The Scribble family in cancer: twentieth anniversary.
Santoni MJ, Kashyap R, Camoin L, Borg JP
More than 90% of adult cancers arise from the malignant transformation of epithelial cells that normally shape organs such as breast and colon. Molecular events that alter the normal function of specific human genes, the so-called oncogenes and tumors suppressor genes, lead to impaired morphology and behavior of epithelial tissues that become tumoral and acquire novel functions such as the capacity to invade other tissues. Dissemination of cancer cells in the human body is a deadly event hardly tackled by the current therapeutic strategies. More research is therefore needed to explore new avenues able to prevent cancer dissemination.
The tumoral phenotype often includes alteration of cell polarity of epithelial tissues, increased cell migration and resistance to apoptosis of tumor cells and, ultimately, formation of metastasis. Cell polarity is a fundamental process initiated during embryonic development and maintained throughout adult life. Epithelial tissues are formed according to two axes of polarity: Apico-Basal Polarity (ABP) building polarized uni- or multi-cellular layers, and Planar Cell Polarity (PCP) which is perpendicular to ABP and required for tissue morphogenesis. The genetic basis of PCP has been historically elucidated in Drosophila which has allowed the characterization of many evolutionarily conserved genes that encode membrane receptors (among others PTK7, VANGL1 and VANGL2), WNT ligands, cytoplasmic adaptors (among others SCRIBBLE and PRICKLE1) and protein kinases (MINK1) belonging to a non-canonical WNT signaling pathway. However, the current understanding of the mode of action of WNT/PCP proteins is so far superficial. As WNT/PCP signaling plays a pivotal role in cancer dissemination and aggressiveness, its components and associated pathways could become very important therapeutic targets in the future. For many years, our team has focused on the study of the WNT/PCP genes SCRIBBLE (and its homologs ERBIN and LANO), PTK7 and VANGL2 as well as on the tumor suppressor LKB1 whose implication in AMPK and WNT signaling is well documented. Our team has significantly contributed to the identification of molecular networks associated to these molecules in physiological and tumoral situations, and have demonstrated the implication of VANGL2, PRICKLE1, PTK7 and ERBIN in various cancers. Our aim is to pursue our fundamental studies on the mode of action of these proteins using cell biology, proteomic and genetic methods and to assess their involvement in cancer using our mouse models and clinical resources/expertise of Paoli-Calmettes Institute.
Our team has characterized a family of cytoplasmic scaffold proteins (LAP for Leucine-Rich Repeats and PDZ domains) implicated in epithelial cell polarity that comprises ERBIN, SCRIBBLE and LANO (reviewed in Oncogene 2020). ERBIN is a protein with LRR (Leucine Rich Repeat) and PDZ domains (PSD95-DLG-ZO1) interacting by its PDZ domain with the carboxy-terminal end of the oncogenic ERBB2/HER2 tyrosine kinase receptor (RTK). We have demonstrated the role of ERBIN in the oncogenicity of ERBB2/HER2 tyrosine kinase receptor in breast (PNAS 2014), skin (in coll. J. Cell Biol. 2015), colon (in coll. Cancer Research 2018), its involvement in severe allergic dermatitis (in coll. J. Allergy Clin. Immunol. 2018) and in cardiac (in coll. PNAS 2014) and nerve (in coll. Nature Neurosc. 2013) functions. The ERBIN-ERBB2/HER2 interaction stabilizes the receptor by increasing its dimerization at the plasma membrane and is required for cell proliferation and oncogenicity in vitro and in vivo (PNAS 2014).
We have contributed to show the role of SCRIBBLE in neuronal plasticity (in coll. Cerebral Cortex 2017) and identified VANGL2, a central actor of the WNT/PCP pathway, as an interactor of SCRIBBLE thanks to the human PDZome (Mol. Cell. Proteomics 2013, in coll. Nat. Methods, 2015). We identified SCRIBBLE as a negative regulator of WNT/b-catenin pathway (Proteomics, 2019).
Another important aim is to study the genetic interactions between the three LAP family members (erbin, scribble and lano) in the mouse. Scribble mutant mice (scrib-/-) exhibit a lethal perinatal neural tube defect (craniorachischisis) whereas erbin-/- and lano-/- mice are viable and fertile.
Most cancers including breast cancers contain a population of self-renewing cancer stem cells (CSC) that drives tumor growth, resists to conventional therapies and initiates metastasis development. We recently identified LANO as a negative regulator for the expansion of normal and cancerous breast stem cells in part by suppressing the secretion of WNT ligands (Stem Cell Reports, 2018). We are currently investigating the molecular mechanism behind this new function of LANO by biochemical and cell biology methods. In addition, CSC chemo- and radio-resistance are due, at least in part, to their residing in micro-environmental niches poorly reached by cytotoxic treatments and considered as the main cause of relapse in association with their involvement in metastasis dissemination. However, the role of mechanical forces in regulating CSC fate has been poorly investigated. A collaborative and interdisciplinary project Capostromex funded by A*MIDEX will define the relationships between stroma mechanical viscoelastic properties on CSC behavior in tumorigenesis and contribution of Scrib and Lano, both associated to CSC pool maintenance.
Scribble genetically interacts in the mouse with vangl2, a gene encoding for a WNT/PCP receptor. We undertook a proteomic strategy to identify protein networks associated to these WNT/PCP components and investigated their contribution in cancer. We found that VANGL2 is overexpressed in basal breast cancer with poor prognosis and co-purifies with the adaptor p62/SQSTM1 and JNK (Nat. Comm. 2016 + patent INSERM) as well as with a membrane trafficking regulator (in coll., Development 2012, Nat. Cell. Biol. 2013). However, the mode of action of VANGL2 is still largely unclear and continues to be explored in the team.
We also carried a comprehensive analysis of the PDZ-binding profile of VANGL2 using a set of cellular, biochemical and biophysical approaches (yeast-two hybrid, holdup assay, SPR, fluorescence polarization). This enabled us to confirmed previous PDZ Vangl2 partners and identify new potential binders currently under investigation. We could also document the pivotal contribution of some residues within the PDZ-binding motif of VANGL2 (Monserrat-Gomez et al., in preparation) and the possible role of phosphorylation events occurring in this VANGL2 region. We recently identified and characterized a new Vangl2 isoform unveiling an unexpected level of complexity in VANGL2 expression, trafficking and function.
PRICKLE1, an adaptor for VANGL2, was also found overexpressed in basal breast cancer and associated to a novel RICTOR-MINK1 complex driving AKT activation and cell dissemination in vitro and in vivo (Dev. Cell 2016, Cell Cycle 2016 + patent INSERM) and ECT2, a Rho-GEF, in basal breast cancer (Br. J. Cancer, 2019). Recently, we found a novel protein complex associated with the prometastatic PRICKLE-MINK1 complex in breast cancer.
Scribble genetically interacts in the mouse with ptk7, a gene encoding for a WNT/PCP receptor. PTK7 is a tyrosine kinase receptor (RTK) overexpressed in poor prognosis acute myeloid leukemia and potentially involved in the retention of leukemia cells in the bone marrow (Blood 2010). Using a ptk7-deficient mouse model, we confirmed this hypothesis by demonstrating the role of this receptor in the homing of murine hematopoietic cells in the bone marrow (J. Immunol., 2016). We have also shown that PTK7 plays a prometastatic role in colon cancer with poor prognosis (PLoS ONE, 2015) and heterodimerizes with ROR2, another RTK of the non-canonical WNT pathway activated by WNT5A (J. Biol. Chem., 2015).
PTK7 contributes to adhesion, migration, and invasion of cancer cells, and is linked to tumor progression and bad prognosis. An antibody drug-conjugate against PTK7 induces sustained tumor regression in mice and is currently entering a phase II clinical trial, demonstrating the potential of targeting PTK7 in cancer. However, its mode of action in cancer is elusive.
We have used proximity biotinylation to investigate the deep proteome associated to PTK7 in colon cancer cells, some of the biotinylated proteins identified by mass spectrometry are currently studied at the molecular and functional levels. We have also generated a conditional ptk7-deficient mouse that will provide an extremely valuable animal model to study ptk7 in specific tissues, especially to investigate its role in colon carcinogenesis (Projet Libre INCa).
In collaboration with Paoli-Calmettes Institute, we continue our exploration of the role of PTK7 in colon cancer using patient samples.
Beside epithelial cells, cancer cells and hematopoietic stem cells, PTK7 is expressed in immune cells such as dendritic cells (DCs). However, its function in immunity and in the cancer-immunity cycle is largely unknown. We have started to investigate this by combining innovative approaches based on scRNA-seq technology, multiparametric spectral cytometry and genetic mouse model. This project is partially funded by the “Emergence 2020” grant from Canceropôle PACA/Gefluc.
Responsible for the Peutz-Jeghers syndrome and considered for a while as a critical kinase for epithelial polarity establishment and maintenance as in worms and flies, we put effort to unveil functions of the serine/threonine kinase LKB1 as tumor suppressor especially important in non-small cell lung carcinoma (NSCLC). Our works have shown that active LKB1 locates at adherens junctions to activate AMPK (Curr. Biol. 2009) which is regulated by mechanical forces submitted to E-Cadherin (coll. Nat. Cell. Biol. 2017). To date, patients affected by NSCLC with loss of LKB1 function are associated with poor prognosis being refractory to chemotherapy as well as immunotherapy. To decipher molecular mechanisms by which LKB1 exerts its tumor suppressor activity, we have chosen to investigate mechanisms regulating its catalytic activity. Indeed, altered LKB1 activity involvement in NSCLC is likely underestimated since, in addition of mutational events leading to its expression or activity loss, those related to its dysregulation are not considered. To define new potential therapeutic targets, we decipher in depth mechanisms and signaling triggered by LKB1activity loss using in vitro and in vivo (knockout mouse) approaches. By SILAC strategy, we have defined that the LKB1 affects phosphorylation of several candidates to repress canonical WNT activity found deregulated in tumors. Complementary proteomic approaches based on proximity labeling (APEX) has also been initiated. Last, involvement of LKB1 activity in mechano-transduction and its relevance in lungs permanently submitted to stretches and tensions associated to breath motions, lead us to complement our investigations through an emerging project aiming to study the role of LKB1 and WNT/PCP molecules in tumoral aggressiveness in conditions where stiffness of tumor microenvironment is modified (A*MIDEX interdisciplinary project Capostromex).
Our team is implicated in the national consortium NANOTUMOR [A1] exploring the molecular and cellular complexities of the tumoral cell. The NANOTUMOR consortium is a French national multi-disciplinary workforce that aims to study cancer initiation and progression at molecular and subcellular level, by combining cutting-edge technologies and expertises in electron and fluorescence 2D-3D imaging, spatial transcriptomics and mass-spectrometry, micropatterning and microfluidics/biomechanics in various cellular and animal models, anti-intrabody/PPI engineering, and high-throughput screening. It will explore several children and adults’ cancers characteristics from the structure of underlying molecular complexes, spatio-temporal genes and proteins expression patterns, all the way up to subcellular organization, tissues morphology/rheology, and ultimately drug design and screening.
Santoni MJ, Kashyap R, Camoin L, Borg JP
Daulat AM, Finetti P, Revinski D, Silveira Wagner M, Camoin L, Audebert S, Birnbaum D, Kodjabachian L, Borg JP, Bertucci F
Colin F, Schauer K, Hamiche A, Martineau P, Borg JP, Bednar J, Bertolin G, Camoin L, Collette Y, Dimitrov S, Fournier I, Hyenne V, Mendoza-Parra MA, Morelli X, Rondé P, Sumara I, Tramier M, Schultz P, Goetz JG