Inhibition of ALK-Signaling Overcomes STRN-ALK-Induced Downregulation of the Sodium Iodine Symporter and Restores Radioiodine Uptake in Thyroid Cells

Abstract

Background:

Radioiodine (RAI) is commonly used for thyroid cancer treatment, although its therapeutic benefits are restricted to iodine-avid tumors. The RAI-refractory disease develops with tumor dedifferentiation involving loss of sodium-iodine symporter (NIS). Thyroid cancers driven by ALK fusions are prone to dedifferentiation, and whether targeted ALK inhibition may enhance RAI uptake in these tumors remains unknown. The aim of this study was to determine the levels of NIS expression during the progression of ALK fusion-driven thyroid cancer, assess the effects of ALK activation on NIS-mediated RAI uptake, and test pharmacological options for its modulation.

Methods:

The expression of NIS at different stages of ALK-driven carcinogenesis was analyzed using a mouse model of STRN-ALK-driven thyroid cancer. For in vitro experiments, a system of doxycycline-inducible expression of STRN-ALK was generated using PCCL3 normal thyroid cells. The STRN-ALK-induced effects were evaluated with quantitative reverse transcription polymerase chain reaction, Western blot, immunofluorescence, RNA sequencing, and gene sets pathways analyses. RAI uptake was measured using ¹³¹I. Treatment experiments were done with FDA-approved ALK inhibitors (crizotinib and ceritinib), MEK inhibitor selumetinib, and JAK1/2 inhibitor ruxolitinib.

Results:

We found that Nis downregulation occurred early in ALK-driven thyroid carcinogenesis, even at the stage of well-differentiated cancer, with a complete loss in poorly differentiated thyroid carcinomas. Acute STRN-ALK expression in thyroid cells resulted in increased MAPK, JAK/STAT3, and PI3K/AKT/mTOR signaling outputs associated with significant ALK-dependent downregulation of the majority of thyroid differentiation and iodine metabolism/transport genes, including Slc5a5 (Nis), Foxe1, Dio1, Duox1/2, Duoxa2, Glis3, Slc5a8, and Tg. Moreover, STRN-ALK expression in thyroid cells induced a significant loss of membranous NIS and a fourfold decrease of the NIS-mediated RAI uptake, which were reversed by ALK inhibitors crizotinib and ceritinib. In addition, a strong dose-dependent restoration of NIS with its membranous redistribution in STRN-ALK-expressing thyroid cells was observed after inhibition of MAPK signaling with selumetinib, which exhibited a cumulative effect with JAK1/2 inhibitor ruxolitinib.

Conclusions:

The findings of this preclinical study showed that ALK fusion-induced downregulation of NIS, the prerequisite of RAI refractoriness, could be reversed in thyroid cells by either direct inhibition of ALK or its downstream signaling pathways.

Introduction

Since the 1950s, radioiodine (RAI) therapy has been the main therapeutic approach for patients with high-risk thyroid cancer. Unfortunately, the benefits of RAI therapy are limited due to the iodine avidity loss in dedifferentiated thyroid tumors. The development of RAI-refractory (RAI-R) thyroid cancer disease is associated with downregulation of thyroid differentiation and iodine transport/metabolism genes, particularly SLC5A5, which encodes sodium (Na)-iodide symporter (NIS), the main plasma membrane channel required for iodine uptake.

Recently, it has been shown that MAPK inhibition restores the expression of thyroid differentiation genes and RAI uptake in BRAF^V600E-driven thyroid tumors in mice.^1,2 Consequently, targeted mono- and polytherapies against MEK and mutant BRAF were reported to resensitize to RAI the RAI-R thyroid carcinomas, mainly BRAF and RAS-mutant, in more than half of patients.^3

–7 Although single case reports showed the enhanced RAI uptake in NTRK and RET fusion-positive thyroid carcinomas by specific inhibitors,^8,9 the effect of targeted therapies on redifferentiation and RAI-resensitization of thyroid tumors driven by gene fusions has not been studied.

Fusions of ALK are identified as drivers in well-differentiated thyroid carcinoma (WDTC), poorly differentiated thyroid carcinoma (PDTC), and anaplastic thyroid carcinoma (ATC) in humans.^10

–15 Furthermore, using mouse model, we confirmed that STRN-ALK-driven thyroid carcinomas are prone to step-wise progression and dedifferentiation.¹⁶ Several generations of ALK inhibitors have been FDA approved for the treatment of nonsmall-cell lung cancer.¹⁷ Promising data on response to ALK inhibitors were reported for single cases of RAI-R ALK fusion-positive thyroid carcinomas treated with first (crizotinib) and second (ceritinib and brigatinib) generations of ALK tyrosine kinase inhibitors.^18

–21 However, whether ALK fusion-driven thyroid carcinomas exhibit loss of functional NIS and treatment of these tumors with ALK inhibitors has effect on RAI avidity remains unknown.

The goal of this study was to assess the expression of NIS in ALK fusion-induced tumors at different stages of progression, dissect the effects of ALK activation in thyroid cells on NIS expression and RAI uptake, and investigate pharmacological options for their modulation.

Materials and Methods

Detailed information on materials and methods is provided in Supplementary Data. In brief, the expression of NIS at different stages of STRN-ALK-driven thyroid cancer was done on samples from the established mouse model¹⁶ using RNA sequencing and immunofluorescence analysis (IFA). Study was approved by Institutional Animal Care and Use Committee, and all experimental procedures were performed in accordance with federal guidelines and institutional policies. An in vitro system of doxycycline (Dox)-inducible expression of STRN-ALK was established using PCCL3 normal thyroid cells. STRN-ALK-dependent effects were evaluated with RNA sequencing, quantitative reverse transcription polymerase chain reaction, Western blot, and IFA.

The antibodies used in this study are listed in Supplementary Table S1. Gene set pathway analysis was performed using gene set enrichment analysis^22,23 (MSidDB database, QIAGEN IPA (QIAGEN),²⁴ and Z-scores (Supplementary Table S2). RAI uptake was determined using 4 μCi of ¹³¹I and competitive NIS inhibitor NaClO₄ (100 μM) as a control. Treatment experiments were done with FDA-approved ALK inhibitors (crizotinib and ceritinib), MEK inhibitor selumetinib, and JAK1/JAK2 inhibitor ruxolitinib. Statistical analysis and graphs generation were done using Prism (GraphPad Software, Inc., San Diego, CA, USA), OriginPro (OriginLab, Northampton, MA, USA), and R Studio (version 2022.02.3).

Results

Downregulation of NIS in ALK fusion-driven thyroid tumors occurs at the stage of well-differentiated cancer

Using thyroid samples from the established mouse model of STRN-ALK-driven thyroid cancer,¹⁶ we evaluated NIS expression in ALK-fusion positive cancers in comparison with normal and hyperplastic thyroid tissues. RNA-sequencing data showed significant 27-fold (p = 0.034) and 45-fold (p = 0.003) median decrease of Nis expression in WDTC and PDTC (represented by two types of PDTC described previously¹⁶), respectively (Fig. 1A). The downregulation of Nis expression, as well as decrease of thyroid differentiation score (TDS), showed a correlation (r = −0.69, p = 0.0006 and r = −0.72, p = 0.00023, respectively) with MAPK signaling output (Fig. 1B and Supplementary Fig. S1).

FIG. 1.

Downregulation of NIS expression in murine ALK fusion-driven tumors and its association with increased MAPK output. (A), Expression of Slc5a5 (Nis) in normal thyroids (n = 3), hyperplastic nodules (HN, n = 3), WDTC (n = 5), and PDTC (type 1, n = 5 [blue dots]; type 2, n = 5 [purple dots]) STRN-ALK-driven thyroid cancers in mice. Box plot with median line. Whiskers represent SD. Kruskal–Wallis with Dunn's posthoc test. (B) Correlation between ERK score and Nis expression in thyroid samples. (C) Representative histopathology and immunofluorescence microphotographs of the STRN-ALK-driven thyroid cancer case showing increased phosphorylation of ERK and loss of membranous NIS in dedifferentiated areas. PDTC was classified as type 1.¹⁶ NIS, sodium-iodine symporter; PDTC, poorly differentiated thyroid carcinoma; WDTC, well-differentiated thyroid carcinoma.

The analysis of ALK fusion-positive murine tumors with WDTC-to-PDTC transition revealed that the increased MAPK signaling output in histologically dedifferentiated zones was accompanied by loss of membranous NIS (Fig. 1C). Therefore, the downregulation of NIS occurred early in ALK fusion-driven thyroid carcinogenesis and was associated with increased MAPK output. For further investigation of the mechanisms of NIS downregulation by ALK signaling in thyroid cells and possibilities to reverse it, we proceeded with an in vitro system of inducible STRN-ALK expression.

Expression of STRN-ALK in thyroid cells leads to ALK fusion-dependent downregulation of NIS. Blocking of ALK signaling with ALK inhibitors restores membranous NIS expression and RAI uptake

The generation of the in vitro system of inducible expression of STRN-ALK (Fig. 2A) was performed using PCCL3 rat normal thyroid cell line that preserves expression of iodine metabolism/transport and thyroid differentiation genes. Genetic constructs encoding TetOn3G and TRE3G-LucTdTomato-T2A-HA-STRN-ALK were delivered, and independent clonal lines (Tet-tdT-ALK) were established. Treatment of transgenic Tet-tdT-ALK cells with Dox resulted in loss of epithelial features and acquiring of epithelial–mesenchymal transition (EMT) phenotype (Fig. 2B). The majority of Tet-tdT-ALK cells coexpressed both tdTomato reporter protein and STRN-ALK chimeric oncoprotein (Fig. 2C). The expressed full-size HA-tagged STRN-ALK exhibited its tyrosine kinase activity, represented by STRN-ALK autophosphorylation and activation of the MAPK signaling (Fig. 2D).

FIG. 2.

STRN-ALK-dependent downregulation of NIS and RAI uptake in thyrocytes can be reversed by ALK inhibitors. (A) Scheme of the genetic constructs and approach of transgenes expression. Luc-tdT, luciferase-tdTomato reporter protein. (B) Phenotypical changes of Tet-tdT-ALK cells after treatment with Dox. (C) Coexpression of the STRN-ALK and tdTomato reporter in majority of Tet-tdT-ALK cells after induction with Dox. (D) Western blot analysis of the representative Tet-tdT-ALK clones and Tet-tdT STRN-ALK-negative control cells. (E) Time-dependent downregulation of the membranous NIS in Tet-tdT-ALK clone 1 after induction with Dox. (F) NIS expression and MAPK signaling output in Tet-tdT-ALK clone 6 at different time points of STRN-ALK induction, its withdrawal and treatment with crizotinib (500 nM) or ceritinib (500 nM). (G) Dose-dependent effects of ALK inhibitors on Nis expression in uninduced (Dox−) and induced (Dox+, pretreated for 48 hours) Tet-tdT-ALK cells mixture (clones 7, 8, and 9). Data presented as mean ± SD. Green dashed line represents the baseline Nis mRNA level in untreated uninduced cells. Experiment was performed in triplicates. (H) Dose-dependent restoration of NIS and inhibition of MAPK signaling in Dox-induced Tet-tdT-ALK cells mixture (clones 7, 8, and 9) treated with crizotinib and ceritinib for 60 hours. For better evaluation of NIS levels, protein samples were pretreated with PNGase F. (I) Restoration of the membranous NIS after 60 hours of treatment with crizotinib (500 nM) and ceritinib (500 nM) in STRN-ALK-expressing cells (based on tdTomato expression). (J) RAI uptake by WT cells and transgenic Tet-tdT-ALK cells mixture (clones 7, 8, and 9). STRN-ALK was induced for 72 hours. Crizotinib and ceritinib were applied for 60 hours. Bars represent mean ± SD. One-way ANOVA with Tukey's post hoc test. Experiment was performed in quadruplicates. ANOVA, analysis of variance; Dox, doxycycline; RAI, radioiodine; SD, standard deviation; WT, wild-type.

Noticeable decrease of NIS expression in thyrocytes was observed after 32 hours of STRN-ALK induction (Fig. 2D, E), which was not seen in wild-type (WT) PCCL3 cells treated with Dox (Supplementary Fig. S2A, B). To determine whether downregulation of NIS in thyrocytes was STRN-ALK dependent, we probed NIS at different time points post-Dox administration, Dox withdrawal, and treatment with ALK inhibitors crizotinib and ceritinib.

Activation of the MAPK signaling pathway was detected as early as 24 hours after STRN-ALK induction, and maximum downregulation of NIS was observed after 72 hours. After withdrawal of Dox, STRN-ALK expression rapidly decreased with no detectable level after 72 hours post-Dox removal. Complete loss of STRN-ALK was accompanied by marked decline in ERK phosphorylation and increase of NIS expression to the level seen in uninduced thyrocytes. Similarly, when ALK inhibitors crizotinib and ceritinib were applied, cells showed a prominent decrease of MAPK signaling and restoration of NIS expression within 72 hours (Fig. 2F).

Next, we established whether the effect of ALK inhibitors on NIS restoration in STRN-ALK-expressing cells was dose dependent. Three independently generated clonal lines with clear NIS downregulation after STRN-ALK induction (Supplementary Fig. S3) were mixed (clones 7, 8, and 9) and subjected to the treatment experiments. Compared with dimethyl sulfoxide (DMSO)-treated cells, Dox-induced Tet-tdT-ALK cells showed significant (p = 0.005) threefold decrease of Nis mRNA expression. When these cells were treated with ALK inhibitors, clear dose-dependent increase of Nis was observed for crizotinib (R ² = 0.34, p = 1.4E-8) and ceritinib (R ² = 0.3, p = 0.001).

Although both inhibitors exhibited more that twofold overstimulation of Nis expression at concentration of 500 nM, the ceritinib showed a more potent effect, especially at lower doses (Fig. 2G). Dose-dependent restoration of NIS levels was accompanied by inhibition of ERK phosphorylation (Fig. 2H). Further IFA showed a clear evidence of the membranous NIS retrieval after treatment with ALK inhibitors (Fig. 2I).

For functional evaluation of NIS, RAI(¹³¹I)-uptake measurements were performed (Fig. 2J). There was no significant (p = 0.059) difference in RAI uptake by DMSO-treated WT and uninduced Tet-tdT-ALK cells. Induction of STRN-ALK resulted in significant fourfold (p < 0.0001) decrease of RAI accumulation in thyrocytes. The RAI uptake was restored to 62% and 98% (compared with uninduced Tet-tdT-ALK cells) after treatment with crizotinib and ceritinib, respectively (p < 0.0001, both). In comparison with crizotinib, ceritinib showed a more potent effect (p < 0.0001) on restoration of RAI uptake when applied at the same dose.

Thus, STRN-ALK expression in thyroid cells was associated with increased MAPK signaling, and resulted in downregulation of NIS and diminished uptake of RAI, which could be reversed by ALK inhibitors.

ALK fusion expression in thyroid cells induces transcriptional profile shift toward dedifferentiation, increase of MAPK, JAK/STAT, and PI3K/AKT/mTOR transcriptional outputs that can be reversed by ALK inhibitors

To explore the transcriptional profile of STRN-ALK signaling and effects of ALK inhibitors on it, we performed RNA-sequencing analysis on three independently generated clonal lines. The principal component analysis of the protein-coding genes revealed no notable variability between cell lines. However, the experimental groups were clustered separately, with STRN-ALK-expressing cells showing the highest level of variance relatively to uninduced cells. Removal of ALK signaling by Dox withdrawal or treatment with ALK inhibitors shifted transcriptional profiles of STRN-ALK-expressing cells, especially after ceritinib treatment, closer to uninduced cells (Fig. 3A).

FIG. 3.

RNA sequencing analysis of the STNR-ALK-dependent transcriptional outputs. (A) Clustering of the Tet-tdT-ALK clonal lines (7, 8, and 9) (n = 3 per each experimental group) based on PCA of the expression of protein-coding genes. ND, no Dox; D, STRN-ALK induction with Dox for 72 hours; dW, Dox withdrawal for 60 hours; dCRZ, treatment with 500 nM crizotinib for 60 hours; dCRT, treatment with 500 nM ceritinib for 60 hours. (B) Overlap of DEG (absLogFC >1, FDR p-value ≤0.05) between treatment groups. (C), Heatmap showing clustering results of the top ranked DEG (baseMean >1000, absLogFC >1, p-value <0.05). (D) Heatmap clustering of thyroid differentiation genes. ns, no significance. (E) TDSs in experimental groups. eTDS, enhanced TDS. (F) Top 20 significantly enriched pathways in Tet-tdT-ALK cells after induction of STRN-ALK, its withdrawal, and treatment with ALK inhibitors. (G) Gene expression scores of MAPK, JAK/STAT3, and PI3K/AKT/mTOR signaling pathways activation in Tet-tdT-ALK cells after induction of STRN-ALK, its withdrawal, and treatment with ALK inhibitors. Box plots with median line. Whiskers represent SD. Kruskal–Wallis with uncorrected Dunn's test. DEG, differentially expressed genes; FDR, false discovery rate; PCA, principal component analysis; TDSs, thyroid differentiation scores.

Differentially expressed genes (DEG) analysis revealed that the STRN-ALK-expressing untreated cells had the highest number of DEG (n = 753) compared with uninduced cells, followed by cells after Dox withdrawal (n = 246), treatment with crizotinib (n = 71) and ceritinib (n = 22) (Fig. 3B; Supplementary Fig. S4A and Supplementary Table S3). Hierarchical clustering of top-ranked highly expressed DEG between STRN-ALK-expressing and uninduced cells showed that 2 out of 3 clones treated with ceritinib were clustered together with uninduced cells (Fig. 3C).

This group of DEG included the upregulated genes of EMT and cancer cells invasion (Vim, Cd44, Hif1a, Fn1), and downregulated thyroid-specific genes (Dio1, Slc5a5, Foxe1). Other less-ranked DEG involved in thyroid differentiation and iodine metabolism/transport were Duox1/2, Duoxa2, Glis3, Slc26a4, Slc5a8, and Tg (Fig. 3D). Although Pax8 and Nkx2-1 expression change by STRN-ALK did not reach statistical significance, the STRN-ALK signaling-dependent loss of nuclear PAX8 and patchy decrease of NKX2-1 (TTF-1) expression was revealed by IFA (Supplementary Fig. S4B). With the exception of Slc26a4, which exhibited a strong upregulation after STRN-ALK expression, the majority of thyroid differentiation genes were downregulated.

TDSs showed significant decrease after induction of STRN-ALK (Fig. 3E). These scores were restored to the normal levels after Dox withdrawal and treatment with ALK inhibitors, particularly ceritinib. To exclude that change in thyroid differentiation genes expression was due to the direct Dox effects, transcriptomes of WT cells with and without Dox treatment were also analyzed, and no significant difference in thyroid differentiation genes expression was revealed (Supplementary Fig. S2C and Supplementary Table S3).

We next performed the gene sets analysis to examine pathways involved in ALK signaling. Among significantly enriched signaling pathways, MAPK and JAK/STAT were identified as main players (Fig. 3F). Further analysis using established ERK score and pancancer MAPK activity score confirmed significant activation of the MAPK pathway after induction of STRN-ALK, and its suppression after STRN-ALK withdrawal and treatment with ALK inhibitors. Testing of the gene sets involved in JAK/STAT signaling axis revealed significant changes in IL6/JAK/STAT3 and STAT3 activation scores.

Both Dox withdrawal and treatment with ALK inhibitors decreased JAK/STAT3 signaling output, although it was not completely inhibited to the levels of uninduced cells after STRN-ALK withdrawal or treatment with crizotinib (Fig. 3G). These findings were confirmed by upstream gene regulator analysis, where STAT1 and STAT3 were listed in top three transcriptional factors activated in response to ALK signaling (Supplementary Fig. S4C).

In concordance with gene set analysis data, Western blot analysis showed that ceritinib treatment completely inhibited the STRN-ALK-induced activation of STAT3 (Supplementary Fig. S5A). Among other studied gene set enrichment scores, only scores representing PI3K/AKT/mTOR signaling pathway showed some degree of activation by STRN-ALK and downregulation after Dox withdrawal or treatment with ALK inhibitors. However, the range of changes was not very high as for MAPK and JAK/STAT3 (Fig. 3G).

Thus, expression of STRN-ALK in thyroid cells resulted in significant changes in transcriptional profiles, including downregulation of thyroid differentiation and iodine metabolism/transport genes such as Nis, which could be reversed after STRN-ALK removal or application of ALK inhibitors. MAPK, JAK/STAT3, and, to less extent, PI3K/AKT/mTOR were identified as main signaling pathways activated by ALK fusion in thyrocytes.

Application of combination therapy for restoration of NIS in thyroid cells expressing ALK fusion

The mentioned experiments demonstrated that ALK inhibitors had strong effect on NIS and RAI uptake restoration in thyrocytes expressing STRN-ALK. Taking into account that ALK fusion-positive cancers are known to develop resistance to ALK inhibitors,^25

–28 we investigated whether alternative ways of blocking the ALK downstream signaling can be applied for redifferentiation of STRN-ALK-expressing thyrocytes. For that, we tested the efficiency of NIS re-expression after MAPK and JAK/STAT3 axes inhibition.

These pathways were selected based on findings using current in vitro model and observations from STRN-ALK-driven murine thyroid cancers where NIS loss was restricted to histologically dedifferentiated zones with high levels of ERK and STAT3 phosphorylation (Fig. 1C; Supplementary Fig. S5B). Treatment of STRN-ALK-expressing cells with MEK inhibitor selumetinib or JAK inhibitor ruxolitinib resulted in dose-dependent downregulation of both pERK and pSTAT3. At the highest tested dose, selumetinib showed an 87% restoration of NIS expression, while ruxolitinib showed a moderate effect of 45% (Fig. 4A, E).

FIG. 4.

Effects of drugs combination on restoration of NIS in STRN-ALK expressing thyroid cells. NIS levels and MAPK, PI3K/AKT/mTOR, and STAT3 signaling in Tet-tdT-ALK clone 8 cells treated with ceritinib, selumetinib, and ruxolitinib alone (A) or their combinations (B–D). (E) Dot blot quantification results of NIS expression after treatment with selumetinib and ruxolitinib or their combinations. Horizontal dashed line shows basic level of NIS in not induced (green) and induced cells (red). Bars represent mean ± SD. Assay was performed three times. (F) Representative microphotographs showing restoration of the membranous NIS in Tet-tdT-ALK clone 7 after treatment with ceritinib, selumetinib, or combination of selumetinib plus ruxolitinib. (G) Putative scheme of the STRN-ALK signaling and effects of ALK, MEK, and JAK1/2 inhibitors on NIS expression in thyroid cells. CRT, ceritinib; RUX, ruxolitinib; SEL, selumetinib.

Compared with ceritinib, which decreased phosphorylation of all ALK signaling effectors (ERK, STAT3, AKT, and S6), selumetinib treatment in a dose-dependent manner increased the levels of phosphorylated AKT, and likely activated upstream regulators of MEK (manifested as pMEK accumulation), while ERK phosphorylation was fully blocked. Ruxolitinib showed a moderated inhibition of AKT phosphorylation and no effect on mTOR activation marker pS6.

Next, we checked whether combination of ALK inhibitor ceritinib with selumetinib or ruxolitinib, and combination of MEK and JAK1/2 inhibitors may have any cumulative effects on the restoration of NIS expression. Application of ceritinib with selumetinib or ruxolitinib did not reveal any cumulative effects, when at least 20% increase in NIS restoration can be achieved for each drug (Fig. 4B, C; Supplementary Fig. S6). However, the combination of selumetinib with ruxolitinib allowed minimizing the effective doses of selumetinib necessary for NIS re-expression, and completely inhibited all STRN-ALK-activated signaling pathways, including PI3K/AKT/mTOR (Fig. 4D, E).

To further test whether combination of ruxolitinib and selumetinib at lower dose results in restoration of NIS expression on plasma membrane, we have performed IFA for NIS using confocal microscopy (Fig. 4F). In concordance with our previous findings, ceritinib treatment resulted in strong restoration of the membranous NIS. Similarly, selumetinib at the highest tested dose, as well as low dose when combined with ruxolitinib, showed an accumulation of NIS on the plasma membrane of the induced Tet-tdT-ALK cells.

Therefore, both ALK inhibitor ceritinib and MEK inhibitor selumetinib demonstrated high redifferentiation efficiency, which manifested in restoration of membranous NIS in STRN-ALK-expressing thyroid cells. The combination of selumetinib with ruxolitinib had a cumulative effect on complete ALK signaling blockage, NIS re-expression, and redistribution to the plasma membrane (Fig. 4G).

Discussion

In this study, we showed that downregulation of NIS occurs early during ALK-driven thyroid cancer progression and that ALK fusion-induced downregulation of NIS and RAI uptake in thyroid cells can be reversed by inhibition of ALK and its downstream effectors.

ALK fusion-positive thyroid carcinomas have a propensity for dedifferentiation^10

–15 that finally manifests by loss of avidity to RAI.^18

–21 Our finding showed that downregulation of NIS, required for RAI uptake, occurs as early as at the stage of WDTC in ALK fusion-driven carcinogenesis. Decreased transcriptional levels of Nis in response to ALK signaling may be explained in part by decline in the expression of thyroid-specific transcriptional factors genes Foxe1 and Pax8. ^29
–31 While we observed decrease of nuclear TTF-1 expression in some of the STRN-ALK-expressing thyroid cells, its role in ALK-signaling-mediated downregulation of Nis remains unclear.

Besides suppression of Nis, we also detected a significant downregulation of iodine metabolism genes, such as Duox1/2, Duoxa2, Dio1, and Tg. These findings suggest that expression of ALK fusion results in functional dedifferentiation of thyrocytes affecting both the uptake and organification of the iodine. Although due to the nonpolarized nature of PCCL3 cells we were not able to experimentally test the iodine organification, we directly demonstrated that ALK signaling in thyrocytes leads to a significant decrease of the NIS-mediated RAI uptake.

The analysis of the STRN-ALK-induced transcriptional changes and signaling pathways revealed the ALK fusion-dependent activation of the MAPK, JAK/STAT3, and PIK3/AKT/mTOR pathways in thyroid cells. Our findings on MAPK-associated suppression of thyroid differentiation genes in thyroid cells expressing ALK fusion are consistent with observations in human¹¹ and murine thyroid cancers¹ driven by other oncogenes such as BRAF^V600E .

In addition to studies on ALK fusions in lymphoma and lung cancer,^32
–34 our findings showed that JAK/STAT3 signaling is one of the main pathways activated downstream of ALK in thyrocytes. Although PIK3/AKT/mTOR was shown to play role in regulation of NIS expression in vitro,³⁵ and we have identified an increased activity of PIK3/AKT/mTOR pathway in STRN-ALK expressing cells, we cannot provide strong evidence that this pathway is crucial in thyroid cells dedifferentiation, particularly NIS downregulation, in the context of ALK activation.

The results of this study raised a possibility that direct or downstream targeting of ALK signaling may be effective for redifferentiation and RAI resensitization of ALK fusion-positive thyroid cancers. To date, the redifferentiation of RAI-R thyroid carcinomas by targeted therapy was studied mostly for BRAF^V600E- and RAS-mutant thyroid tumors.^3

–7 While single cases of ALK fusion-positive RAI-R metastatic thyroid cancers with structural response to the treatment with ALK inhibitors were reported,^18

–21 the effect of inhibition of ALK on thyroid functional differentiation and RAI avidity remained unknown.

In this study, we directly showed that ALK inhibitors crizotinib and, with higher efficiency, ceritinib completely inhibited ALK signaling, reversed the ALK-fusion-induced dedifferentiation and restored the membranous NIS and RAI uptake in thyroid cells expressing ALK fusion. Moreover, our study established that blocking of downstream ALK signaling with MEK inhibitor selumetinib also restored the expression levels and membranous localization of NIS. At selected doses, JAK1/JAK2 inhibitor ruxolitinib, cornerstone in myelofibrosis therapy,³⁶ exhibited a partial restoration of NIS in STRN-ALK-expressing cells, what might be related to the preserved activity of mTOR.^37,38

Furthermore, we revealed a cumulative effect of ruxolitinib with selumetinib, manifested as a complementary blockage of the ALK downstream signaling and restoration of NIS with its membranous redistribution, even at lower doses of selumetinib. Our results suggest that besides ALK inhibitors, MEK inhibitors or combination of MEK and JAK1/JAK2 inhibitors may be potentially exploited for enhancement of redifferentiation and restoration of RAI avidity in ALK fusion-positive thyroid cancers.

In summary, our study reports that ALK fusion-driven thyroid carcinogenesis is associated with early functional thyroid tumor dedifferentiation. We directly show that STRN-ALK expression in thyroid cells induces the transcriptional suppression and loss of the membranous NIS and diminishes RAI uptake, which can be reversed with ALK inhibitors. Moreover, our study established that blocking of downstream ALK signaling through application of MEK inhibitor selumetinib, as well as its combination at lower doses with JAK1/2 inhibitor ruxolitinib, can restore the expression levels and membranous localization of NIS in thyroid cells expressing ALK fusion.

Footnotes

Acknowledgments

We thank Dr. Alexander Sorkin (Department of Cell Biology, University of Pittsburgh, Pittsburgh, PA, USA) for the help with RAI uptake measurement. We thank Dr. Daniel Altschuler (Department of Pharmacology and Chemical Biology, University of Pittsburgh, Pittsburgh, PA, USA) for providing the PCCL3 cells, for technical support, and comments on experimental design.

Authors' Contributions

A.V.N contributed to conceptualization, methodology, investigation, data curation, software, formal analysis, visualization, writing—original draft, and project administration. V.C. and S.S.D. were involved in investigation. Y.E.N. was in charge of conceptualization, supervision, writing—review and editing, and funding acquisition.

Author Disclosure Statement

The authors declare no competing financial interests directly related to this study, while Y.E.N. holds intellectual property rights related to ThyroSeq test and receives royalty related to ThyroSeq from the University of Pittsburgh; he serves as a consultant for Sonic Healthcare USA.

Funding Information

This study was fully supported by the National Institutes of Health grant CA181150. This research was supported in part by the University of Pittsburgh Center for Research Computing through the resources provided (Award No. OAC-2117681).

Supplementary Material

Supplementary Data

Supplementary Figure S1

Supplementary Figure S2

Supplementary Figure S3

Supplementary Figure S4

Supplementary Figure S5

Supplementary Figure S6

Supplementary Table S1

Supplementary Table S2

Supplementary Table S3

References

Chakravarty

, Santos

, Ryder

, et al. Small-molecule MAPK inhibitors restore radioiodine incorporation in mouse thyroid cancers with conditional BRAF activation. J Clin Invest, 2011; 121(12):4700–4711; doi: 10.1172/JCI46382

Nagarajah

, Le

, Knauf

, et al. Sustained ERK inhibition maximizes responses of BrafV600E thyroid cancers to radioiodine. J Clin Invest, 2016; 126(11):4119–4124; doi: 10.1172/JCI89067

, Grewal

, Leboeuf

, et al. Selumetinib-enhanced radioiodine uptake in advanced thyroid cancer. N Engl J Med, 2013; 368(7):623–632; doi: 10.1056/NEJMoa1209288

Rothenberg

, McFadden

, Palmer

, et al. Redifferentiation of iodine-refractory BRAF V600E-mutant metastatic papillary thyroid cancer with dabrafenib. Clin Cancer Res, 2015; 21(5):1028–1035; doi: 10.1158/1078-0432.CCR-14-2915

Jaber

, Waguespack

, Cabanillas

, et al. Targeted therapy in advanced thyroid cancer to resensitize tumors to radioactive iodine. J Clin Endocrinol Metab, 2018; 103(10):3698–3705; doi: 10.1210/jc.2018-00612

Dunn

, Sherman

, Baxi

, et al. Vemurafenib redifferentiation of BRAF mutant, RAI-refractory thyroid cancers. J Clin Endocrinol Metab, 2019; 104(5):1417–1428; doi: 10.1210/jc.2018-01478

Iravani

, Solomon

, Pattison

, et al. Mitogen-activated protein kinase pathway inhibition for redifferentiation of radioiodine refractory differentiated thyroid cancer: An evolving protocol. Thyroid, 2019; 29(11):1634–1645; doi: 10.1089/thy.2019.0143

Groussin

, Clerc

, Huillard

. Larotrectinib-enhanced radioactive iodine uptake in advanced thyroid cancer. N Engl J Med, 2020; 383(17):1686–1687; doi: 10.1056/NEJMc2023094

Lee

, Lee

, Im

, et al. NTRK and RET fusion-directed therapy in pediatric thyroid cancer yields a tumor response and radioiodine uptake. J Clin Invest, 2021; 131(18):e144847; doi: 10.1172/JCI144847

10.

Landa

, Ibrahimpasic

, Boucai

, et al. Genomic and transcriptomic hallmarks of poorly differentiated and anaplastic thyroid cancers. J Clin Invest, 2016; 126(3):1052–1066; doi: 10.1172/JCI85271

11.

Cancer Genome Atlas Research Network. Integrated genomic characterization of papillary thyroid carcinoma. Cell, 2014; 159(3):676–690; doi: 10.1016/j.cell.2014.09.050

12.

Kelly

, Barila

, Liu

, et al. Identification of the transforming STRN-ALK fusion as a potential therapeutic target in the aggressive forms of thyroid cancer. Proc Natl Acad Sci U S A, 2014; 111(11):4233–4238; doi: 10.1073/pnas.1321937111

13.

Chou

, Fraser

, Toon

, et al. A detailed clinicopathologic study of ALK-translocated papillary thyroid carcinoma. Am J Surg Pathol, 2015; 39(5):652–659; doi: 10.1097/PAS.0000000000000368

14.

Bastos

, de Jesus

, Cerutti

. ETV6-NTRK3 and STRN-ALK kinase fusions are recurrent events in papillary thyroid cancer of adult population. Eur J Endocrinol, 2018; 178(1):83–91; doi: 10.1530/EJE-17-0499

15.

Panebianco

, Nikitski

, Nikiforova

, et al. Characterization of thyroid cancer driven by known and novel ALK fusions. Endocr Relat Cancer, 2019; 26(11):803–814; doi: 10.1530/ERC-19-0325

16.

Nikitski

, Rominski

, Condello

, et al. Mouse model of thyroid cancer progression and dedifferentiation driven by STRN-ALK expression and loss of p53: Evidence for the existence of two types of poorly differentiated carcinoma. Thyroid, 2019; 29(10):1425–1437; doi: 10.1089/thy.2019.0284

17.

Elsayed

, Christopoulos

. Therapeutic sequencing in ALK(+) NSCLC. Pharmaceuticals (Basel), 2021; 14(2):80; doi: 10.3390/ph14020080

18.

Godbert

, Henriques de Figueiredo

, Bonichon

, et al. Remarkable response to crizotinib in woman with anaplastic lymphoma kinase-rearranged anaplastic thyroid carcinoma. J Clin Oncol, 2015; 33(20):e84–e87; doi: 10.1200/JCO.2013.49.6596

19.

Leroy

, Bonhomme

, Le Moulec

, et al. Remarkable response to ceritinib and brigatinib in an anaplastic lymphoma kinase-rearranged anaplastic thyroid carcinoma previously treated with crizotinib. Thyroid, 2020; 30(2):343–344; doi: 10.1089/thy.2019.0202

20.

Aydemirli

, van Eendenburg

JDH

, van Wezel

, et al. Targeting EML4-ALK gene fusion variant 3 in thyroid cancer. Endocr Relat Cancer, 2021; 28(6):377–389; doi: 10.1530/ERC-20-0436

21.

de Salins

, Loganadane

, Joly

, et al. Complete response in anaplastic lymphoma kinase-rearranged oncocytic thyroid cancer: A case report and review of literature. World J Clin Oncol, 2020; 11(7):495–503; doi: 10.5306/wjco.v11.i7.495

22.

Subramanian

, Tamayo

, Mootha

, et al. Gene set enrichment analysis: A knowledge-based approach for interpreting genome-wide expression profiles. Proc Natl Acad Sci U S A, 2005; 102(43):15545–15550; doi: 10.1073/pnas.0506580102

23.

Mootha

, Lindgren

, Eriksson

K-F

, et al. PGC-1α-responsive genes involved in oxidative phosphorylation are coordinately downregulated in human diabetes. Nat Genet, 2003; 34(3):267–273; doi: 10.1038/ng1180

24.

Krämer

, Green

, Pollard

Jr.,

et al. Causal analysis approaches in Ingenuity Pathway Analysis. Bioinformatics, 2014; 30(4):523–530; doi: 10.1093/bioinformatics/btt703

25.

Katayama

, Shaw

, Khan

, et al. Mechanisms of acquired crizotinib resistance in ALK-rearranged lung cancers. Sci Transl Med, 2012; 4(120):120ra117; doi: 10.1126/scitranslmed.3003316

26.

Lovly

, McDonald

, Chen

, et al. Rationale for co-targeting IGF-1R and ALK in ALK fusion-positive lung cancer. Nat Med, 2014; 20(9):1027–1034; doi: 10.1038/nm.3667

27.

Crystal

, Shaw

, Sequist

, et al. Patient-derived models of acquired resistance can identify effective drug combinations for cancer. Science, 2014; 346(6216):1480–1486; doi: 10.1126/science.1254721

28.

Doebele

, Pilling

, Aisner

, et al. Mechanisms of resistance to crizotinib in patients with ALK gene rearranged non-small cell lung cancer. Clin Cancer Res, 2012; 18(5):1472–1482; doi: 10.1158/1078-0432.CCR-11-2906

29.

Fernández

, López-Márquez

, Martínez

, et al. New insights into FoxE1 functions: Identification of direct FoxE1 targets in thyroid cells. PLoS One, 2013; 8(5):e62849; doi: 10.1371/journal.pone.0062849

30.

Ohno

, Zannini

, Levy

, et al. The paired-domain transcription factor Pax8 binds to the upstream enhancer of the rat sodium/iodide symporter gene and participates in both thyroid-specific and cyclic-AMP-dependent transcription. Mol Cell Biol, 1999; 19(3):2051–2060; doi: 10.1128/MCB.19.3.2051

31.

Taki

, Kogai

, Kanamoto

, et al. A thyroid-specific far-upstream enhancer in the human sodium/iodide symporter gene requires Pax-8 binding and cyclic adenosine 3',5'-monophosphate response element-like sequence binding proteins for full activity and is differentially regulated in normal and thyroid cancer cells. Mol Endocrinol, 2002; 16(10):2266–2282; doi: 10.1210/me.2002-0109

32.

Zhang

, Raghunath

, Xue

, et al. Multilevel dysregulation of STAT3 activation in anaplastic lymphoma kinase-positive T/null-cell lymphoma. J Immunol, 2002; 168(1):466–474; doi: 10.4049/jimmunol.168.1.466

33.

Yanagimura

, Takeuchi

, Fukuda

, et al. STAT3 inhibition suppresses adaptive survival of ALK-rearranged lung cancer cells through transcriptional modulation of apoptosis. NPJ Precis Oncol, 2022; 6(1):11; doi: 10.1038/s41698-022-00254-y

34.

Tabbo

, Barreca

, Piva

, et al. ALK signaling and target therapy in anaplastic large cell lymphoma. Front Oncol, 2012; 2:41; doi: 10.3389/fonc.2012.00041

35.

Kogai

, Sajid-Crockett

, Newmarch

, et al. Phosphoinositide-3-kinase inhibition induces sodium/iodide symporter expression in rat thyroid cells and human papillary thyroid cancer cells. J Endocrinol, 2008; 199(2):243–252; doi: 10.1677/JOE-08-0333

36.

Mascarenhas

, Hoffman

. Ruxolitinib: The first FDA approved therapy for the treatment of myelofibrosis. Clin Cancer Res, 2012; 18(11):3008–3014; doi: 10.1158/1078-0432.CCR-11-3145

37.

Tavares

, Eloy

, Melo

, et al. mTOR pathway in papillary thyroid carcinoma: different contributions of mTORC1 and mTORC2 complexes for tumor behavior and SLC5A5 mRNA expression. Int J Mol Sci, 2018; 19(5):1448; doi: 10.3390/ijms19051448

38.

Plantinga

, Heinhuis

, Gerrits

, et al. mTOR Inhibition promotes TTF1-dependent redifferentiation and restores iodine uptake in thyroid carcinoma cell lines. J Clin Endocrinol Metab, 2014; 99(7):E1368–E1375; doi: 10.1210/jc.2014-1171