Skip to Main content Skip to Navigation

La Filamine A, acteur central relayant les mécanismes d'invasion des gliomes de haut-grade : implication dans la signalisation et le trafic du récepteur chimiotactique de l'urotensine II

Abstract : Multiform glioblastoma (GBM) are the most frequent and aggressive tumor of the central nervous system (CNS). The standard treatment therapy for GBM follows the “Stupp protocol” consisting in the most complete surgical resection combined with radio/chemotherapy with temozolomide (TMZ). Despite this heavy first line treatment, patients with GBM display a survival median of only 14.6 months. Even if the resection as large as possible, the diffuse properties of GBM cells leads to a quasi-systematic invasion of the margin of the resection cavity. The healthy brain parenchyma invasion by GBM cells, in margin or at distance of the resection cavity, constitutes a main therapeutic issue. These invasion mechanisms are carried by cell transformations caused by the microenvironment such as hypoxia and angiogenesis responsible for mesenchymal transition (MT) mainly controlled by two transcription factors STAT3 and CEBPα which stimulate 70% of secondary mesenchymal genes. Hypoxia associated with MT triggers the expression of chemokine G protein-coupled receptor (GPCRs). The most studied chemotactic GPCR in GBM is CXCR4 which mediates promigratory effects through Gαi/PI3K/PIP3 and Gα13/Rho/ROCK as well as its endocytosis and recycling mediated by β-arrestins. Our team already demonstrated that the UT receptor of the neuropeptide urotensin II (UII) behaves like a chemokine receptor and stimulates GBM directional migration by Gαi/PI3K/PIP3 and Gα13/Rho/ROCK allowing cell polarization, lamellipodia formation, actin stress fiber polymerization and cell contraction. Thus it appears that the expression and coupling redundancy of chemotactic GPCRs constitute a major brake for the development of targeted therapy against these systems. Based on these observations, we proposed to identify new protein partners common to these chemokine GPCRs which could then be targeted by future anti-invasive therapies. First, we validated the systematic redundancy of expression of CXCR4/SDF-1α and UT/UII systems by immunohistochemical studies carried in various patient glioma grades (Collaboration with Pr A. Laquerrière, CHU Rouen Hospital). These systems are more strongly co-expressed in pseudopalisadic peri-necrotic hypoxic GBM areas. In order to identify whether this interaction between FlnA and UT could exert a functional impact in GBM activity, we evaluated the FlnA mRNA expression levels by means of RNAseq data from The Cancer Genome Atlas (TCGA) and IVYgap (IVY glioblastoma atlas project) data bases. These analyses highlighted that FlnA expression levels are correlated with glioma grades, showing the highest levels in IDH1/2 (Isocitrate Dehydrogenase 1 & 2) wild type mesenchymal GBM. FlnA overexpression is shown associated with a decrease of overall survival and earlier recurrence. Moreover IVYgap data analysis combined with immunolabelings performed on glioma patient samples, revealed that is strongly express in pseudopalisades and hypoxic areas as well as in microvascular proliferation areas. Additional analysis FlnA-associated ontological gene networks based on genes which expression is the most correlated with FlnA in glioma, stressed main networks associated with PI3K pathway, cell polarization and adhesion or cytoskeleton regulation. In order to study FlnA intrinsic properties in GBM activity, FlnA expression levels were first confirmed in GBM cell lines under normoxia and hypoxia. Hypoxia triggered FlnA mRNA overexpression as well as increased FlnA and UT protein forms. We generated from the U87 GBM cell line, a KnockOut (KO) U87 cell line deleted for the gene encoding FlnA by using the CRISPR/Cas9 technology. FlnA deletion in U87 cells is associated with a significant decrease of their proliferative capacities, a mechanism characterized by cell accumulation in G1 cell cycle phase, suggesting a slow progression towards S phase and accumulation in senescence. Moreover, FlnA depletion was shown to sensitize U87 cells to TMZ and Irininotecan (CPT-11) tested at low concentrations (10-7M of each). Regarding motility, U87 cells displayed spontaneous production of lamellipodia, actin stress fibers and focal adhesion points, responsible for “mesenchymal endogenous” migration. However, FlnAKO cells exhibited a significant decrease of their motility associated with a reduction of focal adhesion points labeled by an anti phospho_paxillin antibody, associated with a major inhibition of actin stress fiber polymerization. The quantification of the number and the type of membrane protrusions on the two cell lines indicated that U87-FlnAKO are unable to generate and stabilize lamellipodia. These observations show that FlnA is essential for actin stress fiber polymerization, lamellipodia formation and focal adhesion maturation essential for chemotactic directional migration. Previous work of our team demonstrated that UT receptor mediates glioma chemotactic migration through Gαi/PI3K/PIP3 and Gα13/Rho/ROCK allowing lamellipodia expansion and vinculin focal adhesion maturation. In this study we showed that the absence of FlnA prevents adhesion/de-adhesion cycles stimulated by the UII (10-8M)-induced UT activation, as well as the directional migration in response to UII (10-12 M, 10-9 M and 10-7 M) in the Boyden chamber assay. This strongly suggests that FlnA may control GPCR-dependent chemotactic migration in general and UT in particular. We next investigated whether this interaction would play a role in UT couplings and trafficking. In the presence of FlnA, U87 cells expressing recombinant UT-eYFP uncoupled to Gαq, but precoupled within lamellipodia with Gαi , while this colocalization increased within the cytosol after UII treatment. In U87-FlnAKO cells, this Gαi precoupling was not found but UT appeared mainly colocalized with Gαq mainly after UII treatment. This new coupling acquisition was confirmed by specific calcium mobilization after UII treatment in U87-FlnAKO cells. Thus, FlnA seems essential for UT precoupling to Gαi and Gαq exclusion in GBM cells. During a sustain activation, UT receptor is present in membrane ruffles of U87 cells and could be internalized and endocytate after UII treatment (10-7M and 10-9M). These membrane ruffles are absent in FlnAKO cells and UT was expressed at the plasma membrane even after sustained UII treatment. We conclude here that FlnA plays a key role in the transport regulation of UT to the lamellipodia through ruffles expansion and in UT internalization/endocytosis process. Reciprocally, we demonstrated that UT activation in U87 cells triggered FlnA cleavage characterized by the detection of FlnA band at 110-130 kDa in Western blot, associated with a UT relocalization at the plasma membrane. These mechanisms are inhibited in presence of pertussis toxin (PTX) which inhibits Gαi activity while cells still emitted lamellipodia possibly via free  complexes allowing PI3K activation. Results of the receptor distribution quantification suggested that receptor relocation after stimulation is dependent on Gαi activity, PKA inhibition that would favor calpains activity to induce FlnA proteolysis and UT redistribution after UII stimulation. Finally, we performed orthotropic U87 and U87-FlnAKO cells xenograft in the striatum of NMRI/Nude mice. We observed a significant survival improvement of U87-FlnAKO injected mice compared with control mice injected with U87 cells. Immunolabelings performed on brain tumor slices obtained from sampled injected mouse brains at 15 days post-injection, showed a drastic modification of the U87-FlnAKO tumor phenotype. Indeed, U87-GBM displayed a necrotic tumor core associated with a highly invasive phenotype, sometimes multifocal with strong and intense FlnA and MMP-9 labelings, showings colocalisation between FlnA and UT in invasive processes, around UII-expressing components. U87-FlnAKO GBM rather exhibited a highly circumscribe phenotype accompanied by the loss of FlnA, UT and MMP-9 expression as well as by occurrence of a normal vascularization. However the CD44 labeling was not modified in the presence or the absence of FlnA in U87 GBM cells, suggesting that FlnA deletion is not controlling expression all mesenchymal markers, but rather represses downstream GBM invasive properties. Altogether these data show that FlnA plays a major role in proliferation and adhesion processes in GBM cells. Moreover, FlnA directly controls UT receptor signalings by conditioning in one hand its precoupling to Gαi and on the other hand by allowing its rapid and dynamic internalization/endocytosis and its recycling to the lamellipodia. Thus it is possible to develop therapeutic strategies specifically targeting FlnA/UT interaction in order to prevent healthy brain invasion mediated by UT, and more generally, by all chemokine receptors interacting with FlnA.
Complete list of metadata

Cited literature [734 references]  Display  Hide  Download
Contributor : Hélène CASTEL Connect in order to contact the contributor
Submitted on : Wednesday, December 18, 2019 - 4:16:20 PM
Last modification on : Wednesday, March 2, 2022 - 3:48:02 PM


 Restricted access
To satisfy the distribution rights of the publisher, the document is embargoed until : 2023-12-31

Please log in to resquest access to the document


  • HAL Id : tel-02418263, version 1


Alexandre Mutel. La Filamine A, acteur central relayant les mécanismes d'invasion des gliomes de haut-grade : implication dans la signalisation et le trafic du récepteur chimiotactique de l'urotensine II. Sciences du Vivant [q-bio]. Normandie Université, France, 2019. Français. ⟨tel-02418263⟩



Record views