Human primary clear cell renal cell carcinoma and corresponding normal kidney tissue were collected from the Department of Urology, the Second Hospital of Hebei Medical University. All patients were ccRCC patients from July 2015 to June 2020. All patients underwent radical nephrectomy for treatment. Fresh ccRCC and corresponding normal kidney tissue were washed with saline and quickly frozen in liquid nitrogen or formaldehyde to fix for further used. The research protocol has been approved by the Ethics Committee of the Second Hospital of Hebei Medical University, and each patient’s written consent has been obtained.

Human ccRCC cell lines (SW839, A498, Caki-1, 786-0) were purchased from ATCC (Rockville, Maryland, USA), and are preserved in our laboratory. The 293A cell line is deposited in our laboratory. The above-mentioned cells were cultured in low-sugar dulbecco’s modified eagle medium (DMEM) supplemented with 10% fetal bovine serum (Clark Bio, Claymont, DE, USA) and 1% penicillin/streptomycin (Solarbio, Beijing). The cells are cultured in a humidified condition of 95% air and 5% CO₂. All cells were transfected using Lipofectamine 2000 (Invitrogen) according to the manufacturer’s operating manual. Briefly (27), 786-O and SW839 cells (1×10⁵ cells/ml) were seeded into plates in growth medium (DMEM). Until the required number of cells (80% confluence) is obtained at the time of transfection. The cells were washed twice with PBS and the transfected lipoplex (prepared as described in the operating manual) was added to each well. It was mixed gently by rocking the plate back and forth. The cells were cultured in DMEM (without FBS), 5% CO₂, and 37°C humidified incubator for 4-6 hours. Then the cells were washed twice with PBS, replaced with complete medium, and cultured for 48 hours.

Use RNAeasy Mini Elute Kit (QIAGEN) according to the manufacturer’s manual for total RNA isolation. The NanoDrop 2000 system detects RNA concentration and quality. The first strand of cDNA was synthesized using M-MLV First Strand Kit (Life Technologies) with random hexamer primers. Dilute the first strand of cDNA 5-10 times as needed, and use Platinum SYBR Green qPCR Super Mix UDG kit (Invitrogen) to perform real-time quantitative PCR (qRT-PCR) analysis on mRNA in ABI 7500 FAST System (Life Technologies). The relative transcriptional expression level of gene mRNA was standardized with GAPDH as an internal reference gene. The calculation was carried out using the 2^-ΔΔCt formula according to the previous description (28). ZNF304-F:GACCGGGTTCAGAGTTGTGT; ZNF304-R:CCACGTGTGCACAGTTTCTG; FOXO4-F:GGGAAAAGGCCATTGAAAGCG; FOXO4-R:TGTGGCGGATCGAGTTCTTC; GAPDH-F:ATGAATGGGCAGCCGTTAGG; GAPDH-R:TGGAATTTGCCATGGGTGGA; Cyclin D1-F:CTGATTGGACAGGCATGGGT; Cyclin D1-R:GTGCCTGGAAGTCAACGGTA; miR-23-3p-F:GGCACATTGCCAGGGATTTC; miR-96-5p-F:GGTTTGGCACTAGCACATTTTTG; miR-128-3p-F:GGCTCACAGTGAACCGGTCTC; miR-142-3p-F:GGCCGTAGTGTTTCCTACTTTATG; miR-150-5p-F:GCTCTCCCAACCCTTGTACCAG; miR-183-5p-F:GGCTATGGCACTGGTAGAATTCAC; miR-216b-5p-F:GCAAATCTCTGCAGGCAAATGTG; miR-217-F:GTACTGCATCAGGAACTGATTGG; miR-449a-5p-F:GGTGGCAGTGTATTGTTAGCTGG; miR-1271-5p-F:CTTGGCACCTAGCAAGCACTC.

Western blot analysis was performed as previously described (28). Polyvinylidene fluoride (PVDF) membranes (Millipore), were incubated with primary antibodies anti-ZNF304 (1:500, HPA050531), anti-FOXO4 (1:1000, ab128908), anti-cyclin D1 (1:500, 26939-1-AP), and anti-β-actin (1:1000, sc-47778). An HRP-conjugated secondary antibody (1:5000, Rockland) was used and detection was by ECL (enhanced chemiluminescence) Fuazon Fx (Vilber Lourmat). Images were captured and processed by FusionCapt Advance Fx5 software (Vilber Lourmat). All experiments were repeated three times.

Lentiviral pLKO-ZNF304 (shZNF304), pWPI-ZNF304 (overexpression ZNF304, oeZNF304) and pLKO-FOXO4 (shFOXO4) plasmids were designed and constructed from Biocaring Biotechnology Co., Ltd (Shijiazhuang). Briefly (28), ZNF304 cDNA was inserted into EcoRI and XhoI restriction endonuclease digested pWPI empty vector (Addgene, #12254) and then Sanger sequencing for confirming. ZNF304 and FOXO4 shRNA were inserted into BsmBI restriction endonuclease digested pLKO (Addgene, #14748) and then Sanger sequencing for confirming. The 3’ UTR sequences of FOXO4 containing wild-type (WT) or mutant (mut) forms of the miR-183-5p target site were inserted into restriction endonucleases XhoI (Xho1) and SalI (Sal1) along with pmirGLO Dual-Luciferase miRNA Target Expression Vector (Promega, Madison, WI, USA). FOXO4 reporter construct (WT or mut) or the empty reporter vector were co-transfected into SW839 cells with miR-183-5p mimic or mimic control and plasmid Renilla luciferase-thymidine kinase. Luciferase activity was measured by Dual-Glo Luciferase Assay System (Promega) with a Flash and Glow reader (LB955; Berthold Technologies, Bad Wildbad, Germany). The specific target activity was expressed as the relative activity ratio of firefly luciferase to Renilla luciferase.

Human ccRCC and normal kidney tissues were fixed in formalin and then processed for routine paraffin embedding (29). Ten 5-μm-thick consecutive sections were prepared for hematoxylin and eosin staining. Images were acquired using a Leica microscope (Leica DM6000B, Switzerland) and digitized with LAS V.4.4 (Leica).

Cell viability was detected by an MTT [3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide] colorimetric assay as previously described (29). Briefly, SW839 and 786-0 cells were plated and transfected with the indicated vector or treated with AZD6244 for 24 h. Absorbance at 570nm was measured using a microplate reader (Thermo Fisher, USA).

100 cells/well were seeded into six-well plates, cultured for 1 week and fixed in a glacial acetic acid/methanol solution. Then, 0.5% crystal violet was used to stain the colonies. Colony numbers were counted under a microscope (30).

The chromatin immunoprecipitation (ChIP) assay was performed as previously described (30). Briefly, SW839 cells were treated with formaldehyde. The cross-linked chromatin was sonicated to an average size of 400–600 bp. Samples were precleared with protein A-agarose/salmon sperm DNA and the DNA fragments were immunoprecipitated overnight at 4°C with anti-ZNF304 or anti-IgG antibodies. After reversal of cross-linking, ZNF304 occupancy on the miR-183-5p promoter was examined. Results were determined by qRT-PCR with miR-183-prom-F1: GGGCCTTCAGGTGGAGATAGGAG; miR-183-prom-R1:CCCTGCGGAGAGACCAGCGG; miR-183-prom-F2:CAGTCTGGCCCAATCTGGTCTGG; miR-183-prom-R2:GGAGAGGCAAGCAGCTGGCCAG; miR-183-prom-F3:GGGTGCCAGCTCCCCAGAGACC; miR-183-prom-R3:CTGCAGCCACCCAGGTGGCTTG.

Statistical analysis as previously described (28). Data were presented as the mean ± SEM. Student’s t test was used to analyze differences between two groups. Spearman’s correlation analysis was used to evaluate the correlation analysis. Values of p < 0.05 were considered statistically significant. GraphPad Prism 7.0 software was used for the statistical analysis (GraphPad Software).

To investigate the expression level of ZNF304 in ccRCC, firstly, we collected clinical samples and located the ccRCC tumor and an area of normal kidney tissue by using hematoxylin and eosin staining ( Figure 1A ). Compared with normal kidney tissue, the mRNA and protein levels of ZNF304 in ccRCC tissue (T) are significantly reduced ( Figures 1B–D ). Data from the TCGA database show that the mRNA expression level of ZNF304 in ccRCC tissue is significantly lower than that in normal kidney tissue ( Figure 1E ). And the higher levels of ZNF304 indicates a good prognosis for patients in patients with ccRCC ( Figure 1F ). Furthermore, clinicopathologic factors of ZNF304 mRNA expression level was markedly negative associated with tumor grade but not with age, sex and tumor size etc ( Table 1 ). Next, we detected ZNF304 expression in cell lines. 786-0 and SW839 cells expressed much lower levels of ZNF304 compared with the other cell lines ( Figures 1G–I ). These results suggest that downregulation of ZNF304 may be associated with ccRCC progression.

Previous studies revealed ZNF304 to be a transcriptional repressor involved in cancer cell survival (18). Overexpression of ZNF304 with a lentiviral vector markedly increased ZNF304 mRNA and protein levels, whereas transfection with shZNF304 significantly reduced ZNF304 expression ( Figures 2A–C ). The MTT assay showed that ZNF304 overexpression inhibited growth of both 786-0 and SW839 cells, whereas ZNF304 downregulation accelerated cell line proliferation ( Figure 2D ). The colony formation assay confirmed these findings ( Figures 2E, F ). Taken together, these data reveal that ZNF304 inhibits ccRCC cell growth.

To elucidate the molecular pathway underlying ccRCC cell growth regulation by ZNF304, we analyzed the correlation of gene expression with ZNF304 in ccRCC tissues from the TCGA database ( Supplementary Figure 1 ). The transcription factor FOXO4 appeared to be implicated in this. Knockdown of ZNF304 reduced FOXO4 protein level while ZNF304 overexpression upregulated FOXO4 ( Figures 3A, B ). FOXO4 was also detected in ccRCC and normal kidney tissues. Compared to normal kidney tissues, the expression of FOXO4 was obviously decreased in ccRCC tissues (T) ( Figure 3C ). The data from TCGA databases also show that FOXO4 mRNA levels were downregulated in ccRCC ( Figure 3D ). And the lower expression level of FOXO4 was associated with poor overall survival in ccRCC patients ( Figure 3E ). Moreover, we confirmed that FOXO4 has a positive correlation with ZNF304 in ccRCC using PCR data ( Figure 3F ). The MTT assay showed that depletion of FOXO4 facilitated cell proliferation and reversed the inhibition caused by ZNF304 overexpression ( Figure 3G ). Surprisingly, overexpression or depletion of ZNF304 did not affect FOXO4 mRNA levels ( Figure 3H ). These findings show that FOXO4 underlies the regulation of ccRCC cell proliferation by ZNF304.

Regulation of FOXO4 protein expression level by ZNF304, without influence in FOXO4 mRNA, suggests that ZNF304 regulates FOXO4 at the epigenetic regulation. Because microRNAs (miRNAs) are one of critical regulators in the regulation of gene expression at the posttranscriptional level, we predicted miRNAs that may target the FOXO4 3′-UTR from TargetScan, miRanda, and RNA22 database ( Figure 4A ). FOXO4 3′-UTR containing biotin-labeled uracil was generated by a T7 RNA transcriptase. Then, we performed a pull-down assay combined with RT-PCR to detect the enrichment of these candidate miRNA levels ( Figure 4B ). The result showed that 6 high enrichment miRNAs were changed after up- and downregulation of ZNF304. However, only miR-183-5p was increased in shZNF304 transfected cells and decreased in ZNF304 overexpressing cells ( Figure 4C ). In addition, the luciferase reporter assay demonstrated that miR-183-5p can target the 3′UTR of FOXO4 ( Figures 4D, E ). Furthermore, western blot also confirmed that FOXO4 was a directly target of miR-183-5p ( Figures 4F, G ). miR-183-5p mimic facilitated cell proliferation and reversed cell growth inhibition caused by ZNF304 overexpression ( Figure 4H ). Totally, these finding indicated that miR-183-5p, which was regulated by ZNF304, directly targets 3′URT and repressed FOXO4 expression.

In order to confirm whether ZNF304 directly regulated miR-183-5p expression through transcription, we first analyzed the potential ZNF304 binding sites in the miR-183 2-kb 5’-promoter region on the nsembl and PROMO 3.0 websites. The results showed that there were three potential binding sites for ZNF304 in this region. ( Figure 5A ). ChIP analysis proved that ZNF304 mainly bound to the -80 to -65 bp region of the transcription start site in the miR-183 promoter ( Figure 5B ). Furthermore, in RCC cells where ZNF304 was downregulated pre-miR-183-5p expression was promoted and ZNF304 downregulation inhibited pre-miR-183-5p expression ( Figure 5C ). Compare to the normal kidney tissues, miR-183-5p level was significantly upregulation in ccRCC tissues (T) ( Figure 5D ). And the higher expression level of miR-183-5p in ccRCC patients were associated with poor overall survival ( Figure 5E ). Additionally, the mRNA level of miR-183-5p was positively correlated with ZNF304 in ccRCC tissues ( Figure 5F ). These findings reveal that ZNF304 directly promotes miR-183-5p transcription.

Overexpression of ZNF304 increased FOXO4 protein level and reduced cyclin D1 level, and the regulation of FOXO4 was enhanced by miR-183-5p inhibitor co-transfection ( Figures 6A, B ). Depletion of FOXO4 promoted cyclin D1 expression, while miR-183-5p inhibitor co-transfection reversed the promotion of FOXO4 ( Figures 6C, D ). Finally, we used a xenograft model to examine whether disruption of ZNF304/miR-183-5p/FOXO4 axis inhibits cell growth in vivo. Injection of 786-0 cells with stable depletion of miR-183-5p, to block ZNF304/miR-183-5p/FOXO4 axis, yielded smaller tumors in nude mice than in those injected with sham-transfected cells ( Figures 6E, F ). Similarly, stable knockdown of miR-183-5p also resulted in a smaller wet tumor weight ( Figure 6G ). Western blot analysis also revealed that silence of miR-183-5p significantly increased FOXO4 but reduced cyclin D1 proteins level compared to tumors derived from control cells ( Figure 6H ). These findings suggest that disruption of ZNF304/miR-183-5p/FOXO4 axis inhibits cell proliferation in ccRCC cells in vivo.