaction in WP1-4 between dissection and screening have served the strategic goal to develop new research strategies in WP5, which has exploited the scientific achievements. Molecules identified by METAFIGHT to be functionally involved in controlling tumour dissemination constitute potential therapeutic targets. All strategies have been based on the concept that metastasis can be prevented by modulating the core invasive machinery platform. To confirm the usefulness of the molecules identified in the previous WPs, studies were conducted to analyse their expression in human cancer samples using tissue tumour banks. Due to the link between the consortium and the Spanish network of tumour banks together with banks in Torino and Stockholm, the METAFIGHT consortium had a significantly larger pool of samples at its disposal.

WP5 has been a unique interaction of one IND, an small and medium-sized enterprise (SME) and excellent research groups. Exploitation has occurred at different levels. IND Pharmamar has mainly been involved in identifying new molecules from marine extracts that interfere with metastasis in our in vitro and in vivo models. In addition, SME Actar which is specialised in high throughput screening for drug development and has a strong expertise in organic chemistry, medicinal chemistry and preclinical drug development has set up in vitro screening assays where the activity of candidate molecules (e.g. kinase or other regulatory properties) is quantified. A number of members of the consortium have key positions in national and international research networks. Therefore, significant account is taken of other national or international research activities.

The METAFIGHT consortium has contributed to International standard. The Consortium has established internal standards for the generation, breeding, delivery and phenotyping (including extensive analysis of signalling) of mice. Interventions such as xenograft and transplantation have been standardised. These standards have been established as experimental protocols and have been published on the website of the consortium. Numerous alternative protocols exist for a number of the techniques mentioned above and it is not possible to set mandatory national or international standards as researchers are free to decide which method to prefer. However, researchers have taken great care in detailing the experimental protocols to be easy to follow and instructive, including an explanation why we felt a particular protocol would be suited best compared to others, so that researchers would find it convenient to adhere to protocols.


A plan for information dissemination activities for the METAFIGHT project was established at month 8 of the project. It included the set-up of a public website, the organisation and participation in scientific conferences and open door days, the creation of a project flyer and the edition of press releases, articles and publications about METAFIGHT and its results. The objective of the METAFIGHT project was to perform functional studies to understand the dissemination and outgrowth of metastasis through systematic analysis of the core invasive machinery contained within integrin-mediated ECM attachment structure. The aim of the dissemination actions was to reach the target audience and to transfer correct and incisive information about the project activities and achievements. It was the objective of the consortium to communicate around the collaborative actions made feasible thanks to the support of the European Commission (EC).

The METAFIGHT consortium has designed a plan for disseminating knowledge with the objectives:

1. to raise public participation and awareness of the progress made in metastasis research;
2. to enhance exchanges with scientific world;
3. to prepare exploitation of results, create market opportunities;
4. to spread the knowledge gained beyond the consortium.

The dissemination plan is organised in four large parts by target group:

1) dissemination towards the public;
2) dissemination towards the scientific world;
3) dissemination towards the LifeSciHealth Programme.

For each of the target groups the aim, content, target, main message, detailed activities and timing of activities are described in the final summary report.

Exploitation results
To deal with the questions related to dissemination of results, exploitation and intellectual property rights (IPR), a specific work package (WP6) has been set up, focusing on the continuous identification and protection of knowledge generated by the METAFIGHT partners. The WP6 leader, in collaboration with the Executive Board, has been supported partners in their respective valorisation procedures, making sure that the access of METAFIGHT results etc. are dealt with on a fair and viable basis for all. This task has included patent searches, filing of patent (or other intellectual property rights (IPR)) applications, etc.

The results of METAFIGHT are expected to mark an important societal impact thanks to the development of better diagnostic and clinical strategies and to have a potential but significant economic impact, in particular for the IND and the associated SME involved in METAFIGHT project. In the present context, translation and exploitation are referred mainly to the use of the results that have been generated by the project research activities, in particular to their conversion into clinically useful tools for the management of cancer patients and their bringing onto the market for commercial exploitation.

The research activities undertaken under the umbrella of the project have generated results with various commercial potentials. This has raised the issues of IPRs, of protection of the property rights (patenting) and of undertaking to bring a product onto the market using an appropriate commercial vehicle (licensing). The METAFIGHT consortium has been composed of research centres, universities, IND and SME active in the field of cancer metastasis and drug development research and development. They exploited the results of the METAFIGHT project in various ways. The research centres and universities have been mostly benefited from the advance in knowledge which have strengthen their position as leading research institutions in Europe and brought new opportunities for future partnerships. Concrete plans for exploitation covered mainly publications in peer-reviewed international journals and filing of patent applications. The general principles for IPR ownership and IPR protection have been established in the consortium agreement.

The METAFIGHT results are used in two main ways by individual partners:

1. continuous research
2. exploitation of patents.

Publishable exploitable result No. 1: Regulation of expression of p140Cap protein in tumours.
The result was the present invention relates to the area of p140Cap protein. The present invention relates to the expression of human p140Cap in tumours and its regulation.
Partners involved: Unito
Possible market applications: Utilisation of these reagents will be useful for the diagnosis of cancer. The present invention is directed to the use in the tumour therapy of p140Cap protein and/or fragments thereof, which was surprisingly found able to suppress tumourigenic properties of cancer cells and in particular breast cancer cells. In one embodiment, the present invention involves the use of p140Cap protein and/or fragments thereof for the preparation of a medicament for the treatment of tumours and metastases in a patient suffering from a tumour. There is established proof of concept in animal tumour models and in human tumours. IPR: European patent application No. 06 795 477.6 of 30 March 2009 Regulation of p140Cap protein expression in tumours (No. PCT/IB2006/002514 of 8 September 2006).

Publishable exploitable result No. 2: Software for quantitative bioimaging (BioImageXD)
New software tools for quantitative analysis of cell membrane proteins and their internalisation. These can be used to study the components of cell adhesion apparatus and to screen potential inhibitor molecules.
Partners involved: UTURKU
A possible market application is open-source code software. This is freely available from the web site (GNU-GPL-licence required) and can be further targeted for specific purposes, e.g. high throughput screening of molecules with putative anti-adhesion effects. The basic software package is ready and available for scientific community. Special tools and applications (with commercial potential) are being tested and under further development.
IPR: Open-source code software. GNU-GPL-licence required for use.

Publishable exploitable result No. 3: Inhibitors of p21 activated kinase 4 (PAK4) for treatment of cancer progression and metastasis
The present invention is the discovery of novel inhibitors of the p21 activated kinase 4 (PAK4), with demonstrated efficacy in several cell based assays and in vivo.
Partners involved: Actar AB and KI
Possible market applications: In the year of 2007, about 13 % of global deaths (7,6 million people) were caused by malignant tumours and cancer rates are expected to increase dramatically. Systemic metastasis is the primary cause of mortality for most cancer types and represents a major therapeutic challenge in oncology. As a matter of fact approximately 90 % of all cancer deaths arise from the metastatic spread of primary tumours. When cancer is detected at an early stage, before it has spread, it can often be treated successfully by surgery or local irradiation and the patient will be cured. When cancer is detected after it is known to have metastasised, treatments are much less successful. For many patients, in whom there is no evidence of metastasis at the time of their initial diagnosis, metastases will be detected at a later time, years or even decades after apparently successful primary treatment. The cancer market is currently valued to around USD 42 billion and is the fastest growing segment in the Pharma industry. There is a strong trend towards targeted therapies and their total sales now have overtaken cytotoxic therapies. However, cytotoxic drugs are expected to remain a cornerstone of cancer therapy. Nevertheless, there is a medical need for more efficacious and better-tolerated drugs that in addition to killing tumour cells also blocks metastases. Patent application is drafted but no patent has been filed.

Publishable exploitable result No. 4
The siRNA screening has identified different candidate genes that have an impact on tumour cell migration. These genes, including different kinases, could be used for drug development in the pharmaceutical industry as well as diagnostic markers in the clinic. The potential impact is the specific pharmaceuticals development that inhibits cancer cell dissemination. Projects in close collaboration with larger pharmaceutical industries have been submitted. IPR are under investigation.

METAFIGHT partners and contact:

1) Paola Defilippi
Professor of Cell Biology
Department of Genetics, Biology and Biochemistry
Molecular Biotechnology Centre,
University of Torino
Via Nizza 52
10126 Torino
Phone: +39-011-6706434
METAFIGHT website:


2) Jyrki Heino, M.D. Ph.D.
Professor in Biochemistry
Department of Biochemistry
University of Turku
FI-20014 University of Turku

3) Xose R. Bustelo, PhD
Full Professor, CSIC
Centro de Investigacion del Cancer-Cancer Research Center
University of Salamanca Campus Unamuno s/n
E-37007 Salamanca
Phone: +34-923-294802

4) Staffan Stromblad, PhD
Professor in Clinical Molecular Biology
Karolinska Institutet
Department of Biosciences and Nutrition
SE-141 57 Huddinge
Phone: +46-860-89266

5) Joaquin Teixido, PhD
Centro de Investigaciones Biologicas
Cellular and Molecular Medicine Program
Ramiro de Maeztu 9
28040 Madrid
Phone: +34-918-373112

6) Bruno Cucinelli
58A rue du Dessous des Berges
75013 Paris
Phone : +33-153-945461
E-mail :

7) Bob van de Water, Prof. Dr
Division of Toxicology
Leiden/Amsterdam Center for Drug Research
Leiden University
Einsteinweg 55
2333 CC Leiden
The Netherlands
Phone: +31-715-276223

8) Bjorn Kull, Ph.D
Actar AB
Nobels vat 3
SE-171 77 Solna
Phone: +46-852-484801

9) Reinhard Fassler, Prof. Dr
Department of Molecular Medicine
Max Planck Gesellschaft - MPG
Max Planck Institute of Biochemistry
Am Klopferspitz 18
D-82152 Martinsried
Phone: +49-898-5782424

Matthias Mann, Prof. Dr
Max Planck Institute for Biochemistry
Department of Proteomics and Signaltransduction
Am Klopferspitz 18
D-82152 Martinsried
Phone: +49-898-5782557

10) Luis F. Garcia-Fernandez, PhD
Cell Biology and MoA Department, Head
PharmaMar, S.A. (Grupo Zeltia)
Avda. de los Reyes, 1 (PI - La Mina Norte)
E-28770 Colmenar Viejo (Madrid)
Phone: +34-918-234615