Revising the embryonic origin of thyroid C cells in mice and humans.

Current understanding infers a neural crest origin of thyroid C cells, the major source of calcitonin in mammals and ancestors to neuroendocrine thyroid tumors. The concept is primarily based on investigations in quail-chick chimeras involving fate mapping of neural crest cells to the ultimobranchial glands that regulate Ca(2+) homeostasis in birds, reptiles, amphibians and fishes, but whether mammalian C cell development involves a homologous ontogenetic trajectory has not been experimentally verified. With lineage tracing, we now provide direct evidence that Sox17+ anterior endoderm is the only source of differentiated C cells and their progenitors in mice. Like many gut endoderm derivatives, embryonic C cells were found to coexpress pioneer factors forkhead box (Fox) a1 and Foxa2 before neuroendocrine differentiation takes place. In the ultimobranchial body epithelium emerging from pharyngeal pouch endoderm in early organogenesis, differential Foxa1/Foxa2 expression distinguished two spatially separated pools of C cell precursors with different growth properties. A similar expression pattern was recapitulated in medullary thyroid carcinoma cells in vivo, consistent with a growth-promoting role of Foxa1. In contrast to embryonic precursor cells, C cell-derived tumor cells invading the stromal compartment downregulated Foxa2, foregoing epithelial-to-mesenchymal transition designated by loss of E-cadherin; both Foxa2 and E-cadherin were re-expressed at metastatic sites. These findings revise mammalian C cell ontogeny, expand the neuroendocrine repertoire of endoderm and redefine the boundaries of neural crest diversification. The data further underpin distinct functions of Foxa1 and Foxa2 in both embryonic and tumor development.

Linking loss of sodium-iodide symporter expression to DNA damage.

Radiotherapy of thyroid cancer with I-131 is abrogated by inherent loss of radioiodine uptake due to loss of sodium iodide symporter (NIS) expression in poorly differentiated tumor cells. It is also known that ionizing radiation per se down-regulates NIS (the stunning effect), but the mechanism is unknown. Here we investigated whether loss of NIS-mediated iodide transport may be elicited by DNA damage. Calicheamicin, a fungal toxin that specifically cleaves double-stranded DNA, induced a full scale DNA damage response mediated by the ataxia-telangiectasia mutated (ATM) kinase in quiescent normal thyrocytes. At sublethal concentrations (<1nM) calicheamicin blocked NIS mRNA expression and transepithelial iodide transport as stimulated by thyrotropin; loss of function occurred at a much faster rate than after I-131 irradiation. KU-55933, a selective ATM kinase inhibitor, partly rescued NIS expression and iodide transport in DNA-damaged cells. Prolonged ATM inhibition in healthy cells also repressed NIS-mediated iodide transport. ATM-dependent loss of iodide transport was counteracted by IGF-1. Together, these findings indicate that NIS, the major iodide transporter of the thyroid gland, is susceptible to DNA damage involving ATM-mediated mechanisms. This uncovers novel means of poor radioiodine uptake in thyroid cells subjected to extrinsic or intrinsic genotoxic stress.

Differential effects of MAPK pathway inhibitors on migration and invasiveness of BRAF(V600E) mutant thyroid cancer cells in 2D and 3D culture.

Tumor microenvironment influences targeted drug therapy. In this study we compared drug responses to RAF and MEK inhibitors on tumor cell migration in 2D and 3D culture of BRAF(V600E) mutant cell lines derived from human papillary (BCPAP) and anaplastic (SW1736) thyroid carcinomas. Scratch wounding was compared to a double-layered collagen gel model developed for analysis of directed tumor cell invasion during prolonged culture. In BCPAP both PLX4720 and U0126 inhibited growth and migration in 2D and decreased tumor cell survival in 3D. In SW1736 drugs had no effect on migration in 2D but decreased invasion in 3D, however this related to reduced growth. Dual inhibition of BRAF(V600E) and MEK reduced but did not prevent SW1736 invasion although rebound phosphorylation of ERK in response to PLX4720 was blocked by U0126. These findings indicate that anti-tumor drug effects in vitro differ depending on culture conditions (2D vs. 3D) and that the invasive features of anaplastic thyroid cancer depend on non-MEK mechanism(s).

Dual contribution of MAPK and PI3K in epidermal growth factor-induced destabilization of thyroid follicular integrity and invasion of cells into extracellular matrix.

Normal thyrocytes grown as reconstituted follicles in collagen gel were evaluated for drug effects of small molecule kinase inhibitors on growth factor-induced cell migration in a 3D context. MEK inhibition by U0126 only partially antagonized EGF/serum-induced cell migration from the basal follicular surface into the matrix. Combined treatment with U0126 and LY294002, a PI3K blocker, was necessary to abolish migration. However, exposure to only LY294002 facilitated the response to EGF by breakdown of the original follicular structure. In the same time EGF promoted thyroid cell survival that was compromised by LY294002 in absence of EGF. Cells treated with EGF and LY294002 retained the ability to form follicles. The findings indicate that dual inhibition of MAPK and PI3K/AKT pathways is required to fully block matrix invasion of EGF-stimulated thyroid cells. Conversely, single drug treatment with PI3K inhibitor adversely promotes invasiveness probably by destabilizing the follicular epithelium.

Switching from MAPK-dependent to MAPK-independent repression of the sodium-iodide symporter in 2D and 3D cultured normal thyroid cells.

Loss of sodium-iodide symporter (NIS) expression in thyroid tumour cells primarily caused by constitutive MAPK pathway activation is often refractory to small molecule MAPK inhibitors. Suggested mechanisms are rebound MAPK signalling and activation of alternative signalling pathways. Here we provide evidence that failure to recover down-regulated NIS by MEK inhibition is not specific to tumour cells. NIS mRNA levels remained repressed in TSH-stimulated primary thyroid cells co-treated with epidermal growth factor (EGF) and pan-MEK inhibitor U0126 in the presence of 5% fetal bovine serum or, independently of serum, in 3D cultured thyroid follicles. This led to inhibited iodide transport and iodination. In contrast, U0126 restituted thyroglobulin synthesis in EGF-treated follicular cells. Serum potentiated TSH-stimulated NIS expression in 2D culture. U0126 blocked down-regulation of NIS only in serum-starved cells with a diminished TSH response. Together, this suggests that morphogenetic signals modify the expression of NIS and recovery response to MEK inhibition.