MIEAP and ATG5 are tumor suppressors in a mouse model of BRAFV600E-positive thyroid cancer

Mitochondria-eating protein (MIEAP) is a molecule important for non-canonical mitophagy and thought to be a tumor suppressor. Our previous study found that MIEAP expression is defective in thyroid oncocytomas, irrespective of being benign or malignant, and also in non-oncocytic thyroid cancers. Thyroid oncocytomas are composed of large polygonal cells with eosinophilic cytoplasm that is rich in abnormal mitochondria. Thus, our data indicate that, together with increased mitochondrial biogenesis that compensates for the dysfunction of the mitochondria, MIEAP plays a critical role in the accumulation of mitochondria in thyroid oncocytic tumors, whereas a defective MIEAP expression alone is not sufficient for mitochondrial accumulation in non-oncocytic cancers with normal mitochondria. To clarify whether MIEAP is a tumor suppressor in the thyroids and whether MIEAP knockout (KO) alone is sufficient for the oncocytic phenotype and also to extend our effort toward canonical mitophagy (a selective autophagy), we here conducted mouse studies using genetically engineered mice. Braf CA/wt mice developed thyroid cancers 1 year after intrathyroidal injection of adenovirus expressing Cre, while cancer development was observed at 6 months in adenovirus-Cre-injected Braf CA/wt ;Mieap KO/KO and Braf CA/wt ;Atg5 flox/flox mice [where autophagy-related 5 (ATG5) is a component of autophagic machinery], although KO of either molecule alone was not sufficient for cancer development. These data demonstrate that MIEAP or ATG5 KO accelerated thyroid cancer development. However, cancers in adenovirus-Cre-injected Braf CA/wt ; Mieap KO/KO and Braf CA/wt ; Atg5 flox/flox mice were not oncocytic. In conclusion, we here show that MIEAP and ATG5 are both tumor suppressors in thyroid carcinogenesis, but as we have anticipated from our previous data, KO of either molecule does not confer the oncocytic phenotype to BRAF V600E -positive thyroid cancers. The combination of disruptive mitochondrial function and impaired mitochondrial quality control may be necessary to establish a mouse model of thyroid oncocytoma.