DDR1 regulates thyroid cancer cell differentiation via IGF-2/IR-A autocrine signaling loop

in Endocrine-Related Cancer
Correspondence should be addressed to R Malaguarnera or A Belfiore: malaguarnera@unicz.it or antonino.belfiore@unict.it

*(V Vella and M L Nicolosi contributed equally to this work)

Restricted access

Patients with thyroid cancers refractory to radioiodine (RAI) treatment show a limited response to various therapeutic options and a low survival rate. The recent use of multikinase inhibitors has also met limited success. An alternative approach relies on drugs that induce cell differentiation, as the ensuing increased expression of the cotransporter for sodium and iodine (NIS) may partially restore sensitivity to radioiodine. The inhibition of the ERK1/2 pathway has shown some efficacy in this context. Aggressive thyroid tumors overexpress the isoform-A of the insulin receptor (IR-A) and its ligand IGF-2; this IGF-2/IR-A loop is associated with de-differentiation and stem-like phenotype, resembling RAI-refractory tumors. Importantly, IR-A has been shown to be positively modulated by the non-integrin collagen receptor DDR1 in human breast cancer. Using undifferentiated human thyroid cancer cells, we now evaluated the effects of DDR1 on IGF-2/IR-A loop and on markers of cell differentiation and stemness. DDR1 silencing or downregulation caused significant reduction of IR-A and IGF-2 expression, and concomitant increased levels of differentiation markers (NIS, Tg, TSH, TPO). Conversely, markers of epithelial-to-mesenchymal transition (Vimentin, Snail-2, Zeb1, Zeb2 and N-Cadherin) and stemness (OCT-4, SOX-2, ABCG2 and Nanog) decreased. These effects were collagen independent. In contrast, overexpression of either DDR1 or its kinase-inactive variant K618A DDR1-induced changes suggestive of less differentiated and stem-like phenotype. Collagen stimulation was uneffective. In conclusion, in poorly differentiated thyroid cancer, DDR1 silencing or downregulation blocks the IGF-2/IR-A autocrine loop and induces cellular differentiation. These results may open novel therapeutic approaches for thyroid cancer.

Downloadable materials

  • Table 1S. Histotype and molecular characterization of the human thyroid cancer cell lines studied.
  • Supplemental Figure 1: Effect of collagen and of KD mutant K618A DDR1 on IR expression (A) Relative quantification of DDR1 and IR mRNAs in cells stimulated by collagen. SW1736 and 8505C cells were transiently transfected with a pool of four scramble siRNAs or a pool of four siRNA oligos against DDR1 and stimulated with collagen IV (10µg/ml) for 48h. qRT-PCR analysis for DDR1 and IR mRNA expression were performed as described in Methods.
  • Supplemental Figure 2: Effect of collagen and of KD mutant K618A DDR1 on IGF-2 expression. (A-B) Relative quantification of IGF-2 mRNA. SW1736 and 8505C cells were transiently transfected with a pool of four scramble siRNAs or a pool of four siRNA oligos against DDR1 and stimulated with collagen IV (10µg/ml) for 48h (A). Alternatively, cells were transiently transfected with either a constitutive empty (pCMV6-EV) or the human DDR1 (pCMV6-DDR1) or the mutant K618A DDR1 expressing vectors (B). qRT-PCR analysis for DDR1 and IR mRNA expression were performed as described in Methods. Cells were analyzed for IGF-2 mRNA expression by qRT-PCR analysis. (A-B) ns, p > 0.05; *0.01 < p < 0.05; **0.001 < p < 0.01; ***p < 0.001; statistical significance was calculated using one-way ANOVA followed by Bonferroni test.
  • Supplemental Figure 3: Expression of differentiation and stemness markers in cells silenced for DDR1 and stimulated with collagen. (A-B) Relative quantification of differentiation and stemness markers. SW1736 and 8505C cells were transiently transfected with a pool of four scramble siRNAs or a pool of four siRNA oligos against DDR1 and stimulated with collagen IV (10µg/ml) for 48h. Cells were analyzed for differentiation (A) and stemness (B) markers by qRT-PCR analysis. (A-B) ns, p > 0.05; *0.01 < p < 0.05; **0.001 < p < 0.01; ***p < 0.001; statistical significance was calculated using one-way ANOVA followed by Bonferroni test.
  • Supplemental Figure 4: Expression of differentiation and stemness markers in cells expressing the KD mutant K618A DDR1. (A-B) Relative quantification of differentiation and stemness markers. SW1736 and 8505C cells were transiently transfected with either a constitutive empty (pCMV6-EV) or the human DDR1 (pCMV6-DDR1) or the mutant K618A DDR1 expressing vectors. Cells were analyzed for differentiation (A) and stemness (B) markers by qRT-PCR analysis. (A-B) *0.01 < p < 0.05; **0.001 < p < 0.01; ***p < 0.001; statistical significance was calculated using one-way ANOVA followed by Bonferroni test.
  • Supplemental Figure 5: Collagen stimulation does not affect IR, IGF-2 and markers of differentiation, EMT and stemness (A-C) Effect of collagen stimulation on the expression of IR, IGF-2, and markers of thyroid-specific differentiation, EMT, and stemness. SW1736 and 8505C cells were stimulated with collagen IV (10µg/ml) for 48h. Cells were analyzed for DDR1, IR, and IGF-2 (A), differentiation (B) and stemness (C) markers by qRT-PCR analysis. (A-C) ns, p > 0.05.
  • Supplemental Figure 6: DDR1 inhibitors affect the IR-A/IGF-2 loop in thyroid cancer cells. (A) DDR1 and IR expression in cells treated with DDR-IN-1 and 7RH. SW1736 and 8505C cells were treated with DDR1 inhibitors DDR-IN-1 and 7RH for 72h at the indicated concentrations, and DDR1 and IR expression were evaluated by western blot analysis. Blots are representative of three independent experiments. The histograms represent the mean±SEM of densitometric analysis of three independent experiments after normalization against tubulin. (B) Relative quantification of DDR1, IR, and IGF-2 mRNAs in cells treated with DDR1 inhibitors. SW1736 and 8505C cells were treated with DDR1 inhibitors DDR-IN-1 and 7RH for 72h at the indicated concentrations, and DDR1, IR, IGF- mRNA levels were evaluated by qRT-PCR analysis. Values were normalized using human β-actin as housekeeping control gene. (A-B) *0.01 < p < 0.05; **0.001 < p < 0.01; ***p < 0.001; statistical significance was calculated using one-way ANOVA followed by Bonferroni test.
  • Supplemental Figure 7: Collagen does not influence the effects of DDR1 inhibitors. (A-C) Effect of collagen stimulation in cells treated with the combination of DDR-IN-1 and NVP-AEW541. SW1736 and 8505C cells were treated with inhibitors of DDR1 (DDR-IN-1) and IR/IGF-1R (AEW541) at the indicated concentrations, with or without stimulation with collagen IV (10µg/ml), for 48h. Cells were analyzed for DDR1, IR, and IGF-2 (A), differentiation (B) and stemness (C) markers by qRT-PCR analysis. (A-C) *0.01 < p < 0.05; **0.001 < p < 0.01; ***p < 0.001; statistical significance was calculated using one-way ANOVA followed by Bonferroni test.
  • Supplemental Figure 8: IGF-2 induces DDR1 upregulation in thyroid cancer cells. (A-B) Dose-response of DDR1 expression after IGF-2. 8505C and SW1736 cells were starved for 24 h, and then stimulated with IGF-2 at the indicated doses for 24h. Western blot for DDR1 protein expression (a) and qRT-PCR analysis (b) for DDR1 mRNA expression were performed as described in Methods. Each blot is representative of three independent experiments. Values shown in graphs are mean ± SEM of three separate experiments. (A-B) ns, p > 0.05; *0.01 < p < 0.05; **0.001 < p < 0.01; ***p < 0.001; statistical significance was calculated using one-way ANOVA followed by Bonferroni test.

 

      Society for Endocrinology

Related Articles

Article Information

Metrics

All Time Past Year Past 30 Days
Abstract Views 509 509 58
Full Text Views 307 307 2
PDF Downloads 37 37 0

Altmetrics

Figures

  • View in gallery

    DDR1 and IR expression in thyroid cells and human thyroid tissue specimens. (A) DDR1 and IR protein expression in various cells. Human thyroid cancer cell lines (SW1736, C643, Hth74, FF1, 8505C, BcPap, TPC1, 4A1) and normal primary cell cultures (NPC) were analyzed by western immunoblot for DDR1 and IR expression using specific polyclonal antibodies, as indicated. Tubulin antibody was used as control for protein loading. A representative blot of three independent experiments is shown. Histograms represent the mean ± s.e.m. of DDR1 and IR densitometric analysis after normalization against tubulin. (B) DDR1 and IR protein expression in human thyroid specimens. Specimens from human thyroid cancer (C) and the contralateral normal thyroid tissue (N) were analyzed for DDR1 and IR expression levels by western immunoblot, as described in (A). A representative blot of three independent experiments is shown. Histograms represent the mean ± s.e.m. of densitometric analysis after normalization against tubulin. (C) qRT-PCR analysis of DDR1 and IR mRNA in cell lines and NPCs. Human DDR1 and IR mRNA levels were evaluated in all human cell lines shown in (A). NPCs is a mean of five specimens. Normalization was done using human β-ACTIN as housekeeping control gene. Data are presented as the mean ± s.e.m. (error bars) from three independent experiments. (D) qRT-PCR analysis of DDR1 and IR mRNA in human thyroid specimens. Human DDR1 and IR mRNA levels were evaluated in all specimens shown in (B). Normalization was done using human β-ACTIN as housekeeping control gene. Data are presented as the mean ± s.e.m. (error bars) from three independent experiments. (A, B, C and D) ns, P > 0.05; *0.01 < P < 0.05; **0.001 < P < 0.01; ***P < 0.001; statistical significance was calculated using one-way ANOVA followed by Bonferroni test.

  • View in gallery

    mRNA expression levels of IR isoforms in cultured thyroid cells and human thyroid specimens. (A) IR isoform transcripts in cell lines and NPCs. IR isoform (IR-A and IR-B) transcripts were obtained from thyroid cancer cell lines (SW1736, C643, Hth74, FF1, 8505C, BcPap, TPC1, 4A1) and normal primary cultures from five patients (NPCs). mRNA from IGF-1R-deficient mouse fibroblasts transfected with the human IR-A (R-/IR-A) and IR-B (R-/IR-B), expressing 50% of each IR isoforms, were used as positive controls (IR-A/IR-B). Products of PCR amplification were resolved on a 2.5% agarose gel. Images of the PCR product from IR-B (Ex+11) and IR-A (Ex−11) are 167 and 131 bp, respectively. Graphical representation of PCR analysis indicates the percentage of IR-A mRNA. The percentage of IR-A mRNA was calculated as follows: densitometric value of band IR-A/densitometric value of IR-A + IR-B bands. Scanning densitometry was performed using ImageJ software. All results are expressed as means ± s.e. of three independent experiments. (B) IR isoforms transcripts in human thyroid specimens. IR isoform (IR-A and IR-B) transcripts were evaluated in cancer tissues and the contralateral normal tissues of six patients. Results are means ± s.e. of three independent experiments. *0.01 < P < 0.05; **0.001 < P < 0.01; ***P < 0.001; statistical significance was calculated using one-way ANOVA followed by Bonferroni test.

  • View in gallery

    IGFs expression in cultured cells and human specimens. (A and B) mRNA expression of IGF-2 and IGF-1 in thyroid cancer cell lines. PCR analysis of IGF-2 and IGF-1 mRNA were evaluated in SW1736, C643, Hth74, FF1, 8505C, BcPap, TPC1, 4A1 thyroid cells. MCF7/IGF-2 and SK-N-MC cells were used as positive control for IGF-2 and IGF-1, respectively. S9 was used as housekeeping control gene. A representative of three independent experiments is shown. (C) IGF-2 detection in conditioned medium of thyroid cells. IGF-2 protein levels were detected, by ELISA (see ‘Methods’ section), in cell conditioned medium after 72 h of incubation in serum-free medium. Diluted conditioned medium (1:7) from MCF7 transfected with IGF-2 was used as positive control (MCF7/IGF2 (1:7)). Data are presented as the mean ± s.e.m. (error bars) from three independent experiments. (D) IGF-2 mRNA expression in human thyroid specimens. qRT-PCR analysis of IGF-2 mRNA was evaluated in cancer tissues of six patients and in contralateral normal tissues as control. Normalization was done using human β-ACTIN as housekeeping control gene. Data are presented as the mean ± s.e.m. (error bars) from three independent experiments. **0.001 < P < 0.01; ***P < 0.001; statistical significance was calculated using one-way ANOVA followed by Bonferroni test.

  • View in gallery

    DDR1 affects IR expression levels. (A) Western blot after DDR1 silencing. SW1736, 8505C, Hth74 and TPC1 thyroid cancer cell lines were transiently transfected with either a pool of four scramble siRNAs or a pool of four siRNA oligos against DDR1. After 48 h, cells were lysed and analyzed by SDS-PAGE and immunoblotted with the indicated primary antibodies. Tubulin was used as control for protein loading. Blot is representative of three independent experiments. Histograms represent the mean ± s.e.m. of DDR1 and IR densitometric analysis after normalization against tubulin. (B) qRT-PCR after DDR1 silencing. In DDR1-silenced cells (siDDR1), IR mRNA levels were evaluated by qRT-PCR analysis and values were normalized using human β-ACTIN as housekeeping control gene. In parallel, DDR1 mRNA was evaluated by qRT-PCR to confirm DDR1 silencing. (C) Western blot after collagen stimulation. SW1736 and 8505C cells were transiently transfected with either a constitutive empty (pCMV6-EV) or a human DDR1 (pCMV6-DDR1) expressing vectors. After 24 h cells were stimulated with 10 µg/mL collagen IV for 48 h. Cells were then lysed and analyzed by SDS-PAGE and immunoblotted with the indicated primary antibodies. Tubulin was used as control for protein loading. Blot is representative of three independent experiments. Histograms represent the mean ± s.e.m. of densitometric values after normalization against tubulin. (D) Western blot after DDR1 overexpression. Thyroid cancer cell lines were transiently transfected with either a constitutive empty (pCMV6-EV) or a human DDR1 (pCMV6-DDR1) expressing vectors. After 48 h, cells were lysed and analyzed by SDS-PAGE and immunoblotted with the indicated primary antibodies. Tubulin was used as control for protein loading. Blot is representative of three independent experiments. Histograms represent the mean ± s.e.m. of densitometric analysis after normalization against tubulin. (E) qRT-PCR after DDR1 overexpression. In DDR1-transfected cell lines, IR mRNA levels were evaluated by qRT-PCR analysis and values were normalized using human β-ACTIN as housekeeping control gene. DDR1 overexpression was confirmed measuring DDR1 mRNA by qRT-PCR. (F) Thyroid cancer cells were transiently transfected with either a constitutive empty (pCMV6-EV) or the human DDR1 (pCMV6-DDR1) or the mutant K618A DDR1 expressing vectors. After 48 h, cells were lysed and analyzed by SDS-PAGE and immunoblotted with the indicated primary antibodies. Tubulin was used as control for protein loading. Blot is representative of three independent experiments. Histograms represent the mean ± s.e.m. of densitometric values after normalization against tubulin. (G) IR isoforms transcripts after DDR1 silencing. IR isoform (IR-A and IR-B) transcripts were evaluated in all four thyroid cancer cell lines after DDR1 silencing, as described in (A). (H) qRT-PCR expression of IGF-2 after DDR1 silencing. IGF-2 transcript was evaluated in six cell lines. Normalization was done using human β-ACTIN as housekeeping control gene. Data are presented as the mean ± s.e.m. (error bars) from three independent experiments. (A, B, C, D, E and F) *0.01 < P < 0.05; **0.001 < P < 0.01; ***P < 0.001; statistical significance was calculated using one-way ANOVA followed by Bonferroni test.

  • View in gallery

    DDR1 silencing or overexpression affects downstream signaling in human thyroid cancer cells. (A) Downstream signaling after DDR1 silencing. SW1736, 8505C, Hth74 and TPC1 thyroid cancer cells were transiently transfected with a pool of four scramble siRNAs or a pool of four siRNA oligos against DDR1. After 48 h cells were lysed and analyzed by SDS-PAGE and immunoblotted with the indicated primary antibodies. Tubulin was used as control for protein loading. The top panels show representative blots of three experiments. The histograms represent the mean ± s.e.m. of densitometric analysis of three independent experiments after normalization of each phospho-protein against the total non-phosphorylated form and then against tubulin. (B) Downstream signaling after DDR1 overexpression. Thyroid cancer cells were transiently transfected with plasmids encoding either an empty (pCMV6-EV) or a human DDR1 cDNA-expressing vector (pCMV6-DDR1). After 48 h the downstream signaling was assessed as in (A). Blots are representative of three independent experiments. The histograms represent the mean ± s.e.m. of densitometric analysis of three independent experiments after normalization of each phosphor proteins the total non-phosphorylated form and then against tubulin. (A and B) *0.01 < P < 0.05; **0.001 < P < 0.01; ***P < 0.001; statistical significance was calculated using one-way ANOVA followed by Bonferroni test.

  • View in gallery

    DDR1 affects IR promoter activity. (A) DDR1 silencing is associated with reduced activity of the IR promoter. 8505C and SW1736 cells were transiently transfected either with a pool of four scramble siRNAs or a pool of four siRNA oligos against DDR1. After 24 h cells were transfected with a luciferase construct containing a fragment of the IR promoter. 48 h later cells were lysed and luciferase activity measured. tGFP was cotransfected for normalization of transfected DNA content. (B) DDR1 silencing modulates IR-dependent transcription factors. In thyroid cancer cells silenced for DDR1, Sp1 and HMGA1 mRNA levels were evaluated by qRT-PCR. (A and B) **0.001 < P < 0.01; ***P < 0.001; statistical significance was calculated using one-way ANOVA followed by Bonferroni test.

  • View in gallery

    DDR1 affects several splicing factors. DDR1 silencing modulates the splicing factors involved in Ex-11 skipping. SW1736, 8505C, Hth74 and TPC1 thyroid cancer cells were transiently transfected with a pool of four scramble or four siRNA oligos against DDR1. After 48 h splicing factors mRNA levels were evaluated by qRT-PCR analysis. Values were normalized using human β-ACTIN as housekeeping control gene. *0.01 < P < 0.05; **0.001 < P < 0.01; ***P < 0.001; statistical significance was calculated using one-way ANOVA followed by Bonferroni test.

  • View in gallery

    DDR1 silencing affects the biological effects of thyroid cancer cells. (A) Cell proliferation after either DDR1 silencing or DDR1 overexpression. Thyroid cancer cells were transiently transfected with a pool of four scramble siRNAs or of four siRNA oligos against DDR1 (left). Alternatively cells were transiently transfected with plasmids encoding either the empty vector (pCMV6-EV) or the human DDR1 cDNA (pCMV6-DDR1) (right). Cell viability was evaluated by MTT assay. Values are expressed as percentages of untreated scramble-transfected cells (basal) and represent the mean ± s.e.m. of three independent experiments in triplicate. (B) Cell invasion after either DDR1 silencing or DDR1 overexpression. For invasion assay, 48 h after transfection cells were removed from plates with 0.01% trypsin and seeded on polycarbonate filters coated with 25 μg/mL fibronectin. Cells were allowed to migrate for 6 h. Values are expressed as percentages of empty vector-transfected cells (basal) and represent the mean ± s.e.m. of three independent experiments in triplicate. (C) Cell proliferation and invasion after collagen stimulation. SW1736 and 8505C cancer cells were transiently transfected with a pool of four scramble siRNAs or of four siRNA oligos against DDR1. After 24 h they were stimulated with collagen IV for 48 h. Cell viability was evaluated by MTT assay (left). For invasion assay, 48 h after transfection cells were removed from plates with 0.01% trypsin and seeded on polycarbonate filters coated with 10 μg/mL collagen IV. Cells were allowed to migrate for 6 h (right). (D) Thyroid-specific differentiation, EMT, and stemness markers in cells silenced for DDR1 expression. After DDR1 silencing cells showed a significant increase of thyroid cell differentiation markers and two key transcription factors (left). DDR1 silencing caused also a significant decrease of several markers associated with EMT (middle) and of cell stemness markers (right). (A, B, C and D) ns, P > 0.05; *0.01 < P < 0.05; **0.001 < P < 0.01; ***P < 0.001; statistical significance was calculated using one-way ANOVA followed by Bonferroni test.

  • View in gallery

    DDR1 inhibitor affects the IR-A/IGF-2 loop in thyroid cancer cells. (A) Western blot in DDR-IN-1 treated cells. SW1736 and 8505C cells were treated with DDR-IN-1 for 72 h at a concentration of 10 µM. DDR1, pDDR1, and IR expression were evaluated by Western blot analysis. Blots are representative of three independent experiments. The histograms represent the mean ± s.e.m. of densitometric analysis of three independent experiments after normalization against tubulin. (B) Relative quantification of DDR1, IR, and IGF-2 mRNAs. IR, and IGF-2 mRNA levels were evaluated by qRT-PCR analysis in thyroid cancer cells treated with DDR1 inhibitor. Values were normalized using human β-ACTIN as housekeeping control gene. (C) IR isoform transcripts. PCR products were resolved in a 2.5% agarose gel. Images of the PCR product from IR-B (Ex-11) and IR-A (Ex-11) are 167 and 131 bp, respectively. (D) Efficacy of the combination of DDR-IN-1 and the dual IR/IGF-1R inhibitor (AEW541). 8505C and SW1736 cells were incubated with the combined treatment for 72 h. The combination of DDR-IN-1 with NVP-AEW541 gave additive effects in both cell proliferation and invasion assays of thyroid cancer cells. (E) Effect of collagen in cells treated with the combination of DDR-IN-1 and NVP-AEW541. SW1736 and 8505C cells were treated with the DDR1 inhibitor DDR-IN-1 and the IR/IGF-1R inhibitor NVP-AEW541 (AEW541) at the indicated concentrations, in the presence or absence of collagen IV (10 µg/mL), for 48 h. Cell viability was evaluated by MTT assay. (A, B, C, D and E) ns, P > 0.05; *0.01 < P < 0.05; **0.001 < P < 0.01; ***P < 0.001; statistical significance was calculated using one-way ANOVA followed by Bonferroni test.

  • View in gallery

    The crosstalk between the IGF-2/IR-A loop and DDR1 is more relevant in thyrospheres from cancer cells. (A) Relative quantification of DDR1, IR, and IGF-2 mRNAs in cell monolayer compared to thyrospheres. RNA levels were evaluated by qRT-PCR analysis and values were normalized using human β-ACTIN as housekeeping control gene. The experiment shown is representative of three independent experiments. Values are mean ± s.e.m. of three separate experiments. (B) IR isoforms transcripts in cell monolayer (M) compared to thyrospheres (S). IR isoform (IR-A and IR-B) transcripts were evaluated in thyrospheres derived from thyroid cancer cells, as described in Methods. (C, D and E) Effect of DDR-IN-1, NVP-AEW541 and vemurafenib on DDR1, IR and IGF-2 expression in thyrospheres. 8505C cells were grown as thyrospheres in the presence or absence of collagen IV (10 µg/mL), and exposed to DDR1-IN-1 (10 µM), NVP-AEW541 (AEW541) (1 µM), and vemurafenib (1 µM) either as single agents or in combination. Cells were analyzed for DDR1, IR and IGF-2 (C), differentiation markers (D), and stemness markers (E) expression. All the effects described in (C), (D), and (E) were not influenced by the presence of collagen. (A, B, C, D and E) *0.01 < P < 0.05; **0.001 < P < 0.01; ***P < 0.001; statistical significance was calculated using one-way ANOVA followed by Bonferroni test.

Cited By

PubMed

Google Scholar