Thyroid myopericytoma harboring the BRAFV600E mutation: a novel in vitro and in vivo model of thyroid tumor microenvironment using vemurafenib therapy

Background: Myopericytoma is a mesenchymal tumor characterized by perivascular proliferation of oval to spindled cells with myoid and pericytic features. BRAFV600E mutation can occur in epithelial or mesenchymal cancers with a crucial role in tumor cell-extracellular matrix (ECM) adhesion (microenvironment). Experimental Design: We report the first case (45 years old woman) of myopericytoma arising in the thyroid and harboring the BRAFV600E mutation. Thyroid imaging and 18FDG PET/CT were performed to assess disease stage. DNA large scale genotyping and immunohistochemistry were performed on excised thyroid myopericytoma (TM) tissue and tissue-derived primary TM cell cultures. Mutant BRAFV600E mRNA was knocked-down in TM cells by specific short hairpin (sh) RNA with cell adhesion/migration and senescence assays performed. ECM genes were assesses by qPCR/multiprofiling gene expression, and anchorage (Matrigel) assays were developed. TM cell viability was determined with vemurafenib (a novel orally available selective inhibitor of BRAFV600E). Angiogenesis in vitro assays were performed using microvascular blood endothelial cells, three-dimensional (3D) cell cultures with Matrigel (tumor microenvironment system), and confocal microscopy. We have developed a patient derived xenograft mouse model using primary TM cells engineered to express green fluorescent protein (GFP). Results: Thyroid imaging revealed a goiter associated with an irregular/hypoechoic/hypervascularized/hypofunctional mass in the left thyroid lobe with mediastinal extension and high 18FDG uptake. The mass was diagnosed as TM and harbored the heterozygous BRAFWT/V600E mutation, and showed extension to the peri-thyroidal fibroconnective tissue but no neck lymph nodes/distant metastases were detected. Both TM tissue and established primary TM cells harbored BRAFWT/V600E and were positive for PDGFRB, ?-smooth muscle actin, phosphoERK1/2, Ki67, and stem cell markers (e.g. CD44). A higher vascular density (CD31 expression) was also found in the TM vs normal thyroid. PTH and thyroid-specific markers were negative. Sh-control TM cells expressed high levels of ECM molecules (mRNA and protein: e.g. integrins, collagens, etc.); as a result, they grew as cell aggregates in Matrigel. Sh-BRAFV600E TM cells showed a significant reduction of cell adhesion/migration and increased senescence vs. controls. Vemurafenib significantly down-regulated ECM genes expression (e.g. integrins and collagens) and reduced viability and migration of TM cells vs controls. Furthermore, TM cells triggered angiogenesis (tubles formations) by recruiting microvascular blood endothelial cells in vitro in a 3D cell co-culture system that recapitulates tumor microenvironment in vitro; treatment with vemurafenib significantly disrupted angiogenesis compared to vehicle (control) treatment. Furthermore, our preliminary results suggest that TM-derived xenograft SCID mice show a significant therapeutic response to vemurafenib compared to vehicle-treated mice. Conclusions: We report the first TM and primary TM cell culture with the heterozygous BRAFWT/V600E mutation. We show a causal role for BRAFWT/V600E/ hyper activation of ERk1/2 signaling in myopericytoma development. We demonstrate BRAFV600E-dependent cell adhesion, migration, and angiogenesis. These results also suggest that vemurafenib might have a potential therapeutic efficacy in patients with BRAFWT/V600E-positive thyroid myopericytoma or at other locations inhibiting both cell viability and angiogenesis.