Thyroid cancer (TC) is the most common endocrine malignancy. The sodium–iodide symporter (NIS), responsible for active transport of iodide into thyroid cells, allows the use of radioactive iodine (RAI) as the systemic treatment of choice for TC metastatic disease. Still, patients with advanced forms of TC often lose the ability to respond to RAI therapy, which results in worse survival rates. We have shown that the overexpression of RAC1b, a tumor-related RAC1 splice variant, is associated with less favorable clinical outcomes in differentiated TCs derived from the follicular epithelial (DTCs). RAC1b overexpression is also significantly associated with the presence of MAPK-activating BRAFV600E mutation, which has been previously implicated in the loss of NIS expression. Here, we show that increased RAC1b levels are associated with NIS downregulation in DTCs and demonstrate that ectopic overexpression of RAC1b in non-transformed thyroid cells is sufficient to decrease TSH-induced NIS expression, antagonizing the positive effect of the canonically spliced RAC1 GTPase. Moreover, we clearly document for the first time in thyroid cells that both NIS expression and iodide uptake are hampered by RAC1 inhibition, highlighting the role of RAC1 in promoting TSH-induced NIS expression. Our findings support a role for RAC1 and RAC1b signaling in the regulation of NIS expression in thyroid cells and suggest that RAC1b in cooperation with other cancer-associated signaling cues may be implicated in the response of DTCs to RAI therapy.
Figure 1 supplemental: NIS and RAC1b expression data retrieved from TCGA database. NIS and RAC1b isoform expression data was collected from The Cancer Genome Atlas (TCGA) project [Thyroid Carcinoma (THCA) project] using the TSVdb tool (Sun et al. 2018). NIS and RAC1b RNA-Seq data was retrieved in the RSEM format (Li & Dewey 2011). Results are depicted as Tukey’s boxplots of the LOG2 RSEM values of each sample, comparing primary PTC (n=501) and normal solid thyroid tissue (n=59) data sets. The existence of significant differences between the two datasets was evaluated using a two-tailed Mann-Whitney’s u-test, assuming a significance level of α=0.05.