Extracellular vesicle-carried GTF2I from mesenchymal stem cells promotes the expression of tumor-suppressive FAT1 and inhibits stemness maintenance in thyroid carcinoma

Abstract

Through bioinformatics predictions, we identified that GTF2I and FAT1 were downregulated in thyroid carcinoma (TC). Further, Pearson’s correlation coefficient revealed a positive correlation between GTF2I expression and FAT1 expression. Therefore, we selected them for this present study, where the effects of bone marrow mesenchymal stem cell-derived EVs (BMSDs-EVs) enriched with GTF2I were evaluated on the epithelial–to–mesenchymal transition (EMT) and stemness maintenance in TC. The under-expression of GTF2I and FAT1 was validated in TC cell lines. Ectopically expressed GTF2I and FAT1 were found to augment malignant phenotypes of TC cells, EMT, and stemness maintenance. Mechanistic studies revealed that GTF2I bound to the promoter region of FAT1 and consequently upregulated its expression. MSC-EVs could shuttle GTF2I into TPC-1 cells, where GTF2I inhibited TC malignant phenotypes, EMT, and stemness maintenance by increasing the expression of FAT1 and facilitating the FAT1-mediated CDK4/FOXM1 downregulation. In vivo experiments confirmed that silencing of GTF2I accelerated tumor growth in nude mice. Taken together, our work suggests that GTF2I transferred by MSC-EVs confer antioncogenic effects through the FAT1/CDK4/FOXM1 axis and may be used as a promising biomarker for TC treatment.

Keywords

thyroid carcinoma / mesenchymal stem cell / extracellular vesicle / GTF2I / FAT1 / CDK4

1 Introduction

Thyroid carcinoma (TC) is a prevalent malignancy, and its incidence has been increasing worldwide [1,2]. An increasing body of evidence points to the involvement of the epithelial–to–mesenchymal transition (EMT) and the production of cancer stem cells in the resistance, relapse, and distant metastasis of malignant diseases [3–5]. Cancer stemness has been extensively recognized as a pivotal culprit in the pathophysiology of recurrence and metastasis, although the clinical evidence remains to be further studied [6]. Recent knowledge concerning the mechanistic basis of EMT and stemness maintenance may provide new biomarkers for early diagnosis, prognosis, and treatment.

New data are accumulating on the potential of mesenchymal stem/stromal cell-derived extracellular vesicles (MSC-EVs) as alternative, cell-free therapeutic approaches in a variety of cancers [7,8]. They have been suggested to influence the disease course by shuttling such bioactive cargoes as DNA, proteins/peptides, mRNAs, miRNAs, lipids, and organelles to target cells [9,10]. GTF2I, the human gene encoding TFII-I, is a general transcription factor, and its mutation is recurrent in thymic epithelial tumors yet rare in other malignancies [11,12]. GTF2I mutation has been reported to increase the protein expression of EMT-related genes in thymic epithelial cells [13]. Pearson’s correlation coefficient based on bioinformatic data predicts that GTF2I expression may be positively correlated with FAT1 expression, which encodes a protocadherin and is among the most frequently mutated genes in human cancer [14]. Knockdown of FAT1 has been documented to facilitate the proliferative capacities of sporadic medullary TC cells [15]. Moreover, FAT1 deficiency can augment resistance to the CDK4 inhibitor [16], and downregulation of CDK4 suppresses papillary TC cell proliferative, migratory, and invasive phenotypes, as well as tumor growth [17]. Meanwhile, CDK4 inhibition reduces FOXM1 phosphorylation in bladder cancer cells [18]. FOXM1 forms a member of the forkhead box family of transcription factors participating in the regulation of cell proliferative, angiogenic, and invasive potential, whereas knockdown of FOXM1 can impair these malignant phenotypes of anaplastic TC cells [19].

In this work, we hypothesized that GTF2I shuttled by MSC-EVs may repress the malignant characteristics of TC involving the interaction with FAT1, CDK4, and FOXM1. For validation, we isolated EVs from bone marrow mesenchymal stem cells (BMSCs), cocultured them with TC cells, and conducted in vitro and in vivo experiments.

2 Materials and methods

2.1 Ethics statement

The study was conducted under the approval of the Ethics Committee of Huashan Hospital, Fudan University, China (Ethical code: 2019-031). The current study was approved by the Animal Ethics Committee of Huashan Hospital, Fudan University (Ethical code: 2019 JS-102) and performed according to the Guide for the Care and Use of Laboratory Animals published by the US National Institutes of Health.

2.2 In silico prediction

TC-related TCGA data were downloaded from the GEPIA2 database, followed by differential analysis using the R “limma” package to identify the differentially expressed genes (DEGs) with |log₂FC| > 0.5 and adjusted P value (q value) < 0.01 as the threshold. TC-related genes were retrieved from the GeneCards database with “thyroid cancer” as the keyword, and the top 5000 genes in the relevance score were screened. The Phenolyzer database was used to retrieve TC-related genes with “thyroid cancer” as the keyword, and the top 2000 genes in score were screened. The differentially poorly expressed genes were intersected with TC-related genes from the GeneCards and Phenolyzer databases using the jvenn tool to identify the candidate genes. These candidate genes were then subjected to GO and KEGG enrichment analyses using the R “clusterProfiler” package (version 3.14.3). The P value correction method is “BH.” An adjusted P < 0.05 was used as the screening criterion for significant enrichment of the pathway.

RNAseq data and patient survival data in level 3 HTSeq-Count format of the TC project were obtained from the TCGA database. The related genes were calculated according to the expression of a single gene using Pearson’s correlation coefficient (low expression group: 0%–50% and high expression group: 50%–100%), with r > 0.06 and P < 0.05 indicating a significant positive correlation. The expression of genes in TC-related TCGA data (58 normal samples and 510 tumor samples) was tested using the Mann–Whitney U test.

The R “ggplot2” package (version 3.3.3) was used for visualization, and box plots and volcano plots were drawn. The potential binding between transcription factors and target gene promoters was predicted by the hTFtarget database.

2.3 Tissue collection

Cancer tissues and adjacent tissues were collected from 63 patients with TC undergoing surgical procedures in Huashan Hospital, Fudan University, from June 2019 to December 2020. All patients had complete clinicopathological data. The specimens were diagnosed as TC, and the patients had not received any anti-cancer treatment before surgery [20]. Those who had received non-surgical treatment, such as chemotherapy, radiotherapy, molecular targeted therapy, or other thyroid disease treatments before surgery, were excluded from this study [21].

2.4 Cell culture

Human normal thyroid cell line Nthy-ori 3-1 and TC cell lines TPC-1 (CBP60257) and CAL62 (CBP60709) were purchased from COBIOER BIOSCIENCES CO., LTD. (Nanjing, Jiangsu, China), 8305C cell line (CL-0613) was procured from Procell Life (Wuhan, China), and IHH4 cell line (CBP61201) was acquired from COBIOER. These cells were cultured in DMEM (PM150210, Procell Life, Wuhan, China) with 10% FBS and 1% penicillin–streptomycin (PB180120, Procell Life) in a 5% CO₂ incubator at 37 °C. The cells were passaged upon reaching approximately 80% confluence.

2.5 Isolation and identification of EVs

Human BMSCs (CP-H166, COBIOER) were cultured in DMEM containing EV-free serum for 2 days, and the supernatant was collected. The supernatant was centrifuged at 500 g for 10 min and then at 2000 g for 10 min to remove dead cells and other impurities. The resulting supernatant was centrifuged at 100 000 g and 4 °C for 1 h and resuspended in PBS. A second ultracentrifugation was performed under the same conditions. The pellet was stored at −80 °C for later use or used immediately.

The morphology of the isolated EVs [22] was then observed under transmission electron microscopy (TEM). EVs (3 μL) were added dropwise on the copper wire, followed by treatment of phosphotungstic acid solution (pH 6.8, 30 μL). The samples were dried with an incandescent lamp and photographed under TEM (H-7650, Hitachi, Tokyo, Japan).

Nanoparticle tracking analysis (NTA) [23] with a Nanosight LM10 Particle Tracking Analysis System (Malvern Instruments, Malvern, the UK) was used to measure the size distribution of EVs based on Brownian motion and diffusion coefficient. EVs were resuspended and mixed in 1 mL of PBS, and the filtered PBS was used as the control.

The expression of EV surface marker proteins (rabbit anti-CD63 (ab216130, 1:2000, Abcam Inc., Cambridge, the UK), rabbit anti-CD9 (ab223052, 1:500, Abcam), rabbit anti-TSG101 (ab30871, 1:1000, Abcam), and rabbit anti-Calnexin (ab22595, 1:100, Abcam)) was detected by Western blot analysis [24].

2.6 Uptake of EVs by TPC-1 cells

BMSC-EVs were labeled with green fluorescent dye PKH67 (UR52303, Umibio (Shanghai) Co., Ltd., Shanghai, China). About 10 µg of labeled EVs was resuspended in 100 µL of PBS and incubated with TPC-1 cells (1 × 10⁵ cells/well) for 12 h at 37 °C. The cells were collected and stained with DAPI (C1005, Beyotime Biotechnology Co., Shanghai, China), followed by immunofluorescence analysis (Leica, Carl Zeiss, Jena, Germany). A GFP vector (K483001, Thermo Fisher Scientific Inc., Waltham, MA, USA) containing the target gene was constructed and then used to transfect the MSCs. The EVs were isolated, and MSC-EVs with GFP-tagged GTF2I were obtained. The TPC-1 cells (1 × 10⁵ cells/well) were incubated for 24 h, collected, and stained with DAPI (C1005, Beyotime). Finally, immunofluorescence analysis (Leica) was performed.

2.7 Cell transduction

TPC-1 cells or MSCs at the logarithmic growth phase were plated in a 6-well cell culture plate at a density of 4 × 10⁵ cells/well. When reaching 70%–80% confluence, the TPC-1 cells were transduced as per the instructions of Lipofectamine 2000 reagent (11668-019, Invitrogen Inc., Carlsbad, CA, USA) with lentivirus carrying oe-NC, sh-NC, oe-FAT1, oe-GTF2I, oe-FOXM1, sh-GTF2I-1, sh-GTF2I-2, oe-GTF2I + sh-NC, oe-GTF2I + sh-FAT1-1, oe-GTF2I + sh-FAT1-2, oe-FAT1 + oe-NC, and oe-FAT1 + oe-CDK4. MSCs were transduced with lentivirus carrying sh-NC, sh-GTF2I-1, sh-GTF2I-2, oe-NC, and oe-GTF2I. The transfection sequence and plasmid were purchased from Shanghai GenePharma Co., Ltd. (Shanghai, China; Table S1).

A lentiviral packaging system was constructed through LV5-GFP (the lentiviral vector for gene overexpression) and pSIH1-H1-copGFP (the lentiviral vector for gene silencing). The packaging virus and the target vector were co-transduced into 293T cells using Lipofectamine 2000 upon cell confluence reaching 80%–90%. The supernatant was collected after 48 h of cell culture, and the virus particles contained in the supernatant were filtered. The virus at the exponential phase was collected, and the virus titer was detected. The cells at the exponential phase were trypsinized, pipetted to prepare a cell suspension containing 5 × 10⁴ cells/mL, seeded into a 6-well plate (2 mL per well), and cultured at 37 °C overnight.

2.8 RNA isolation and quantitation

Total RNA was extracted from cells with TRIzol reagent (Invitrogen). The extracted RNA was reverse transcribed into cDNA using the cDNA Reverse Transcription Kit (RR047A, TaKaRa). RT-qPCR was conducted using a Fast SYBR Green PCR Kit (RR820A, TaKaRa). GAPDH was used as a loading control, and the fold changes were calculated using relative quantification (the 2^-ΔΔCt method). The primer sequences are listed in Table S2.

2.9 Western blot analysis

Total protein was extracted from tissues, cells, and EVs using RIPA lysis buffer (P0013B, Beyotime) containing PMSF, and the concentration was determined by the BCA kit (P0028, Beyotime). The protein was separated using SDS–PAGE and electrotransferred onto a PVDF membrane. The membrane was treated with 5% BSA at room temperature for 1 h, and it underwent overnight incubation at 4 °C with primary rabbit antibodies (Abcam) against GTF2I (ab129025, 1:10 000), FAT1 (ab190242, 1:2000), N-cadherin (ab76011, 1:1000), slug (ab27568, 1:1000), snail (ab216347, 1:1000), E-cadherin (ab40772, 1:1000), Vimentin (ab92547, 1:1000), CD133 (Ab222782, 1:2000), Oct4 (ab200834, 1:10 000), ALDH1A1 (ab134188, 1:1000), c-Myc (ab32072, 1:1000), and GAPDH (ab9485, 1:1000; loading control). The next day, the membrane was re-probed with HRP-labeled secondary antibody IgG (ab6721, 1:5000, Abcam) for 2 h. The ECL reagent was used to visualize the results with the Image Quant LAS 4000C Gel Imager.

2.10 ChIP assay

EZ-Magna ChIP Kit (17-295, Sigma) was used for this assay. TPC-1 cells were fixed with 1% formaldehyde to produce DNA–protein cross-linking. The cells were then lysed with cell lysis buffer and nuclear lysis buffer and subjected to ultrasonic treatment to produce 200–1000 bp chromatin fragments. Thereafter, immunoprecipitation was performed using ProteinA Agarose/Salmon Sperm DNA. The precipitate was washed, eluted, and incubated with 20 μL of 5 mol/L NaCl to unlock the cross-linking. DNA was recovered, and the enriched chromatin fragments were detected by fluorescence-based quantitative PCR. FAT1 primer information is as follows: forward: 5′-TCTTGTCCGGGCTGCGTC-3′ and reverse: 5′-AAGTTCCTAGGCAGAGGGGAA-3′.

2.11 Dual-luciferase reporter assay

The wild-type dual-luciferase reporter gene plasmid containing full-length FAT1 promoter and mutant dual-luciferase reporter gene plasmid were constructed: pmirGLO-WT, pmirGLO-MUT1 (GTF2I binding site 1 mutation, TCCCGCTGCCCTAGGA), pmirGLO-MUT2 (GTF2I binding site 2 mutation, AGTTGCTGCCCCCGAGAAA), and pmirGLO-MUT3 (GTF2I binding site 3 mutation, GGCTGCCGG). The dual-luciferase reporter gene plasmids were co-transfected with Vector and GTF2I into 293T cells for 48 h. The luciferase activity was detected using a Dual-Luciferase Reporter Assay System (E1910, Promega Corporation, Madison, WI, USA).

2.12 Colony formation assay

A single cell suspension was seeded in a 6-well plate at a density of 100 cells/well and cultured for 2 weeks to form colonies. With the supernatant discarded, the cells were fixed with 4% paraformaldehyde and stained with 0.5% crystal violet solution. The cells were air-dried in a ventilated place at room temperature, counted, and photographed. The number of colony (cells > 50) formed in each well was calculated.

2.13 Transwell assay

Transwell chamber (CLS3396, Corning Incorporated, Corning, NY, USA) in 24-well plates was used for cell migration and invasion assays. The lower Transwell chamber uncoated (migration assay) with Matrigel (E1270, Sigma) or coated (invasion assay) with Matrigel was pre-added with 600 mL of 10% FBS culture medium and equilibrated at 37 °C for 1 h. The TPC-1 cells, after different treatments, were resuspended in FBS-free culture medium, seeded into the upper chamber at a density of 1 × 10⁶ cells/mL, and cultured at 37 °C with 5% CO₂ for 24 h. The Transwell chamber was fixed with 5% glutaraldehyde and stained with 0.1% crystal violet for 5 min at 4 °C. The cells were observed under an inverted fluorescence microscope (TE2000, Nikon, China).

2.14 Sphere formation assay

TPC-1 cells were plated in a 6-well plate (CLS3335, Corning) with very low attachment at a density of 1 × 10³ cells/well and kept in serum-free medium as previously reported [25]. After 10–14 days of culture using the above method, the number of cell spheres with a diameter greater than 75 μm was counted when there were more cells with a spherical diameter of 75 μm.

2.15 Xenograft tumor in nude mice

A total of 48 six-week-old specific pathogen-free female BALB/c nude mice (Hunan SJA Laboratory Animal Co., Ltd., Human, China) were housed in a laboratory at 22 °C–25 °C with humidity of 60%–65% under a 12 h light/dark cycle, with ad libitum access to food and water. The mice were acclimatized for 1 week before the experiment.

About 200 μL of TPC-1 cells (5 × 10⁶ cells) resuspended in serum-free DMEM was subcutaneously inoculated into the right back of 24 randomly selected mice. The tumor volume was measured every 7 days. The nude mice were randomly divided into four groups: PBS, MSC-EVs (without any treatment), MSC-EVs/sh-NC (EVs derived from sh-NC-transduced MSCs), and MSC-EVs/sh-GTF2I (EVs derived from sh-GTF2I-transduced MSCs), with 6 mice in each group. The corresponding EVs (100 μg; 50 μg/100 μL) were injected into the tumor center of mice via the tail vein every 4 days [26]. After 4 weeks, the mice were euthanized, and the primary tumor was excised, with the tumor volume and weight recorded.

Construction of the in vivo lung metastasis model: 24 nude mice were randomly selected and injected with 2 × 10⁶ TPC-1 cells treated with PBS, MSC-EVs, MSC-EVs/sh-NC, and MSC-EVs/sh-GTF2I via the tail vein, with 6 mice following each treatment. The corresponding EVs (100 μg; 50 μg/100 μL) were injected into the mice via the tail vein every 4 days. After 6 weeks, the mice were euthanized, and the lung tissue was excised for HE staining. Metastatic nodules were counted [27].

2.16 Immunohistochemical staining

Tumor tissue samples were fixed in 10% neutral formalin solution, paraffin-embedded, and sectioned. The sections were heated in 10 mM sodium citrate, blocked with 10% normal goat serum, and probed with primary rabbit antibody against Ki67 (ab15580, 1:100, Abcam) at 4 °C overnight. The next day, the sections were re-probed with secondary antibody goat anti-rabbit IgG (ab97057, 1:5000, Abcam) for 30 min and treated with HRP-labeled streptavidin protein working solution (A0303, Beyotime) for 30 min at 37 °C. Thereafter, DAB (P0202, Beyotime) was added for development. The sections were counterstained with hematoxylin (C0107, Beyotime) for 30 s. Finally, the sections were observed under an upright microscope (BX63, Olympus Japan Co., Ltd., Tokyo, Japan).

2.17 HE staining

Lung tissues of mice were fixed in 4% paraformaldehyde, dehydrated, paraffin-embedded, and sectioned. The sections were stained with hematoxylin solution for 5 min and counterstained with 5% eosin solution for 3 min. Finally, the sections were observed under a microscope (Olympus) to count the metastatic nodules [28].

2.18 Statistical analysis

SPSS 21.0 statistical software (IBM Corp. Armonk, NY, USA) was used for data processing, and statistical significance was set at P < 0.05. The measurement data were described as mean ± standard deviation. Data between two groups were compared using paired t-test (paired data) or unpaired t-test (unpaired data). Differences among multiple groups were statistically analyzed using one-way ANOVA with Tukey’s post hoc tests. Tumor volume data among multiple groups at different time points were compared using repeated measures ANOVA, followed by Bonferroni post hoc test for multiple comparisons.

3 Results

3.1 Ectopic expression of FAT1 inhibits the oncogenic phenotypes, EMT, and stemness maintenance of TC cells

Differential analysis of TC-related TCGA data from the GEPIA2 database yielded 9815 DEGs, consisting of 1908 highly expressed DEGs and 7907 poorly expressed DEGs (Table S3). In addition, 5000 and 3000 TC-related genes were retrieved from the GeneCards and Phenolyzer databases, respectively, and then intersected with the poorly expressed DEGs, with 345 candidate genes obtained (Fig.1). GO and KEGG enrichment analyses suggested that the 345 candidate genes were mainly enriched in urogenital system development, activation of protein kinase activity, epithelial cell proliferation, protein kinase B signaling, and epithelial cell development (Fig.1). The gene network involved in related pathways is shown in Fig.1, where 21 genes were involved in the epithelial cell development pathway. Further analysis of TC-related GSE66783 showed that FAT1 was poorly expressed in TC (Fig.1). In addition, the mRNA expression of FAT1 was significantly downregulated in TCGA_TC samples (Fig.1).

Compared with adjacent tissues, TC tissues presented lower FAT1 expression (Fig.1). This result demonstrated that the poor expression of FAT1 occurred in TC cell lines (TPC-1, 8305C, CAL62, and IHH4) relative to the normal thyroid cell line Nthy-ori 3-1 (Fig.1), with TPC-1 cells showing low expression. Thus, TPC-1 cells were chosen for further research.

TPC-1 cells were transduced with oe-FAT1, the efficiency of which was verified by RT-qPCR and Western blot analyses (Fig.1). As shown in Fig.1 and 1J, ectopically expressed FAT1 decreased the clonogenic potential and migratory and invasive capacities of TPC-1 cells. Western blot analysis was adopted to detect the expression of EMT-related proteins E-cadherin (epithelial cadherin), N-cadherin (neural cadherin), Vimentin (mesenchymal cell marker) [29,30], snail (zinc-finger transcription factor), and snail2 (slug; zinc-finger transcription factor) [31]. The results showed a decline in the expression of Vimentin, slug, snail, and N-cadherin but an increase in that of E-cadherin in TPC-1 cells overexpressing FAT1 (Fig.1). In the presence of FAT1 overexpression, sphere formation ability was found to be reduced (Fig.1). Furthermore, the protein expression of stemness-related proteins CD133, Oct4, ALDH1A1, and c-Myc was reduced in TPC-1 cells upon FAT1 overexpression (Fig.1). The above results indicated that FAT1 was downregulated in TC, but its overexpression could suppress TC cell oncogenic phenotypes, EMT, and stemness maintenance.

3.2 GTF2I, downregulated in TC, binds to the promoter region of FAT1 to augment its expression

On the basis of the expression of FAT1 in TCGA_TC samples, we used Pearson’s correlation coefficient and found that 847 genes were significantly positively correlated. Using the hTFtarget database, 150 transcription factors that may regulate FAT1 were obtained. Following intersection analysis of related genes and hTFtarget prediction results, 11 candidate transcription factors were identified (Fig.2). The expression of these 11 candidate transcription factors in the TCGA_THCA dataset is shown in Fig.2. Among them, GTF2I had a larger fold change and a high overall expression level. GTF2I expression was positively correlated with FAT1 expression, as analyzed by Pearson’s correlation coefficient (Fig.2). Thus, GTF2I was chosen for further research.

RT-qPCR results showed diminished expression of GTF2I in TC tissues as compared with adjacent tissues (Fig.2). We also noted that the expression of GTF2I in TC cell lines (TPC-1, 8305C, CAL62, and IHH4) was lower than that in the normal thyroid cell line Nthy-ori 3-1 (Fig.2). The transduction efficiency of oe-GTF2I and sh-GTF2I in TPC-1 cells was confirmed by RT-qPCR, with sh-GTF2I-1 exhibiting superior efficiency (Fig.2), so it was used for the subsequent experiment. The expression of FAT1 was detected to be elevated in oe-GTF2I-transduced TPC-1 cells, but it was reduced in the absence of GTF2I (Fig.2).

hTFtarget predicted the presence of binding sites of the GTF2I and FAT1 promoter region (Table S4), and the top three were as follows: site1: 1243–1258 (TCCCCTCTGCCTAGGA), site2: 1928–1946 (AGTTTCCGCGCCCGAGAAA), and site3: 1975–1983 (GGGCGGAGG). The results of dual-luciferase reporter assay suggested that ectopically expressed GTF2I increased the luciferase activity of promoter WT, promoter MUT-site2, and promoter MUT-site3 but failed to alter that of promoter MUT-site1 (Fig.2). Thus, the binding site of GTF2I to the FAT1 promoter region was site1. Meanwhile, the results of ChIP assay showed that GTF2I was enriched in the FAT1 promoter region in TPC-1 cells (Fig.2). These results demonstrated that GTF2I was downregulated in TC, and it could bind to the promoter region of FAT1 to increase the expression of FAT1.

3.3 GTF2I represses the oncogenic phenotypes, EMT, and stemness maintenance of TC cells by upregulating FAT1 expression

Furthermore, we sought to explore the effect of GTF2I regulating FAT1 expression on TC. The results of RT-qPCR showed an increase in the expression of GTF2I and FAT1 in the oe-GTF2I-treated cells, whereas the expression of FAT1 was reduced following transduction with oe-GTF2I + sh-FAT1-1 or oe-GTF2I + sh-FAT1-2. Among them, oe-GTF2I + sh-FAT1-1 presented superior silencing efficiency (Fig.3), so it was chosen for subsequent experiments.

Ectopic expression of GTF2I decreased the clonogenic potential and migratory and invasive capacities of TPC-1 cells, which was negated by silencing of FAT1 (Fig.3 and 3C). Western blot analysis results indicated a decline in the expression of Vimentin, slug, snail, and N-cadherin but an increase in that of E-cadherin in TPC-1 cells overexpressing GTF2I. However, opposite results were caused following FAT1 silencing (Fig.3). Moreover, sphere formation ability was reduced in response to FAT1 overexpression, but it was enhanced following FAT1 silencing (Fig.3). Furthermore, Western blot analysis results illustrated that the expression of CD133, Oct4, ALDH1A1, and c-Myc was diminished in TPC-1 cells upon GTF2I overexpression, the effect of which was undermined by FAT1 silencing (Fig.3). Altogether, these data indicated that GTF2I could arrest TC cell oncogenic phenotypes, EMT, and stemness maintenance by upregulating FAT1 expression.

3.4 MSC-EVs can transfer GTF2I into TPC-1 cells

Under TEM, MSC-EVs exhibited round or oval membranous vesicles (Fig.4). The results of NTA analysis showed that the diameter of EVs was 30–120 nm (Fig.4). Western blot analysis revealed the expression of CD9, CD63, and TSG101 in EVs without calnexin expression (Fig.4). These results indicated the successful isolation of EVs. Moreover, in MSC-EVs, we witnessed the poor expression of GTF2I mRNA. In the presence of sh-GTF2I, GTF2I protein expression was decreased in MSCs and EVs (Fig. S1).

Coculture data of PKH67-labeled EVs and TPC-1 cells suggested the obvious fluorescence distribution in TPC-1 cells under a fluorescence microscope (Fig.4), which indicated that PKH67-labeled EVs could be effectively internalized by TPC-1 cells. MSCs were transfected with GTF2I labeled with GFP to obtain MSC-EVs with GFP-labeled GTF2I. After incubation with TPC-1 cells for 24 h, GFP-labeled GTF2I was present in TPC-1 cells (Fig.4). Western blot analysis results further demonstrated the internalization of GTF2I in EVs by TPC-1 cells, as evidenced by increased GTF2I expression in TPC-1 cells cocultured with MSC-EVs; however, downregulated GTF2I expression was witnessed upon further GTF2I silencing (Fig.4). These results indicated that MSC-EVs could deliver GTF2I into TPC-1 cells.

3.5 MSC-EVs transfer GTF2I to upregulate FAT1 expression, thereby restricting the oncogenic phenotypes, EMT, and stemness maintenance of TC cells

To evaluate the effect of MSC-EVs loaded with GTF2I on TC, we first treated TPC-1 cells with sh-FAT1-1, followed by MSC-EVs. The results of RT-qPCR and Western blot analyses showed increased GTF2I protein expression and FAT1 mRNA and protein expression, but GTF2I mRNA expression was not altered in cells treated with MSC-EVs. By contrast, GTF2I mRNA and protein expression was found to be unchanged, whereas FAT1 mRNA and protein expression decreased in cells treated with MSC-EVs + sh-FAT1-1 (Fig.5).

Treatment with MSC-EVs led to decreased clonogenic potential and migratory and invasive capacities of TPC-1 cells, which was negated by further treatment with sh-FAT1-1 (Fig.5 and 5C). The results of Western blot analysis presented a decline in the expression of Vimentin, slug, snail, and N-cadherin but an increase in that of E-cadherin in TPC-1 cells cocultured with MSC-EVs. However, opposite results were noted following FAT1 silencing (Fig.5). Sphere formation assay results demonstrated that the sphere formation ability of TPC-1 cells cocultured with MSC-EVs was weakened, but it was enhanced following FAT1 silencing (Fig.5). Fig.5 depicts a decline in the protein expression of CD133, Oct4, ALDH1A1, and c-Myc in TPC-1 cells cocultured with MSC-EVs, which was reversed by FAT1 silencing. In summary, MSC-EVs could deliver GTF2I to TC cells where GTF2I upregulated FAT1 expression and impaired the oncogenic phenotypes, EMT, and stemness maintenance of TC cells.

3.6 FAT1 downregulates CDK4/FOXM1 to inhibit the oncogenic phenotypes, EMT, and stemness maintenance of TC cells

We extended our mechanistic findings to determine the inhibiting effect of FAT1 on the TC cell malignant phenotype. Analysis of the TC-related TCGA data indicated that CDK4 and FOXM1 were highly expressed in TC samples (Fig.6). In addition, RT-qPCR results showed higher expression of CDK4 and FOXM1 in TC tissues than in adjacent tissues (Fig.6). Consistently higher expression of CDK4 and FOXM1 was noted in TC cell lines (TPC-1, 8305C, CAL62, and IHH4) than in the normal thyroid cell line Nthy-ori 3-1 (Fig.6). In TPC-1 cells, oe-FAT1 increased the expression of FAT1 but suppressed that of CDK4 and FOXM1. However, further addition of oe-CDK4 led to the augmented expression of CDK4 and FOXM1 (Fig.6).

The results of colony formation and Transwell assays suggested the decreased clonogenic potential and migratory and invasive capacities of TPC-1 cells transduced with oe-FAT1, which was negated by further treatment with oe-CDK4 (Fig.6 and 6F). The results of Western blot analysis exhibited a decline in the expression of Vimentin, slug, snail, and N-cadherin but an increase in that of E-cadherin in TPC-1 cells overexpressing FAT1. However, opposite results were induced following CDK4 overexpression (Fig.6). Moreover, the sphere formation ability of oe-FAT1-treated TPC-1 cells was attenuated, but CDK4 overexpression reversed this result (Fig.6). As illustrated in Fig.6, the protein expression of CD133, Oct4, ALDH1A1, and c-Myc was inhibited in TPC-1 cells transduced with oe-FAT1, which was annulled by CDK4 overexpression.

3.7 Overall, FAT1 inhibited the oncogenic phenotypes, EMT, and stemness maintenance of TC cells by disrupting CDK4/FOXM1.

GTF2I delivered by MSC-EVs elevates FAT1 expression and inhibits CDK4/FOXM1, thereby blunting the oncogenic phenotypes, EMT, and stemness maintenance of TC cells

To characterize the effect of MSC-EVs enriched with GTF2I on TC by regulating the FAT1/CDK4/FOXM1 axis, we first transduced TPC-1 cells with oe-FOXM1 and then treated them with EVs isolated from MSCs transduced with oe-NC or oe-GTF2I (Fig. S2). Western blot analysis results showed higher GTF2I expression in TPC-1 cells treated with MSC-EVs/oe-GTF2I than in cells treated with MSC-EVs/oe-NC (Fig.7). Further RT-qPCR results demonstrated enhanced FAT1 expression but decreased CDK4 and FOXM1 expression in response to MSC-EVs/oe-GTF2I. However, treatment with MSC-EVs/oe-GTF2I + oe-FOXM1 increased FOXM1 expression without altering FAT1 and CDK4 expression (Fig.7).

Additionally, the results of colony formation and Transwell assays exhibited that MSC-EVs/oe-GTF2I decreased the clonogenic potential and migratory and invasive capacities of TPC-1 cells, and these changes were reversed by further treatment with oe-FOXM1 (Fig.7 and 7D). The results of Western blot analysis indicated a decline in the expression of Vimentin, slug, snail, and N-cadherin but an increase in that of E-cadherin in TPC-1 cells treated with MSC-EVs/oe-GTF2I. However, contrary results were caused by MSC-EVs/oe-GTF2I + oe-FOXM1 (Fig.7). Moreover, the sphere formation ability of MSC-EVs/oe-GTF2I-treated TPC-1 cells was reduced, but additional FOXM1 overexpression reversed this result (Fig.7). As illustrated in Fig.7, MSC-EVs/oe-GTF2I suppressed the protein expression of CD133, Oct4, ALDH1A1, and c-Myc, which was annulled by additional FOXM1 overexpression. Cumulatively, the abovementioned findings mentioned suggested that MSC-EVs could shuttle GTF2I to upregulate FAT1 expression and inactivate CDK4/FOXM1, thereby attenuating the oncogenic phenotypes, EMT, and stemness maintenance of TC cells.

3.8 GTF2I delivered by MSC-EVs inhibits the tumorigenesis of TC cells in vivo

Finally, we proceeded to dissect the effect of MSC-EVs packaged with GTF2I on TC in vivo. As shown in Fig.8–8C, tumor volume and weight of MSC-EV-treated mice were reduced, whereas an increase was found in the presence of MSC-EVs/sh-GTF2I. Western blot analysis results showed that GTF2I expression was augmented in the tumor tissue of MSC-EVs-treated mice, but this augmentation was reversed following treatment with MSC-EVs/sh-GTF2I (Fig.8). In addition, RT-qPCR data exhibited elevated FAT1 expression and decline of CDK4 and FOXM1 expression in the tumor tissue of mice treated with MSC-EVs. MSC-EVs/sh-GTF2I led to opposite results (Fig.8). Fig.8 depicts a reduction in the positive rate of Ki67 expression in the tumor tissue of mice treated with MSC-EVs, whereas silencing of GTF2I induced an increase in the positive rate of Ki67 expression. HE staining results suggested that MSC-EVs reduced the number of metastatic nodules in the lung tissue of mice, whereas further treatment with sh-GTF2I increased the number of metastatic nodules (Fig.8). Overall, the results described above demonstrated that GTF2I delivered by MSC-EVs attenuated the tumorigenesis of TC cells in vivo.

4 Discussion

MSC-EVs have demonstrated potential application in clinical settings as new biomarkers and therapeutic targets in human diseases [32]. The findings collected from this study indicated that MSC-EVs enriched with GTF2I could potentially arrest the oncogenic phenotypes, EMT, and stemness maintenance of TC cells via regulating the FAT1/CDK4/FOXM1 axis.

Our results provided evidence suggesting that ectopically expressed FAT1 impaired the oncogenic phenotypes, EMT, and stemness maintenance of TC cells. Consistent with this finding, FAT1 knockdown contributes to promoting the proliferative capacities of sporadic medullary TC cells [15]. Meanwhile, FAT1 knockdown has been reported to accelerate tumor initiation, malignant progression, invasiveness, stemness, and EMT [14]. These findings suggested that FAT1 represents an important player in TC progression and a potential therapeutic target for this cancer.

GTF2I is a ubiquitously expressed transcription factor that can regulate gene expression positively or negatively [11]. A transcription factor E2F1 has been reported to occupy the promoter region of FAT1 and then elevate its transcription activity and mRNA levels [33]. Here, the current study revealed that GTF2I could bind to the promoter region of FAT1 and promote its expression. Additionally, the present results showed that GTF2I overexpression repressed the oncogenic phenotypes, EMT, and stemness maintenance of TC cells by upregulating FAT1 expression. In partial agreement with our results, a recent study revealed that GTF2I mutation is responsible for the increased protein expression of EMT-related genes in thymic epithelial cells [13]. The mRNA expression of GTF2I is significantly decreased in breast cancer tissues and cells, and this decrease can lead to the promotion of the proliferative and migratory phenotypes of breast cancer cells [34]. CD133, Oct4, ALDH1A1, and c-Myc are well-established cancer stem cell-related markers, and their increased expression indicates the enhanced stemness of cancer cells [5,35]. The current study revealed that GTF2I overexpression in TPC-1 cells could reduce the expression of CD133, Oct4, ALDH1A1, and c-Myc, thereby suggesting the inhibiting effect of GTF2I on the stemness of TC cells, as well as highlighting its probable potential as a therapeutic target for TC treatment. However, despite these results, further investigation is still required on the role of GTF2I in the biological functions of TC cells due to the scarcity of supporting literature.

Mechanistic investigations showed that the anti-tumor properties of GTF2I on TC were achieved by MSC-EVs, which could serve as a delivery system to transfer GTF2I into TPC-1 cells where GTF2I exerted potent anti-tumor roles. MSC-derived EVs have therapeutic potential in cancer treatment due to their involvement in intercellular communication through the transfer of proteins, RNA, DNA, and bioactive lipids [36]. For instance, exosomes isolated from BMSCs have been shown to deliver miR-152 into TC cells and inhibit the proliferative, invasive, and migratory phenotypes of TC cells [37]. Thus, MSC-EVs can transfer GTF2I to upregulate FAT1 expression and consequently attenuate the oncogenic phenotypes, EMT, and stemness maintenance of TC cells.

Further analysis exhibited that FAT1 downregulated CDK4/FOXM1 to inhibit the oncogenic phenotypes, EMT, and stemness maintenance of TC cells. Knockdown of FAT1 results in the upregulation of CDK4 expression [38]; meanwhile, CDK4 can stabilize and activate FOXM1 [39]. These features indicate the possible adverse correlation of FAT1 with FOXM1. In addition, inhibition of CDK4 triggers cell death and suppresses cell viability in anaplastic TC cells [40]. Elevated expression of FOXM1 has been confirmed in papillary TC cell lines, and this elevation can augment the proliferative capacities of papillary TC cells [41]. Considering the published reports combined with the current results, we concluded that GTF2I delivered by MSC-EVs could elevate FAT1 expression and inhibit CDK4/FOXM1, arresting the oncogenic phenotypes, EMT, and stemness maintenance of TC cells.

Overall, our study indicated that MSC-EVs could transfer GTF2I to TC cells where GTF2I elevated FAT1 and impaired CDK4/FOXM1, thereby preventing oncogenic phenotypes, EMT, and stemness maintenance of TC cells (Fig.9). These results can provide a thorough understanding of the onset and progression of TC, which further aids in the development of early detection molecular markers for TC treatment. However, other factors may be involved in TC progression in different ways, which will be further explored in future studies. Further studies with specimens from patients diagnosed with TC are essential to validate these findings and expand the translational potential of this direction.

References

[1] Bray F, Ferlay J, Soerjomataram I, Siegel RL, Torre LA, Jemal A. Global cancer statistics 2018: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin 2018; 68(6): 394–424

[2] Kim J, Gosnell JE, Roman SA. Geographic influences in the global rise of thyroid cancer. Nat Rev Endocrinol 2020; 16(1): 17–29

[3] Aleksakhina SN, Kashyap A, Imyanitov EN. Mechanisms of acquired tumor drug resistance. Biochim Biophys Acta Rev Cancer 2019; 1872(2): 188310

[4] Bocci F, Gearhart-Serna L, Boareto M, Ribeiro M, Ben-Jacob E, Devi GR, Levine H, Onuchic JN, Jolly MK. Toward understanding cancer stem cell heterogeneity in the tumor microenvironment. Proc Natl Acad Sci USA 2019; 116(1): 148–157

[5] Pai S, Bamodu OA, Lin YK, Lin CS, Chu PY, Chien MH, Wang LS, Hsiao M, Yeh CT, Tsai JT. CD47-SIRPα signaling induces epithelial-mesenchymal transition and cancer stemness and links to a poor prognosis in patients with oral squamous cell carcinoma. Cells 2019; 8(12): 1658

[6] Zhang Z, Wang ZX, Chen YX, Wu HX, Yin L, Zhao Q, Luo HY, Zeng ZL, Qiu MZ, Xu RH. Integrated analysis of single-cell and bulk RNA sequencing data reveals a pan-cancer stemness signature predicting immunotherapy response. Genome Med 2022; 14(1): 45

[7] Sanmartin MC, Borzone FR, Giorello MB, Yannarelli G, Chasseing NA. Mesenchymal stromal cell-derived extracellular vesicles as biological carriers for drug delivery in cancer therapy. Front Bioeng Biotechnol 2022; 10: 882545

[8] Hassanzadeh A, Rahman HS, Markov A, Endjun JJ, Zekiy AO, Chartrand MS, Beheshtkhoo N, Kouhbanani MAJ, Marofi F, Nikoo M, Jarahian M. Mesenchymal stem/stromal cell-derived exosomes in regenerative medicine and cancer; overview of development, challenges, and opportunities. Stem Cell Res Ther 2021; 12(1): 297

[9] Rani S, Ryan AE, Griffin MD, Ritter T. Mesenchymal stem cell-derived extracellular vesicles: toward cell-free therapeutic applications. Mol Ther 2015; 23(5): 812–823

[10] Qiu G, Zheng G, Ge M, Wang J, Huang R, Shu Q, Xu J. Mesenchymal stem cell-derived extracellular vesicles affect disease outcomes via transfer of microRNAs. Stem Cell Res Ther 2018; 9(1): 320

[11] Gurumurthy A, Wu Q, Nar R, Paulsen K, Trumbull A, Fishman RC, Brand M, Strouboulis J, Qian Z, Bungert J. TFII-I/Gtf2i and erythro-megakaryopoiesis. Front Physiol 2020; 11: 590180

[12] Nathany S, Tripathi R, Mehta A. Gene of the month: GTF2I. J Clin Pathol 2021; 74(1): 1–4

[13] Kim IK, Rao G, Zhao X, Fan R, Avantaggiati ML, Wang Y, Zhang YW, Giaccone G. Mutant GTF2I induces cell transformation and metabolic alterations in thymic epithelial cells. Cell Death Differ 2020; 27(7): 2263–2279

[14] Pastushenko I, Mauri F, Song Y, de Cock F, Meeusen B, Swedlund B, Impens F, Van Haver D, Opitz M, Thery M, Bareche Y, Lapouge G, Vermeersch M, Van Eycke YR, Balsat C, Decaestecker C, Sokolow Y, Hassid S, Perez-Bustillo A, Agreda-Moreno B, Rios-Buceta L, Jaen P, Redondo P, Sieira-Gil R, Millan-Cayetano JF, Sanmatrtin O, D’Haene N, Moers V, Rozzi M, Blondeau J, Lemaire S, Scozzaro S, Janssens V, De Troya M, Dubois C, Pérez-Morga D, Salmon I, Sotiriou C, Helmbacher F, Blanpain C. Fat1 deletion promotes hybrid EMT state, tumour stemness and metastasis. Nature 2021; 589(7842): 448–455

[15] Qu N, Shi X, Zhao JJ, Guan H, Zhang TT, Wen SS, Liao T, Hu JQ, Liu WY, Wang YL, Huang S, Shi RL, Wang Y, Ji QH. Genomic and transcriptomic characterization of sporadic medullary thyroid carcinoma. Thyroid 2020; 30(7): 1025–1036

[16] Li Z, Razavi P, Li Q, Toy W, Liu B, Ping C, Hsieh W, Sanchez-Vega F, Brown DN, Da Cruz Paula AF, Morris L, Selenica P, Eichenberger E, Shen R, Schultz N, Rosen N, Scaltriti M, Brogi E, Baselga J, Reis-Filho JS, Chandarlapaty S. Loss of the FAT1 tumor suppressor promotes resistance to CDK4/6 inhibitors via the hippo pathway. Cancer Cell 2018; 34(6): 893–905.e8

[17] Wang X, Sun Z, Tian W, Piao C, Xie X, Zang J, Peng S, Yu X, Wang Y. S100A12 is a promising biomarker in papillary thyroid cancer. Sci Rep 2020; 10(1): 1724

[18] Rubio C, Martínez-Fernández M, Segovia C, Lodewijk I, Suarez-Cabrera C, Segrelles C, López-Calderón F, Munera-Maravilla E, Santos M, Bernardini A, García-Escudero R, Lorz C, Gómez-Rodriguez MJ, de Velasco G, Otero I, Villacampa F, Guerrero-Ramos F, Ruiz S, de la Rosa F, Domínguez-Rodríguez S, Real FX, Malats N, Castellano D, Dueñas M, Paramio JM. CDK4/6 inhibitor as a novel therapeutic approach for advanced bladder cancer independently of RB1 status. Clin Cancer Res 2019; 25(1): 390–402

[19] Bellelli R, Castellone MD, Garcia-Rostan G, Ugolini C, Nucera C, Sadow PM, Nappi TC, Salerno P, Cantisani MC, Basolo F, Gago TA, Salvatore G, Santoro M. FOXM1 is a molecular determinant of the mitogenic and invasive phenotype of anaplastic thyroid carcinoma. Endocr Relat Cancer 2012; 19(5): 695–710

[20] Feng L, Yang B, Tang XD. Long noncoding RNA LINC00460 promotes carcinogenesis via sponging miR-613 in papillary thyroid carcinoma. J Cell Physiol 2019; 234(7): 11431–11439

[21] Xia F, Chen Y, Jiang B, Du X, Peng Y, Wang W, Huang W, Feng T, Li X. Long noncoding RNA HOXA-AS2 promotes papillary thyroid cancer progression by regulating miR-520c-3p/S100A4 pathway. Cell Physiol Biochem 2018; 50(5): 1659–1672

[22] Liao Z, Luo R, Li G, Song Y, Zhan S, Zhao K, Hua W, Zhang Y, Wu X, Yang C. Exosomes from mesenchymal stem cells modulate endoplasmic reticulum stress to protect against nucleus pulposus cell death and ameliorate intervertebral disc degeneration in vivo. Theranostics 2019; 9(14): 4084–4100

[23] Pritchard A, Tousif S, Wang Y, Hough K, Khan S, Strenkowski J, Chacko BK, Darley-Usmar VM, Deshane JS. Lung tumor cell-derived exosomes promote M2 macrophage polarization. Cells 2020; 9(5): 1303

[24] Fang T, Lv H, Lv G, Li T, Wang C, Han Q, Yu L, Su B, Guo L, Huang S, Cao D, Tang L, Tang S, Wu M, Yang W, Wang H. Tumor-derived exosomal miR-1247-3p induces cancer-associated fibroblast activation to foster lung metastasis of liver cancer. Nat Commun 2018; 9(1): 191

[25] Shao XJ, Xiang SF, Chen YQ, Zhang N, Cao J, Zhu H, Yang B, Zhou Q, Ying MD, He QJ. Inhibition of M2-like macrophages by all-trans retinoic acid prevents cancer initiation and stemness in osteosarcoma cells. Acta Pharmacol Sin 2019; 40(10): 1343–1350

[26] Dai X, Xie Y, Dong M. Cancer-associated fibroblasts derived extracellular vesicles promote angiogenesis of colorectal adenocarcinoma cells through miR-135b-5p/FOXO1 axis. Cancer Biol Ther 2022; 23(1): 76–88

[27] Yang Y, Mao F, Guo L, Shi J, Wu M, Cheng S, Guo W. Tumor cells derived-extracellular vesicles transfer miR-3129 to promote hepatocellular carcinoma metastasis by targeting TXNIP. Dig Liver Dis 2021; 53(4): 474–485

[28] Cui Y, Wang D, Xie M. Tumor-derived extracellular vesicles promote activation of carcinoma-associated fibroblasts and facilitate invasion and metastasis of ovarian cancer by carrying miR-630. Front Cell Dev Biol 2021; 9: 652322

[29] Yang XZ, Cheng TT, He QJ, Lei ZY, Chi J, Tang Z, Liao QX, Zhang H, Zeng LS, Cui SZ. LINC01133 as ceRNA inhibits gastric cancer progression by sponging miR-106a-3p to regulate APC expression and the Wnt/β-catenin pathway. Mol Cancer 2018; 17(1): 126

[30] Wei C, Yang C, Wang S, Shi D, Zhang C, Lin X, Liu Q, Dou R, Xiong B. Crosstalk between cancer cells and tumor associated macrophages is required for mesenchymal circulating tumor cell-mediated colorectal cancer metastasis. Mol Cancer 2019; 18(1): 64

[31] Liang H, Yu T, Han Y, Jiang H, Wang C, You T, Zhao X, Shan H, Yang R, Yang L, Shan H, Gu Y. LncRNA PTAR promotes EMT and invasion-metastasis in serous ovarian cancer by competitively binding miR-101-3p to regulate ZEB1 expression. Mol Cancer 2018; 17(1): 119

[32] Park KS, Bandeira E, Shelke GV, Lässer C, Lötvall J. Enhancement of therapeutic potential of mesenchymal stem cell-derived extracellular vesicles. Stem Cell Res Ther 2019; 10(1): 288

[33] Wang Y, Wang G, Ma Y, Teng J, Wang Y, Cui Y, Dong Y, Shao S, Zhan Q, Liu X. FAT1, a direct transcriptional target of E2F1, suppresses cell proliferation, migration and invasion in esophageal squamous cell carcinoma. Chin J Cancer Res 2019; 31(4): 609–619

[34] Zhou SW, Su BB, Feng YQ, Du XQ, Zhao H. Expression of GTF2IP23 in breast cancer and it mediated regulation of GTF2I. Chin J Oncol (Zhonghua Zhong Liu Za Zhi) 2019; 41(12): 918–922

[35] Chen Y, Shao Z, Jiang E, Zhou X, Wang L, Wang H, Luo X, Chen Q, Liu K, Shang Z. CCL21/CCR7 interaction promotes EMT and enhances the stemness of OSCC via a JAK2/STAT3 signaling pathway. J Cell Physiol 2020; 235(9): 5995–6009

[36] Xunian Z, Kalluri R. Biology and therapeutic potential of mesenchymal stem cell-derived exosomes. Cancer Sci 2020; 111(9): 3100–3110

[37] Tang M, Wang Q, Wang K, Wang F. Mesenchymal stem cells-originated exosomal microRNA-152 impairs proliferation, invasion and migration of thyroid carcinoma cells by interacting with DPP4. J Endocrinol Invest 2020; 43(12): 1787–1796

[38] Hu XL, Zhai YF, Li GD, Xing JF, Yang J, Bi YH, Wang J, Shi RY. FAT1 inhibits cell proliferation of esophageal squamous cell carcinoma through regulating the expression of CDK4/CDK6/CCND1 complex. Zhonghua Zhong Liu Za Zhi 2018; 40(1): 14–20

[39] Anders L, Ke N, Hydbring P, Choi YJ, Widlund HR, Chick JM, Zhai H, Vidal M, Gygi SP, Braun P, Sicinski P. A systematic screen for CDK4/6 substrates links FOXM1 phosphorylation to senescence suppression in cancer cells. Cancer Cell 2011; 20(5): 620–634

[40] Lopes-Ventura S, Pojo M, Matias AT, Moura MM, Marques IJ, Leite V, Cavaco BM. The efficacy of HRAS and CDK4/6 inhibitors in anaplastic thyroid cancer cell lines. J Endocrinol Invest 2019; 42(5): 527–540

[41] Wang G, Wang X, Jin Y. LINC01410/miR-3619-5p/FOXM1 Feedback Loop Regulates Papillary Thyroid Carcinoma Cell Proliferation and Apoptosis. Cancer Biother Radiopharm 2019; 34(9): 572–580

RIGHTS & PERMISSIONS

Higher Education Press