Analysis of differentially expressed genes (DEGs) in thyroid cancer and normal tissues was performed using gene expression profiling interactive analysis (GEPIA, http://gepia.cancer-pku.cn/index.html). The criteria were $|$log2FC$|$ Cutoff: 1, p-value Cutoff: 0.01, THCA, Match TCGA normal and GTEx data, and 512 thyroid cancer tissue samples and 337 normal control tissue samples were included. The same method was used to analyze differences in the expression level of fibronectin 1 (FN1) between thyroid cancer tissues and normal tissues. The R package (version 3.4.4, Auckland, New Zealand) was used to divide samples into high expression and low expression groups according to the median expression level of MAZ, and they were then analyzed. Kaplan-Meier test was utilized to analyze the overall survival (OS) and disease-free survival (DFS) of thyroid cancer patients in The Cancer Genome Atlas (TCGA) database.

A total of 58 pairs of paraffin-embedded papillary thyroid cancer tissue microarray chips were provided by Xinchao Company (Shanghai, China), which were routinely dried, dewaxed, hydrated, and the antigens were repaired under high pressure with citrate repair buffer in 3% hydrogen peroxide. The chips were soaked for 10 min to remove endogenous peroxidase. Then, goat serum was added for blocking for 20 min, diluted primary antibody (MAZ, #ab85725, 1:200; Abcam, Cambridge, UK) was added dropwise, and placed into a refrigerator at 4 °C overnight. On the next day, the secondary antibody (enzyme-labeled anti-mouse/rabbit IgG polymer) was added into drops and placed at room temperature for 30 min. Afterwards, 3,3${}^{\prime }$-diaminobenzidine (DAB) was added for developing the color, and visualization was performed under a microscope for 5–10 min until the appropriate color appeared. Finally, the chips were counterstained with hematoxylin, dehydrated, dried, and fixed.

All immunostained sections were double-blinded, and five fields of view were randomly selected under a high-magnification microscope for scoring. Staining intensity was scored according to the nuclear staining: 0 indicated no intensity, 1 indicated light intensity, 2 indicated medium intensity, and 3 indicated high intensity that was dark brown. More than 500 cells were counted to determine the percentage of immunostained cells in the total cells.

The percentage of positively stained cells was scored as follows: 0 ($\le$5%), 1 (6–25%), 2 (26–50%), 3 (51–75%), and 4 ($>$75%). The product of the two scores was the final score of the staining. In the present study, a score of 4 was indicative of negative expression of MAZ, and a score of $\ge$4 was representative of positive expression of MAZ.

Four thyroid cancer cell lines (human thyroid squamous cell carcinoma cell line SW579, human papillary thyroid cancer cell lines B-CPAP and TPC-1, and human anaplastic thyroid cancer cell line 8505C) were purchased and validated by STR authentication and mycoplasma test from Shanghai Yihe Application Biological Technology Co., Ltd. (Shanghai, China). The cells were cultured in a complete Dulbecco’s modified Eagle’s medium (DMEM), consisting of fetal bovine serum (FBS) and penicillin/streptomycin mixture (Solarbio Science & Technology Co., Ltd., Beijing, China) in an 89:10:1 ratio. The growth conditions of the cells were 5% CO${}_2$ in a 37 °C incubator.

Cells at the logarithmic growth phase were inoculated into six-well plates and transfected with liposome 2000 (Invitrogen, Carlsbad, CA, USA) according to the manufacturer’s instructions when the cells reached a confluence of 80%. After 6 h of transfection, the cells were stored in a fresh medium for subsequent experiments.

TRIzol reagent (Tiangen, Beijing, China) was used to extract the total RNA of thyroid cancer cell line 8505C after 48 h of transfection. Total RNA concentration was measured by Nanodrop 2000 (Thermo Fisher Scientific, Waltham, MA, USA). The transcriptome sequencing bioinformatics analysis was performed by GeneChem Co., Ltd. (Shanghai, China). The libraries were sequenced using a 150 bp double-ended sequencing strategy on the Illumina-HiSeq2000 platform. The raw imaging data were converted from sequence data generated by base calling (Illumina pipeline CASAVA ver. 1.8.2, Illumina Inc., San Diego, CA, USA). Then, quality control standards were used to remove reads that did not meet the following criteria: (1) readings are paired with joints and primers, at most two mismatches; (2) exceeding 5% of unknown bases at the time of reading; (3) over 50% of the low-quality bases. Finally, after quality control, filtered reads were retained for further analysis.

Analysis of DEGs in two groups (two biological replicates per group) was performed using DESeq2 (version 1.16.1, R package, Auckland, New Zealand). DESeq2 shares information across genes to generate more accurate estimates of variation based on the mean expression level of the gene. Benjamini-Hochberg method was adopted to adjust the p and control the false discovery rate. Using DESeq2, it was found that DEGs met p 0.05 after adjustment. Corrected p and $|$log2FC$|$ were used as thresholds for significant DEGs.

The Gene Ontology (GO) enrichment analysis of DEGs was performed by the clusterProfiler package of R 3.4.4 software, in which the gene length bias was corrected. GO terms with corrected p 0.05 indicated significant DEGs. The Kyoto Encyclopedia of Genes and Genomes (KEGG) is a database enabling researchers to systematically analyze gene function and genomes. The clusterProfiler package of R 3.4.4 software was utilized to analyze DEGs in the KEGG pathway.

Total RNA was extracted using TRIzol reagent, and was then reversely transcribed into a cDNA template, and PCR was performed using the SYBR Green Mixture kit (Tiangen). The sequences of primers used for RT-qPCR are shown in Supplementary Table 1. The PCR reaction conditions were summarized as follows: denaturation at 95 °C for 10 min, extension at 95 °C for 15 s, and annealing at 65 °C for 60 s for 40 cycles. Relative RNA expression was calculated using the 2${}^{-\mathrm{\Delta }\mathrm{\Delta }Ct}$ method.

Cells were harvested and lysed using RIPA Lysis Buffer (Beyotime, Shanghai, China). The concentration of total protein was determined by BCA Protein Assay Kit (Beyotime). 10 µg of protein samples were subjected to SDS-PAGE. After that, PVDF membrane was incubated with MAZ and GAPDH (#AF0006, 1:1000; Beyotime) primary antibody to determine the protein expression of MAZ and GAPDH.

Cells were cultured in 6-well plates. After 12 h of culture, the cells were scratched vertically in one direction with a 10 µL tip, and were then washed with phosphate-buffered saline (PBS) before adding fresh serum-free culture medium. The cells were observed and photographed under a microscope at 0, 24, and 48 h, respectively, and the cell migration was measured using ImageJ software (version 1.8.0, National Institutes of Health, Bethesda, MD, USA). The difference was determined between the scratching width with a mobility of 0 h minus the scratching width of 24 or 48 h and divided by the scratching width of 0 h.

Transwell chambers (BD, Franklin Lakes, NJ, USA) with or without Matrigel were used to detect cell migration and invasion. Place the chamber into a 24-well plate, add culture medium, and incubate in an incubator for 30 min. Add 100 µL serum-free culture medium containing 2 $\times$ 10${}^4$ cells to the upper chamber, and add 600 µL of culture medium with 10% FBS to the lower chamber. After 30 h, fix the cells with 4% paraformaldehyde for 40 min, and perform staining with 1% crystal violet for 1 h. After washing with PBS, wipe off the Matrigel and remaining cells with a cotton swab. Take a picture of the cells and count them.

In the present study, SPSS 28.0 software (IBM, Armonk, NY, USA) was used to perform statistical analysis. The comparison between two groups was carried out by the Chi-square test. Multivariate logistic regression was used to analyze the relationship between the expression level of MAZ and clinicopathological features of patients with thyroid cancer. Survival analysis was performed using the Kaplan-Meier method. The Spearman correlation test was used to analyze the correlation between the expression levels of MAZ and FN1 in thyroid cancer tissues. p 0.05 was considered statistically significant.

In order to determine the expression level of MAZ in thyroid cancer tissues, bioinformatics analysis was initially conducted. In TCGA database, the expression level of MAZ in thyroid cancer tissues was significantly higher than that in the adjacent non-tumor tissues (p 0.05, Fig. 1A). Then, to confirm the results of the bioinformatics analysis, 58 paired papillary thyroid cancer and adjacent non-tumor thyroid tissue samples were collected, and IHC was carried out. As shown in Fig. 1B, MAZ protein was mainly located in the nucleus and cytoplasm. The positive expression rate of MAZ in papillary thyroid cancer tissues was 75.86% (44/58), and the positive expression rate in the paired adjacent non-tumor tissues was 18.96% (11/58). The difference in the expression level of MAZ between papillary thyroid cancer and adjacent non-tumor tissues was statistically significant (p 0.001, Table 1). The Wilcoxon test was used to compare the immunostaining scores of MAZ in papillary thyroid cancer and adjacent non-tumor tissues. The results showed that the MAZ staining scores in the adjacent non-tumor tissues were lower than those in the papillary thyroid cancer tissues. The median difference between the two groups was –3.5 (95% confidence interval (CI): –4~–2.5, p 0.001, Fig. 1C). These results suggested that the expression level of MAZ was upregulated in papillary thyroid cancer tissues.

Fig. 1.

Expression of MAZ in thyroid cancer tissues. (A) Expression of MAZ in thyroid cancer and adjacent tissues in TCGA database. (B) Representative results of immunohistochemical staining (red arrows indicate the position of image at $\times$400 magnification). (C) Staining score of MAZ in papillary thyroid cancer tissues and adjacent tissues (n = 58). *, p 0.05, and ***, p 0.001. MAZ, Myc-associated zinc finger protein; TCGA, The Cancer Genome Atlas.

To further explore the effects of the expression level of MAZ on the prognosis of patients with thyroid cancer, the relationship between clinicopathological characteristics and the expression level of MAZ in 58 papillary thyroid cancer patients was clarified. Clinicopathological variables presented in Table 2 were included in the univariate analysis. The results showed that the expression level of MAZ was correlated with tumor, node, and metastasis (TNM) stage, tumor diameter, and tumor capsule (p 0.05, Table 2). Then, variables with p 0.1 in univariate analysis were imported into the multivariate logistic regression analysis. As a result, it was found that the expression level of MAZ was significantly associated with tumor diameter (OR = 6, 95% CI: 1.2~29.998, p = 0.029) and tumor capsule (OR = 8.4, 95% CI: 1.345~52.476, p = 0.023) (Table 3).

According to the clinical data downloaded from TCGA database, the Kaplan-Meier method was used to analyze the relationship between the expression level of MAZ and the prognosis of patients with thyroid cancer. According to the expression level of MAZ, all patients were divided into two groups. Compared with the low MAZ expression group, thyroid cancer patients in the high MAZ expression group had lower OS and DFS (p 0.05, Fig. 2A,B). These results demonstrated that the expression level of MAZ was correlated with prognosis of thyroid cancer patients, indicating the role of the expression level of MAZ in thyroid cancer progression.

Fig. 2.

Relationship between MAZ expression and thyroid cancer prognosis. (A) Kaplan-Meier plotter analysis of OS and (B) DFS of MAZ was conducted in patients with thyroid cancer. TPM, transcripts per million

To investigate the biological functions of MAZ in thyroid cancer cells, the mRNA expression levels of MAZ in four thyroid cancer cell lines (SW579, B-CPAP, TPC-1, and 8505C) were detected by RT-qPCR. It is noteworthy that 8505C cell line was selected for the following experiment, because it had the highest MAZ mRNA expression level among the four cell lines (Fig. 3A). Two short hairpin RNAs (shRNAs) targeting MAZ were purchased from GeneChem Co., Ltd., including KD-1# and KD-2#, to knockdown the expression level of MAZ in cells. Two shRNAs were introduced into 8505C cells, and the expression level of MAZ in 8505C cells was detected after 48 h. Compared with the negative control (NC) group, both the mRNA and protein expression level of MAZ in the two groups (KD-1# or KD-2#) was significantly reduced (p 0.01, Fig. 3B). Therefore, the two shRNAs could effectively interfere with the expression level of MAZ in thyroid cancer cells.

Fig. 3.

Role of MAZ in metastasis of thyroid cancer. (A) RT-qPCR detection of relative expression of MAZ mRNA in thyroid cancer cell lines. (B) MAZ mRNA and protein expression after MAZ knockdown in 8505C cells. (C,D) Wound healing assay. (E,F) Transwell assay. *, p 0.05, and **, p 0.01. NC, negative control; KD, knock-down.

The migratory ability of 8505C cells with MAZ knockdown was detected by wound healing assay. Compared with the NC group, the cell migratory ability was significantly reduced after MAZ knockdown in the KD-1# and KD-2# groups, and the wound healing speed in the KD-1# group was slower than that in the KD-2# group (p 0.05, Fig. 3C,D). The Transwell assay was performed in 8505C cells with MAZ knockdown to confirm MAZ role in metastasis of thyroid cancer cells. As shown in Fig. 3E,F, compared with the NC group, the invasive and migratory abilities in the KD1# and KD2# groups were significantly reduced, and the effect was more significant in the KD1# group (p 0.05). These results indicated that MAZ could play a role in the regulation of thyroid cancer metastasis.

RNA sequencing of thyroid cancer cells after MAZ knockdown was performed using the GeneChip Human Gene 1.0 ST array. Then, according to the normalized expression levels, the correlation values between each sample were calculated and the correlation heat map was drawn. As shown in Fig. 4A, downregulation of MAZ changed the expression levels of various genes at the transcriptome level. Compared with the NC cells, there were 25,535 DEGs in MAZ knockdown cells, including 2129 upregulated genes and 2237 downregulated genes (Fig. 4B). To understand the relevant functions of DEGs in thyroid cancer after MAZ downregulation, GO enrichment analysis was conducted based on biological process (BP), cell component (CC), and molecular function (MF). The results indicated that MAZ-related DEGs were mainly enriched on ribonucleoprotein complex biogenesis, ribosome biogenesis, and rRNA metabolism in BP; in CC, they were mainly associated with focal adhesion, cell-substrate junction, and cell-substrate adhesion; in MF, they were mainly correlated with cadherin binding, structural composition of ribosomes, cell adhesion molecule binding, and mRNA binding (Fig. 4C). The results of KEGG pathway analysis showed that MAZ-related DEGs were mainly enriched in cancer pathway, neurotrophic factor signaling pathway, regulation of actin cytoskeleton, apoptosis pathway, protein processing in endoplasmic reticulum, and other pathways (Fig. 4D). The GO and KEGG pathway enrichment analyses further confirmed the involvement of MAZ in thyroid cancer progression.

Fig. 4.

RNA-seq results of MAZ knockdown thyroid cancer cells. (A) Heatmap of the differentially expressed genes (DEGs). (B) Volcano plot. (C) The Gene Ontology (GO) enrichment analysis. (D) The Kyoto Encyclopedia of Genes and Genomes (KEGG) enrich distribution point diagram. BP, biological process; CC, cell component; MF, molecular function.

EMT-related factors participate in the metastatic progression of a great deal of malignant tumors. The expression level of EMT-related factor FN-1 was obtained from RNA-seq data of MAZ knockdown in thyroid cancer cells. The expression level of FN1 after MAZ downregulation was then examined to determine the EMT process. As shown in Fig. 5A, RT-qPCR assay showed that the expression level of FN1 was elevated in MAZ knockdosswn in thyroid cancer cells, and the expression level of FN1 in KD1# groups was higher than that in KD2# group (p 0.05). Moreover, the expression level of MAZ in thyroid cancer tissues was negatively correlated with the expression level of FN1 (r = –0.31, p 0.001, Fig. 5B). Data downloaded from TCGA database revealed that the expression level of FN1 in thyroid cancer tissues was higher than that in adjacent thyroid tissues (p 0.05, Fig. 5C). These results indicated that MAZ may promote thyroid cancer metastasis through regulating the expression level of FN1.

Fig. 5.

Relationship between MAZ expression and FN1 expression in thyroid cancer. (A) RT-qPCR detection of relative expression of FN1 mRNA after MAZ down-regulation. (B) Correlation analysis between MAZ and FN1. (C) Expression of FN1 in thyroid cancer and adjacent tissues in TCGA database. *, p 0.05. FN1, fibronectin 1.