Sustained ERK inhibition maximizes responses of BrafV600E thyroid cancers to radioiodine

Abstract

Radioiodide (RAI) therapy of thyroid cancer exploits the relatively selective ability of thyroid cells to transport and accumulate iodide. Iodide uptake requires expression of critical genes that are involved in various steps of thyroid hormone biosynthesis. ERK signaling, which is markedly increased in thyroid cancer cells driven by oncogenic BRAF, represses the genetic program that enables iodide transport. Here, we determined that a critical threshold for inhibition of MAPK signaling is required to optimally restore expression of thyroid differentiation genes in thyroid cells and in mice with BrafV600E-induced thyroid cancer. Although the MEK inhibitor selumetinib transiently inhibited ERK signaling, which subsequently rebounded, the MEK inhibitor CKI suppressed ERK signaling in a sustained manner by preventing RAF reactivation. A small increase in ERK inhibition markedly increased the expression of thyroid differentiation genes, increased iodide accumulation in cancer cells, and thereby improved responses to RAI therapy. Only a short exposure to the drug was necessary to obtain a maximal response to RAI. These data suggest that potent inhibition of ERK signaling is required to adequately induce iodide uptake and indicate that this is a promising strategy for the treatment of BRAF-mutant thyroid cancer.

Figure 1. Profound inhibition of MAPK signaling is required to restore differentiated gene expression in thyroid PCCL3-BRAF cells and in murine Braf^V600E-induced PTCs.

(A) Western blots of PCCL3-BRAF cells treated with dox for 4 days to induce BRAF^V600E, and then for 3 days with the indicated concentrations (μM) of PLX4032 or U0126. (B) Graph shows loading-adjusted p-ERK vs. NIS levels from the Western blot. The large green and orange dots indicate the –dox and +dox conditions, respectively. (C and D) Western blots of PTCs from TPO-Cre LSL-Braf^V600E mice treated with vehicle, AZD6244 (50 mg/kg twice per day), or CKI (1.5 mg/kg/d) for 3 days. On the fourth day tissues were harvested at the indicated times after dosing while remaining on the same treatment schedule. Vehicle lanes represent mice that never received active compound (they do not represent a time 0). (E) Western blots of TPO-Cre LSL-Braf^V600E mouse PTCs (n = 3) treated with the indicated compounds for 4.5 days. Thyroid lobes were collected 2 hours after the final dose. (F and G) Quantitative RT-PCR of MAPK transcriptional output markers (F) or iodine metabolism–related genes (G) in thyroid tissues from mice treated with the indicated doses of AZD6244 (n = 3) or CKI (n = 5) for 4.5 days. Data represent percentage change in β-actin–normalized expression compared with vehicle-treated TPO-Cre LSL-Braf^V600E (F) or wild-type mice (G). **P = 0.008, ***P = 0.0003, ****P < 0.0001, Mann-Whitney test. QD, once per day; BID, twice per day; tERK, total ERK.

Figure 2. ¹⁸F-TFB uptake and kinetic analysis in IEC6 intestinal rat cells and in TPO-Cre LSL-Braf^V600E mouse PTCs in response to MAPK inhibitors.

(A) Rates of TFB uptake (2-minute time points) were determined at 140 mM Na⁺ with various concentrations of TFB. The K_m and V_max were 9.4 ± 1.1 μM and 17 ± 1.1 pmol/μg DNA/2 min, respectively. (B) ¹⁸F-TFB uptake in IEC6 cells in the absence or presence of ClO₄^– or sodium (mean ± SD). (C) Fluorine is not transported via NIS. (D) Time course of ¹⁸F-TFB uptake by PET of PTCs of TPO-Cre LSL-Braf^V600E mice treated with vehicle (n = 5), AZD6244 (50 mg/kg twice per day, n = 5) or CKI (1.5 mg/kg once per day, n = 5) for 21 days. Graph shows thyroid uptake of ¹⁸F-TFB normalized for tumor volume given in % uptake of injected activity (*P = 0.008, CKI vs. AZD6244). Upper row shows representative axial PET thyroid images of mice treated with CKI (magnification ×2.5). Serum TSH was markedly increased in TPO-Cre LSL-Braf^V600E compared with wild-type mice, as these animals become hypothyroid upon Braf^V600E expression, and were not significantly different in vehicle- vs. CKI/AZD6244-treated animals (not shown). Hence, all ¹⁸F-TFB uptake experiments were performed under TSH-stimulated conditions.

Figure 3. Profound MAPK pathway blockade with CKI maximizes ¹²⁴I uptake and response to RAI therapy in Braf-induced PTCs.

(A) Protocol to explore effects of MAPK pathway inhibitors on ¹²⁴I incorporation in TPO-Cre LSL-Braf^V600E mice. (B and C) Thyroid uptake of ¹²⁴I normalized for tumor volume was used to calculate: (B) maximum uptake of ¹²⁴I and (C) predicted lesional radiation dose if given 1 mCi (37 MBq) of ¹³¹I. **P = 0.0043, *P = 0.017, Mann-Whitney test. (D) Tumor volume measured by ultrasound at time 0 (4.5 days after treatment with vehicle, AZD6244, or CKI, and prior to administration of ¹³¹I), and at the indicated times after RAI therapy. The green line shows PTC volume in age-matched untreated mice. Five mice per group were treated with the indicated drugs. Two mice of the vehicle group died after 12 weeks. Volume differences in all groups were statistically significant (at 24 weeks: ****P < 0.0001 by Kruskal-Wallis test; **P = 0.008 for AZD6244 vs. CKI by Mann-Whitney test). (E–G) Histology of PTCs of TPO-Cre LSL-Braf^V600E mice at 24 weeks. Scale bars: 1,000 μm. (E) H&E, (F and G) IHC for pan-cytokeratin (F) or p-ERK (G) of representative sections from untreated mice, or mice treated with vehicle + ¹³¹I, AZD6244 + ¹³¹I, or CKI + ¹³¹I. Black arrows point to vestigial remains of PTC in thyroid lobes after CKI + ¹³¹I.