m6A-mediated upregulation of miR-3690 drives HNSCC progression by regulating nuclear-cytoplasmic signaling pathway

Abstract

MicroRNAs (miRNAs) are involved in carcinogenesis. However, the biological roles and underlying mechanism of miR-3690 in head and neck squamous cell carcinoma (HNSCC) progression are far from elucidated. In this study, we found that the expression level of miR-3690 in HNSCC tissues was significantly higher and correlated with poor clinical prognosis. HNSCC cells proliferation, migration, and invasion were promoted by miR-3690 overexpression, both in vitro and in vivo. Mechanistically, miRNA pulldown DNA-seq and luciferase reporter assays revealed that miR-3690 could directly activate CKS2 expression through targeting its promoter. Meanwhile, RNA pull down and mass spectrometry (MS) analysis suggested that upregulation of CKS2 by miR-3690 correlated with increased BPTF occupancy and H3K4me3 at CKS2 promoter. Additionally, luciferase reporter assays showed that miR-3690 facilitated Wnt/β-catenin signaling in HNSCC by repressing NKD1 expression through directly targeting its 3’-UTR. Finally, methylated RNA Immunoprecipitation (meRIP) and RNA pull down indicated that METTL3 and METTL14-mediated m6A modification accelerated pri-miR-3690 maturation through regulating the processing of pri-miR-3690 by DGCR8. In conclusion, miR-3690 may be a prognostic indicator and potential therapeutic target for HNSCC.

Supplementary Information

The online version contains supplementary material available at 10.1007/s00018-025-05875-7.

Introduction

Head and neck squamous cell carcinoma (HNSCC) is the sixth most common cancer worldwide with an annual incidence of approximately 650,000 new cases and 350,000 deaths [1]. It is more common in males than in females and is usually diagnosed between 50 and 70 years of age [2]. The multidisciplinary treatment of HNSCC consists of surgery, anticancer drug therapy, and radiotherapy. Although therapeutic strategies have advanced rapidly in recent years, the 5-year overall survival rate of patients with HNSCC remains approximately 66% [1]. The major clinical challenges associated with HNSCC therapy include the occurrence of distant metastases in 10–30% and tumor relapse in 30–50% of patients [3], often resulting in treatment failure. Therefore, identifying novel targets for the diagnosis and treatment of HNSCC is imperative.

MicroRNAs (miRNAs), a class of single-stranded non-coding RNAs composed of 21–25 nucleotides, modulate neoplastic processes in a variety of cancers, including HNSCC [4]. Their dysregulation has been used to diagnose and predict HNSCC prognosis. MiRNA-targeted therapeutics could also offer an alternative strategy for the prevention and treatment of HNSCC, especially in patients for whom few or no treatment options are available. For example, miR-130b-3p is dramatically elevated in HNSCC tissues and enhances the angiogenic and tumorigenic abilities of cancer cells [5], and alcohol-induced dysregulation of miR-30a and miR-934 may play crucial roles in HNSCC pathogenesis and progression [6]. However, these reported miRNAs downregulated the target genes during post-transcriptional processes by targeting the 3’-UTR to degrade mRNA or suppress translation. Apart from that, miRNAs could also target 5’-UTRs [7], promoters [8], or coding regions [9] to silence the gene. Moreover, miRNAs also exert positive effects on gene expression, thus adding to the complexity of miRNA-mediated gene regulation. For example, miRNA-551b-3p could directly upregulate STAT3 expression to promote growth and metastasis of ovarian cancer [10], and miR-10a has been shown to enhance translation by interacting with the 5’-UTR of ribosomal protein mRNAs [11]. However, few studies have focused on the functional importance of miR-3690 in the pathogenesis of HNSCC. The molecular mechanisms underlying the role of miR-3690 in HNSCC progression remain unclear, and further studies are required to explore these aspects.

In the present study, we investigated the expression of miR-3690 in HNSCC tissues and determined its potential prognostic value particularly the underlying mechanism by which miR-3690 facilitates the progression of HNSCC.

Methods and materials

Patient samples

All HNSCC and corresponding normal tissues were sourced from the Head and Neck Surgery Department of the Eye and ENT Hospital, Fudan University. As soon as the tissues were collected, they were stored at −80 °C until use. Tissues were collected for qRT-PCR, western blotting, fluorescence in situ hybridization, immunofluorescence, and immunohistochemistry analysis. This study was approved by the Ethics Committee of the Fudan University Eye & ENT Hospital (No. 2018036). Written consent was obtained from all patients.

Cell culture

Human HNSCC cell lines AMC-HN-8 and TU686 (Chinese Academy of Science Cell Bank, Shanghai, China) were used in this study. All cells were cultured in RPMI-1640 medium (HyClone, USA) supplemented with 10% fetal bovine serum (FBS, Gibco, USA) at 37 °C in 5% CO2.

Western blot

RIPA lysis buffer (Cat#P0013B, Beyotime) consists of 50mM Tris (pH 7.4), 150mM NaCl, 1% Triton X-100, 1% sodium deoxycholate, 0.1% SDS, supplemented with protease and phosphatase inhibitors including sodium orthovanadate, sodium fluoride, EDTA, and leupeptin. Total cell or tissue proteins were extracted using RIPA lysis buffer containing 1% phenylmethanesulfonyl fluoride (PMSF, Cat#ST506, Beyotime). The proteins were separated by SDS-PAGE (Beyotime, China), transferred onto PVDF membranes (Millipore, USA), and incubated overnight with the primary antibodies at 4 °C. The following diluted primary antibodies were used: β-actin (Cat#ab8226), MMP7 (Cat#ab207299), MYC (Cat#ab32072), Axin-2 (Cat#ab109307), β-catenin (Cat#ab32572) were obtained from Abcam (1:1000), E-Cadherin (Cat#3195T), N-Cadherin (Cat#13116T), Vimentin (Cat#5741T), Snail (Cat#3879T) were obtained from Cell Signaling Technology (1:1000), HRP-labeled Goat Anti-Rabbit IgG(H + L) (Cat#A0208) and (HRP-labeled Goat Anti-Mouse IgG(H + L) (Cat#A0216) were obtained from Beyotime (1:1000). An electrochemiluminescence kit (Tanon Science & Technology Co.) was used to visualize the target protein bands. β-actin served as a control.

Quantitative reverse transcription PCR (qRT-PCR)

Total RNA was extracted from cells or tissues using TRIzol reagent (Thermo Fisher Scientific) according to the manufacturer’s protocol. The mRNA was reverse-transcribed into cDNA using the RevertAid First Strand Synthesis Kit (Thermo Fisher Scientific, USA), and the target genes were amplified with gene-specific primers. To conduct qRT-PCR analysis of miRNAs, the All-in-One™ miRNA qRT-PCR kit (GeneCopoeia Inc., USA) was used to reverse-transcribe total RNA into cDNA, which was amplified using primers specific for miRNAs. The primers were as follows: β-actin, forward: 5′-GCACTCTTCCAGCCTTCCTTCC-3′ and reverse :5′-GCGGATGTCCACGTCACACTTC-3; CKS2, forward :5′-ACGAGTACCGGCATGTTATGTTACC-3′ and reverse :5′-TCCTCCACTCCTCTTCAGACATCAG-3; BPTF, forward :5′-GTCAACAAAGTGGTGTACGATGA-3′ and reverse :5′-TGCAGTAACTGGCGTCGTC-3; Myc, forward :5′-AGCAGCGACTCTGAGGAGGAAC-3′ and reverse :5′-TCCAGCAGAAGGTGATCCAGACTC-3; MMP-7, forward :5′-GAGGATGAACGCTGGACGGATG-3′ and reverse :5′-AGGATCAGAGGAATGTCCCATACCC-3; Axin-2, forward :5′-CACCACCACCATTCGCAGTACC-3′ and reverse :5′-ACATGCTTCGTCGTCTGCTTGG-3; NKD1, forward :5′-CGCCTTGGTGGTGTATGAGAGC-3′ and reverse :5′-AGTGGTGGTAATGGTGGTGATGTTC-3; and Pri-miR-3690, forward :5′-CCCAGTGTAGACAGGAGAAGACTC-3′ and reverse :5′-GCAGGGTCTCCTCCCATCTA-3. ’ Primers for miR-3690 (HmiRQP1976) and U6 (HmiRQP9001) were purchased from GeneCopoeia (Rockville, MD, USA). The mRNA and miRNA expression levels were normalized to actin and U6 levels, respectively.

Transwell migration and invasion assays

The migration and invasion of cells were studied in 24-well transwell chambers equipped with an insert made of 8 μm pore polycarbonate membranes (Corning, USA). After miR-3690 overexpression or knockdown lentiviruses transfection, 5 × 10⁴ cells with 200 µl of serum-free media were added to the upper compartment and 600 µl of complete media was added to the lower compartment. Invasion assays were conducted in a chamber pre-coated with Matrigel (BD Biosciences) in serum-free medium at a ratio of 1:8. Following incubation for 24 h for the migration assay and 36 h for the invasion assay, migratory and invasive cells that adhered to the lower chamber membrane were stained with crystal violet (Beyotime, China). A Leica microscope was used to quantify the cell numbers.

5-Ethynyl-2′-deoxyuridine (EdU) assay: The EdU assay was conducted using an EdU kit (C0075S, Beyotime, China). According to the reagent instructions, dilute EdU (10 mM) 1:500 with cell culture medium to prepare the EdU working solution (20 µM). The transfected cells were co-cultured with EdU working solution at 37 °C in 5% CO2 for 2 h. After fixing with 4% paraformaldehyde for 15 min, the cells were washed three times with PBS and permeabilized with 0.1% Triton X-100 for 10 min. Following the manufacturer’s protocol, the cells were incubated with click reaction solution for 30 min in the dark. The nuclei were counterstained with diamidino-2-phenylindole (DAPI). A Leica microscope was used to capture the images.

Immunohistochemistry and Immunofluorescence

For immunohistochemistry, tissues from patients with HNSCC and tumor-bearing nude mice were fixed in 4% paraformaldehyde prior to paraffin embedding, and cut into 4 μm thick sections. The tissue sections were dewaxed by baking at 60 °C for 30 min followed by immersion in xylene I and II (10 min each) for complete paraffin removal. Rehydration was then performed through a graded ethanol series (100%→95%→80%→70%, 5 min each) with a final rinse in distilled water to complete the process. And then were blocked overnight with 10% BSA before being incubated overnight with primary antibodies (CKS2, Cat#ab155078, Abcam, 1:100; BPTF, Cat#ab288159, Abcam, 1:500; METTL3, Cat#ab195352, Abcam, 1:500; METTL14, Cat#ab220030, Abcam, 1:1000) at 4 °C, followed by incubation with secondary antibodies (HRP-labeled Goat Anti-Rabbit IgG(H + L), Cat#A0208, Beyotime, 1:50; HRP-labeled Goat Anti-Mouse IgG(H + L), Cat#A0216, Beyotime, 1:50) and 3,3-diaminobenzidine (DAB) visualization. Hematoxylin and eosin (H&E) were used as a counterstain. Images were acquired using a microscope (Leica). For immunofluorescence, the cells were cultured on slides for 48 h and fixed with 4% paraformaldehyde, permeabilized with 0.1% Triton X-100, blocked with 1% BSA, incubated with the primary antibody (Ecadherin, Cat#ab40772, Abcam, 1:500; Vimentin, Cat#ab92547, Abcam, 1:100;Snail, Cat#ab224731, Abcam, 1:500; Ki67, Cat#ab16667, Abcam, 1:250; BPTF, Cat#ab288159, Abcam, 1:50; β-catenin, Cat# ab32572, Abcam, 1:250) at 4 °C overnight, followed by incubation with the appropriate secondary antibodies (AF488-labeled Goat Anti-Rabbit IgG (H + L), Cat#A0423, Beyotime, 1:500; AF555-labeled Donkey Anti-Mouse IgG (H + L), Cat#A0460, Beyotime, 1:500). Nuclei were counterstained with DAPI antifade mounting medium (Beyotime, China). Images were captured under a microscope (Leica DM IL LED, Germany).

The slides stained by IHC were assessed by two pathologists who had no knowledge of the clinical information. The staining intensity of the slides was graded as 0 indicates negative; 1 indicates weak; 2 indicates moderate; 3 indicates strong staining. The percentage of positive cells was graded as 0, 0–5%; 1, 6–25%; 2, 26–50%, 3, 51–75%; 4, > 75%. The histological score (H-score) is computed by multiplying the staining percentage with the intensity score.

Luciferase reporter assay

HNSCC cells transfected with lentiviruses or 293 T cells were seeded into 24-well plates. Cells were transfected with plasmids or siRNAs. Cells were collected after 48 h and analyzed for luciferase activity using the Dual-Luciferase activity with a Dual-Luciferase^® Reporter Assay System (Promega, USA).

RNA pull-down

HNSCC cells (2.0 × 10⁷) were collected and lysed (Cat#Bes5109, Bersin Bio) by adding lysis buffer, protease inhibitor, RNase inhibitor, and DTT, followed by incubation on ice for 20 min with intermittent vortexing (3 times, 5 s each). After lysis, the sample was centrifuged at 13,000 g for 5 min at 4 °C, and the supernatant was carefully transferred to the RNase-free tube. Biotinylated pri-miR-3690 and miR-3690 probes were synthesized by Bersin Bio (Guangzhou, China) and incubated with the total protein extracts. RNA probe–protein complexes were obtained by coupling streptavidin magnetic beads. For protein elution, Protein Elution Buffer and DTT were added to the bead-bound complexes, followed by gentle pipetting to mix. The mixture was incubated at 37 °C for 2 h with intermittent mixing to promote elution. After incubation, the tube was placed on a magnetic rack for 1 min to immobilize the beads, allowing the supernatant containing the eluted proteins to be carefully transferred to a new centrifuge tube. Proteins were identified using silver staining and collected for mass spectrometry and western blotting.

Transfection

The siRNAs against METTL3 (5′-CUGCAAGUAUGUUCACUAUGA-3′), METTL14 (5′-AAGGAUGAGUUAAUAGCUAAA-3′), and BPTF (5′-GCCATATCCTTCTCCTAGA-3′) and control siRNAs were purchased from Genomeditech (Shanghai, China), NKD1 siRNA (SC-93414) and negative control siRNA (Cat. SC-36869) were purchased from Santa Cruz Biotechnology (Dallas, TX, USA). NKD1 overexpression and empty plasmids were purchased from GeneChem (Shanghai, China). Transfection with siRNAs or plasmids was performed using Lipofectamine 2000 (Thermo Fisher Scientific). MiR-3690 overexpression and knockdown lentiviruses (5′- CTTTGTCTACGCTGGGTCCAGGT-3′) were obtained from GeneChem (Shanghai, China) and transfected with polybrene (5 µg/mL).

Methylated RNA immune precipitation (MeRIP)

MeRIP assays were performed according to the manufacturer’s instructions using a MeRIP kit (Bes5203-1, BersinBio, China). Briefly, RNA was isolated from HNSCC cells and fragmented by sonication. The RNA fragments were incubated with an m6A antibody conjugated to protein A/G magnetic beads. The bound RNA was extracted using phenol-chloroform-isoamylol after the beads were washed three times with the elution buffer. qRT-PCR was used to analyze the immunoprecipitated RNA.

RNA Immunoprecipitation assay (RIP)

RIP assays were performed using a RIP kit (Bes5101; Bersin Bio, China). Briefly, the cells were lysed with RIP lysis buffer. Anti-METTL3 antibody (5 µg, ab195352, Abcam), anti-METTL14 antibody (5 µg, A8530, ABclonal), anti-DGCR8 antibody (5 µg, ab191875, Abcam) or normal rabbit IgG (5 µg, Bes5101, Bersin Bio) conjugated protein A/G beads were incubated with the lysates of the cells at 4 °C overnight. For pri-miR-3690 extraction and purification from IP samples containing magnetic beads, TRIzol was added and mixed by inversion for 10 s followed by 5 min incubation on ice. After adding chloroform, samples were vigorously shaken for 15 s and incubated on ice for 2 min. Following centrifugation at 13,000 g for 10 min at 4 °C, the aqueous phase was transferred to a new RNase-free tube. RNA was precipitated by adding glycogen, sodium acetate and 100% ethanol with thorough mixing, then incubated at −80 °C overnight. After centrifugation at 16,100 g for 30 min at 4 °C, the supernatant was discarded. The pellet was washed with chilled 75% ethanol, centrifuged again at 16,100 g for 10 min at 4 °C, air-dried for 10 min at room temperature, and finally dissolved in RNase-free water and pri-miR-3690 was examined using qRT-PCR.

Fluorescence in situ hybridization (FISH)

Probes specific for miR-3690 were designed and synthesized by Service Bio (Wuhan, China). FISH was performed using a FISH kit (Service Bio) according to the manufacturer’s instructions. Cells were fixed with in situ hybridization fixative (Cat#G1113, Service Bio), followed by washes in PBS. Proteinase K (Cat#G1234, Service Bio) was applied for digestion, then rinsed with pure water and washed in PBS. Pre-hybridization buffer was added and incubated. After removing the pre-hybridization buffer, probe-containing hybridization buffer (Cat#G3045, Service Bio) was added for overnight hybridization. Post-hybridization washes included: 2×→1×→0.5× Saline Sodium Citrate (SSC, Cat#G3015, Service Bio) for 10 min. For signal amplification, pre-warmed branch probe hybridization solution was added and hybridized, followed by sequential washes with 2×→1×→0.5×→0.1× SSC. Signal probe hybridization solution was then applied and incubated, followed by additional SSC washes. Finally, nuclei were counterstained with DAPI (Cat#G1012, Service Bio) and imaged by microscopy (Leica DM IL LED).

MiRNA pulldown DNA assay

MiRNA pulldown DNA assay was performed using the miRNA pulldown kit (Bersin Bio, Cat#Bes5204) according to the manufacturer’s instructions. Briefly, HNSCC cells transfected with biotin-labeled miR-3690 or biotin-labeled NC were crosslinked with 1% formaldehyde and quenched with glycine. Ultrasonication was used to obtain DNA fragments of 200–500 bp in size. The lysates were immunoprecipitated using streptavidin-conjugated magnetic beads. Immunoprecipitated DNA was analyzed by sequencing and qRT-PCR.

In-situ hybridization (ISH) for miR-3690 detection

Following deparaffinization and rehydration, tissue sections were digested with proteinase K (Cat# G1234, Service bio) and fixed. After prehybridization, digoxigenin (DIG)-labeled miR-3690-specific probes (5’- CTTTGTCTACGCTGGGTCCAGGT − 3’) (Service bio) were hybridized to the sections at 60 °C overnight. Post-hybridization washes were conducted, followed by serum blocking. The sections were then incubated with HRP-conjugated mouse anti- DIG antibody (Cat#200-032-156, Jackson ImmunoResearch Laboratories) at 40 °C for 2 h. After PBS washes, signals were visualized using DAB chromogenic substrate (Cat# G1212, Service bio). Finally, nuclei were counterstained with hematoxylin (Cat#G1004, Service bio), and the sections were dehydrated through a graded ethanol series and mounted with neutral balsam.

TOP/FOP luciferase activity analysis

Cells were seeded in 24-well plates at 50% confluence and transfected with either TOPflash reporter plasmid (containing wild-type TCF/LEF binding sites) (Cat#D2501, Beyotime) or FOPflash control plasmid (containing mutated sites) (Cat#D2503, Beyotime). After 24 h of transfection, cells were lysed and centrifuged at 10,000×g for 10 min at 4 °C to obtain supernatants. The supernatants were sequentially mixed with Firefly and Renilla luciferase assay reagents (Cat#RG027, Beyotime) to measure relative light units. The Wnt/β-catenin signaling activity was quantitatively analyzed by calculating the Firefly/Renilla luciferase activity ratio.

Nucleoplasmic RNA separation

Nucleoplasmic RNA separation assay was performed using PARIS™ Kit (Cat# AM1921) from Thermo Fisher Scientific according to the manufacturer’s instructions. Collect up to 10⁷ fresh cultured cells, wash once with PBS, and keep on ice. Resuspend cells in 200 µL ice-cold Cell Fractionation Buffer, incubate on ice for 10 min, then centrifuge at 500 × g for 5 min at 4 °C. Carefully aspirate the cytoplasmic fraction away from the nuclear pellet, wash the nuclear pellet with ice-cold Cell Fractionation Buffer, and lyse in Cell Disruption Buffer. Mix the lysate with an equal volume of 2x Lysis/Binding Solution, reduce viscosity if needed, then add 100% ethanol equal to the sample volume. Pass the mixture through a Filter Cartridge, wash once with 700 µL Wash Solution 1 and twice with 500 µL Wash Solution 2/3. Elute RNA with 60 µL of 95 °C Elution Solution, followed by a second 60 µL aliquot. The purified RNA is ready for subsequent qRT-PCR analysis.

Nuclear protein extraction

Nuclear protein extraction assay was performed using Nuclear and Cytoplasmic Protein Extraction Kit (Cat# P0027) from Beyotime according to the manufacturer’s instructions. Wash and pellet cells with ice-cold PBS, then add cytoplasmic extraction reagent containing PMSF, vortex vigorously for 10 s, and lyse on ice for 10–15 min. Centrifuge at 12,000×g for 5 min to collect the cytoplasmic protein supernatant. Resuspend the pellet with nuclear extraction reagent containing PMSF, vortex vigorously for 15 s, and lyse on ice for 30 min. Centrifuge at 12,000×g for 10 min, and the resulting supernatant contains the nuclear protein extract.

Animal experiments

All animal procedures were approved by the Eye & ENT Hospital, Fudan University. Six-week-old BALB/c nude mice, were provided by the Shanghai Laboratory Animal Company (Shanghai, China). Mice were handled in accordance with the guidelines provided by the National Institutes of Health for the care and use of laboratory animals. A subcutaneous xenograft tumor model was established by injecting transfected cells (5 × 10⁶) into the flanks of mice (n = 5 per group). The mice were sacrificed approximately 30 days later and their tumor size and weight were measured. The tumor tissues were then subjected to H&E, immunofluorescence, or immunohistochemical staining.

Statistical analysis

Stata 13.0 (StataCorp, USA) and GraphPad Prism 9 software (GraphPad Software, USA) were used for statistical analyses. Data are presented as means ± standard deviation (SD). Student’s unpaired t-test was used to compare two groups, and one-way analysis of variance was used to compare three groups. Statistical significance was set at P < 0.05. significant.

Results

MiR-3690 expression increases during HNSCC and is associated with poor prognosis in patients with HNSCC

To investigate miRNA expression in HNSCC tissues, we analyzed three pairs of HNSCC tissues and matched adjacent non-tumor tissues using RNA-seq and found that miR-3690 expression was significantly higher in HNSCC tissues than in matched adjacent non-tumor tissues (Fig. 1A and B). This finding was also consistent with the qRT-PCR results of miR-3690 expression in HNSCC tissues relative to its expression in matched non-tumorous tissues (Fig. 1C and D). Histochemical analyses of the tissue microarray revealed that miR-3690 was expressed in the cytoplasm and nucleus; miR-3690 was expressed significantly higher in HNSCC samples than in matched non-tumor tissues (Fig. 1E and F). Patients with high levels of miR-3690 expression had shorter survival times and higher recurrence rates than those with low levels of miR-3690 expression (Fig. 1G and I). In addition, high miR-3690 levels were associated with distant metastasis and TNM stage (Table 1). Taken together, these results demonstrate that increased miR-3690 expression in HNSCC correlates with poor clinical outcomes in patients and is likely involved in the progression of HNSCC.

Fig. 1.

MiR-3690 expression is upregulated in HNSCC and correlates with poor prognosis. (A) RNA-seq analysis was performed on three pairs of HNSCC and adjacent tissues. (B) Volcano plot of differentially expressed genes. Red stands for upregulated genes, green represents downregulated genes. (C–D) qRT-PCR was used to quantify miR-3690 expression in 30 matched pairs of HNSCCs. (E) MiR-3690 expression was measured by in-situ hybridisation in TMAs from 79 patients with HNSCC and representative images are provided, the scale bar: 200 μm, 50 μm. MiR-3690 expression level was analyzed by H-score (F). (G-I) Survival and recurrence analysis of miR-3690 expression in 79 patients with HNSCC, the scale bar: 200 μm, 50 μm. *P < 0.05. **P < 0.01

Table 1.

Correlation between miR-3690 and clinicopathological characteristics in 79 HNSC patients

Variables	Number of patients
	41 miR-3690high	38 miR-3690low	P value
Gender
Male	41	36	0.228
Female	0	2
Year
≤ 65	22	23	0.538
>65	19	15
Smoke
Yes	37	30	0.162
No	4	8
Drinking
Yes	24	20	0.598
No	17	18
Hypertension
Yes	15	11	0.470
No	26	27
Tumor size (cm)
≤ 3.5	14	22	0.034
> 3.5	27	16
Lymph nodes metastasis
Yes	23	12	0.028
No	18	26
Tumor differentiation
Moderate to poor	29	30	0.401
High	12	8
Clinical stages
I - III	14	24	0.010
IV	27	14

Increased miR-3690 levels promote HNSCC progression in vivo and in vitro

Next, to understand the biological functions of miR-3690 in HNSCC development, we determined miR-3690 expression in three HNSCC cell lines and normal epithelial cells. Thereafter, miR-3690 shRNAs were transfected into TU686 cells for high expression and transfected miR-3690 vectors into AMC-HN-8 cells for low expression, which was confirmed by qRT-PCR (Fig. 2A and C). Transwell migration and invasion assays and EdU assay revealed significant inhibition of the metastasis and growth of TU686 cells after miR-3690 knockdown, whereas overexpression of miR-3690 promoted the metastasis and growth of AMC-HN-8 cells (Fig. 2D and F). MiR-3690 may participate in EMT in addition to promoting HNSCC proliferation. TU686 cells adopted the cobblestone-like morphology of normal epithelial cells after miR-3690 knockdown, whereas AMC-HN-8 cells developed the spindle-like morphology of fibroblasts after overexpressing miR-3690 (Fig. 2G). Next, we detected changes in the expression levels of EMT markers in transfected HNSCC cells. Western blot analysis showed that miR-3690 knockdown increased the expression of the epithelial marker E-cadherin in TU686 cells and decreased the levels of the mesenchymal markers Vimentin and Snail; on the contrary, miR-3690 overexpression inhibited the expression of E-cadherin and upregulated Vimentin and Snail in AMC-HN-8 cells (Fig. 2H). Furthermore, these findings were confirmed by immunofluorescence analysis of transfected HNSCC cells (Fig. 2I). To further explore the effects of miR-3690 in vivo, a subcutaneous xenograft model was established, which indicated that miR-3690 knockdown suppressed tumor growth substantially, as proven by reduced tumor size and weights after 4 weeks in TU686 xenografts (Fig. 2J and L). Immunofluorescence staining for Ki67, Ecadherin and Vimentin showed that miR-3690 knockdown resulted in a significant decrease in Ki67 and Vimentin protein levels, but an increase in Ecadherin protein levels, confirming the promoting effects of miR-3690 in HNSCC progression (Fig. 2M). Taken together, these data suggested that miR-3690 upregulation promotes HNSCC progression both in vivo and in vitro.

Fig. 2.

MiR-3690 promoted HNSCC cell migration, invasion and proliferation in vitro and in vivo. (A) The level of miR-3690 in three HNSCC cells and a normal nasopharyngeal epithelial cell was determined by qRT-PCR. (B–C) The efficiency of transfection in HNSCC cells was assessed by qRT-PCR. (D-F) Transwell and EdU assays were used to investigate migration, invasion, and proliferation abilities of transfected HNSCC cells, the scale bar: 50 μm. (G) In the transfected cells, cellular morphology was detected, the scale bar: 50 μm. (H) EMT markers were determined by western blotting. (I) Immunofluorescence staining of EMT markers in transfected cells, the scale bar: 50 μm. (J) Tumors derived from nude mice that were subcutaneously transplanted with transfected TU686 cells. (K-L) Volume and weight measurement of subcutaneous tumors. (M) Immunofluorescence staining was used to measure Ki67, Ecadherin and Vimentin expression in serial sections of xenograft tumors, the scale bar: 50 μm. *P < 0.05. **P < 0.01

miR-3690 effectively activates CKS2 expression by targeting its promoter

Some miRNAs have been reported to be located not only in the cytoplasm, but also in the nucleus [12]. Nucleocytoplasmic separation and miRNA in situ hybridization assays revealed that miR-3690 is distributed in the nucleus and cytoplasm (Fig. 3A). To investigate the target genes regulated by miR-3690 and further decipher the mechanism underlying the biological effects of miR-3690 on HNSCC progression, miRNA pulldown DNA-seq was performed on AMC-HN-8 cells biotin-labeled miR-3690. MACS2 was introduced to perform peak calling filtered based on the P-value and peak enrichment, and the peak regions ranging from the TSS 3000 bp upstream to 3000 bp downstream were annotated using CHIPSeeker (Fig. 3B). As shown in Fig. 3C, the peak regions in the whole genome were enriched in intergenic regions (85.39%), promoter regions (7.41%), and intron regions (5.59%). GO enrichment revealed that their molecular functions were mainly focused on the processes of the structural constituents of ribosomes, nuclease activity, and molecular adaptor activity, with specific biological processes mainly focused on the positive regulation of the cell cycle (Fig. 3D). Cyclin-dependent kinase regulatory subunit 2 (CKS2) plays an essential role in regulating multiple cellular processes in diverse human malignancies [13–15]. Here, we found a putative miR-3690-binding peak at the promoter region of CKS2, based on which the potential motif bound by miR-3690 was identified using the MEME motif software (Fig. 3E and F). Furthermore, the miRNA pulldown DNA assay performed in AMC-HN-8 cells transfected with biotin-labeled miR-3690 revealed that biotin-labeled miR-3690 pulled down the CKS2 promoter DNA more effectively than did the negative control (Fig. 3G). The dual-luciferase reporter assay demonstrated that with an increase in the concentration of miR-3690, the transcriptional activity of CKS2 was enhanced more significantly (Fig. 3H). qRT-PCR and Western blotting showed that miR-3690 knockdown reduced the expression of CKS2 in TU686 cells; conversely, miR-3690 overexpression had the opposite effect in AMC-HN-8 cells (Fig. 3I and J). The correlation between miR-3690 and CKS2 mRNA expression was further analyzed via a scatter plot, the results suggested that there was significant positive correlation between miR-3690 and CKS2 mRNA expression (Fig. 3K). Moreover, we found that CKS2 protein expression was significantly higher in HNSCC tissues than that in matched adjacent non-tumor tissues, and Pearson correlation analysis showed that miR-3690 was positively related to CKS2 protein expression in these patients with HNSCC (Fig. 3L and M). These findings indicated that CKS2 may be a direct target of miR-3690 in HNSCC.

Fig. 3.

MiR-3690 activates CKS2 expression by targeting its promoter. (A) Intracellular localization of miR-3690 was detected by nucleocytoplasmic separation assay and in situ hybridization assay, the scale bar: 10 μm. (B) Heatmap showing miRNA pulldown DNA-seq read densities around the miR-3690-bound regions in AMC-HN-8 cells; the representative image is presented. (C) The distribution of miR-3690 peaks across the genome is shown. (D) GO enrichment analysis of biological processes, cellular components, and molecular function categories. (E) The CKS2 promoter region contains a motif bound by miR-3690. (F) The distribution of miR-3690 binding peaks at the CKS2 promoter. (G) The miRNA pulldown DNA assay was performed to verify miR-3690 enrichment at the target regions of CKS2 promoter in AMC-HN-8 cells. (H) Dual luciferase reporter assay was applied to assess the transcriptional activity of CKS2 in AMC-HN-8 cells overexpressing miR-3690. (I-J) CKS2 expression was detected in the indicated cells using qRT-PCR and western blotting analyses. (K) Pearson correlation analysis between CKS2 mRNA and miR-3690 was performed in 25 HNSCC tissues. (L) CKS2 expression was measured by IHC in TMAs and CKS2 expression level was analyzed by H-score, the scale bar: 200 μm, 50 μm. (M) MiR-3690 and CKS2 expression were determined by in-situ hybridisation in TMAs and their relationship was assessed by Pearson correlation analysis, the scale bar: 200 μm, 50 μm. **P < 0.01

Upregulation of CKS2 by miR-3690 correlates with increased BPTF occupancy at CKS2 promoter

To further explore the molecular mechanism by which miR-3690 activates CKS2 expression, RNA pull-down and mass spectrometry (MS) were performed to identify potential miR-3690-binding proteins in AMC-HN-8 cells biotin-labeled miR-3690. Our findings identified the Bromodomain PHD Finger Transcription Factor (BPTF), the largest subunit of the nuclear remodeling factor (NURF) complex [16], among the candidate miR-3690-interacting partners (Fig. 4A and B). BPTF was selected for further investigation because BPTF is widely recognized to catalyze nucleosome sliding in an ATP-dependent manner to assist transcriptional activation of target genes [17–19]; however, its role in the transcriptional activation of CKS2 mediated by miR-3690 has not been well studied. The interaction between miR-3690 and BPTF was determined using RNA pull-down and co-localization assays (Fig. 4C and D). Furthermore, the ChIP-PCR assay demonstrated significant enrichment of BPTF in the CKS2 promoter region (Fig. 4E). Both qPCR and Western blot analysis confirmed that BPTF siRNA effectively reduced BPTF expression (Fig. 4F). Notably, treatment with different concentrations of BPTF siRNA (50 nM and 100 nM) significantly attenuated the miR-3690-mediated enhancement of CKS2 transcriptional activity (Fig. 4G). Whereas miR-3690 overexpression increased BPTF enrichment at the CKS2 promoter (Fig. 4H). Histone H3 lysine 4 trimethylation (H3K4me3), a histone modification, has been reported to be associated with the transcriptionally active state of genes and is a primary ligand for the plant homeodomain (PHD) finger of BPTF [16, 20]. In this study, we found that miR-3690 overexpression increased the enrichment of H3K4me3 in the promoter region of CKS2 (Fig. 4H), which further suggested that the occupancy of BPTF at the CKS2 promoter increased after miR-3690 overexpression. Additionally, qRT-PCR and western blotting showed that knockdown of BPTF attenuated the upregulation of CKS2 induced by miR-3690 overexpression (Fig. 4I and J). As a basic component in the regulation of cell cycle progression, CKS2 has been reported to participate in the control of the cell cycle and to facilitate cancer cell proliferation in multiple tumor types [13–15]. Here, we performed an EdU assay to determine the function of CKS2 in HNSCC, as shown in Fig. 4K; miR-3690 overexpression promoted the growth of AMC-HN-8 cells, and knockdown of BPTF or CKS2 abrogated cell proliferation induced by overexpression of miR-3690. Finally, immunohistochemistry analysis results implied that the level of BPTF protein in HNSCC samples was significantly higher than that in matched non-tumor tissues, and that the level of BPTF protein was positively correlated with the level of miR-3690 in the HNSCC tissue microarray (Fig. 4L and N). Taken together, these results demonstrate that miR-3690 upregulates CKS2 expression by increasing BPTF enrichment at the CKS2 promoter.

Fig. 4.

BPTF associates with the promoter of CKS2. (A-B) A combination of RNA pulldown and MS was used to identify the miR-3690 binding partners. (C) RNA pulldown and western blotting were conducted to verify the interaction between miR-3690 and BPTF. (D) Immunofluorescence and in situ hybridization assay were applied to identify the colocalization of miR-3690 and BPTF, the scale bar: 10 μm. (E) CHIP–qPCR was utilized to quantify enrichment of BPTF at the CKS2 promoter. (F) The knockdown efficiency of BPTF siRNA was verified by qRT-PCR and Western blot analysis. (G) The dual luciferase reporter assay was performed to evaluate the transcriptional activity of CKS2 in AMC-HN-8 cells transfected with miR-3690 and different concentrations (e.g., 50 nM, 100 nM) of BPTF siRNA. (H) BPTF and H3K4me3 enrichment at the CKS2 promoter was measured using CHIP-qPCR. (I-J) CKS2 expression was analyzed by qRT-PCR and western blotting in transfected AMC-HN-8 cells. (K) The proliferation ability of transfected AMC-HN-8 cells was determined using EdU assays, the scale bar: 50 μm. (L) BPTF expression was measured by IHC in TMAs and BPTF expression level was analyzed by H-score, the scale bar: 200 μm, 50 μm. (M-N) MiR-3690 and BPTF expression were determined by in-situ hybridisation and IHC in TMAs and their relationship was assessed by Pearson correlation analysis, the scale bar: 200 μm, 50 μm. **P < 0.01

MiR-3690 facilitates Wnt/β-catenin signaling in HNSCC by repressing NKD1 expression through direct targeting of its 3’-UTR

The classical mechanism for miRNAs to regulate gene expression is by binding to the 3’-untranslated region (3’-UTR) of target genes for post-transcriptional control. Here, by comparing the target genes regulated by differentially expressed miRNAs in HNSCC tissues and matched adjacent non-tumor tissues, we found that their biological functions focused on the Wnt signaling pathway, metabolic pathway, endocytosis, MAPK signaling pathway, and axon guidance, according to Kyoto Encyclopedia of Genes and Genomes (KEGG) analysis (Fig. 5A). Naked cuticle 1 (NKD1), widely considered a negative regulator of the canonical Wnt/β-catenin pathway, is abnormally expressed in many types of malignant tumors [21, 22]. Upon searching the TargetScan website, we found that the NKD1 mRNA transcript had a putative target sequence for miR-3690 in the 3’-UTR and was predicted to be a direct target of miR-3690, which was further verified using luciferase reporter (Fig. 5B). qRT-PCR and western blotting analyses revealed that miR-3690 knockdown increased NKD1 expression, whereas miR-3690 overexpression yielded the opposite result (Fig. 5C and D). NKD1 negatively regulates the Wnt/β-catenin pathway, whereas miR-3690 suppressed the expression of NKD1. Therefore, we speculated that miR-3690 may also regulate the Wnt/β-catenin pathway by targeting NKD1. To test this hypothesis, nuclear protein extraction combined with western blot and immunofluorescence (IF) were performed; Fig. 5E and F show that miR-3690 knockdown repressed the levels of nuclear β-catenin, whereas the overexpression of miR-3690 yield the opposite effect. Similar changes in the expression of downstream target genes of the Wnt/β-catenin pathway such as Myc, MMP7, and Axin2 were also observed in miR-3690 knockdown or overexpression cells (Fig. 5G and H). TOP/FOP luciferase activity analysis further suggested miR-3690 knockdown suppressed the Wnt/β-catenin pathway activity, which was regained after NKD1 knockdown. In contrast, the increased Wnt/β-catenin pathway activity induced by miR-3690 overexpression was attenuated by the overexpression of NKD1(Fig. 5I). Finally, we observed that NKD1 knockdown restored cell migration and invasion, which were impaired by miR-3690 knockdown in TU686 cells. NKD1 overexpression abrogated the migration and invasion induced by miR-3690 overexpression in AMC-HN-8 cells (Fig. 5J and M). Taken together, these results reveal that miR-3690 promotes HNSCC cell migration and invasion by inhibiting NKD1 and activating the Wnt/β-catenin pathway.

Fig. 5.

MiR-3690 represses NKD1 expression by directly targeting its 3′-UTR. (A) KEGG pathway analysis of the target genes regulated by differentially expressed miRNAs in HNSCC tissues and matched adjacent nontumor tissues. (B) The putative miR-3690 binding sequence in the 3′-UTR of NKD1 (top panel). Luciferase activity of 293 T cells co-transfected with WT or MT luciferase reporter plasmids and miR-3690 or negative control (bottom panel). (C-D) NKD1 expression was analyzed by qRT-PCR and western blotting in transfected HNSCC cells. (E-F) β-catenin expression in the nucleus was detected by western blotting and immunofluorescence in transfected HNSCC cells, the scale bar: 10 μm. (G-H) Myc, MMP7, and Axin-2 expression were analyzed by qRT-PCR and western blotting in transfected HNSCC cells. (I) TOP/FOP-Flash luciferase reporter assay was used to analyze the activity of the Wnt/β-catenin pathway. (J-M) The transwell assay was used to assess the migration and invasion ability of transfected HNSCC cells, the scale bar: 50 μm. **P < 0.01

METTL3 and METTL14-dependent m6A methylation modulates the processing of pri-miR-3690 by DGCR8

Several recent advances in the epigenetic regulation of tumors have clarified the role of m6A methylation of RNA in the initiation and progression of tumors [23–25]. It remains unclear whether and how miR-3690 is modulated by m6A methylation in HNSCC cells. 5-AZA-CdR is a DNA methyltransferase inhibitor. We found that miR-3690 expression was not upregulated in HNSCC cells following 5-AZA-CdR treatment (Fig. 6A). The SRAMP database was used to predict potential m6A sites located in pri-miR-3690, which revealed that m6A was significantly enriched within these predicted m6A sites (Fig. 6B). Methyltransferases METTL3 and METTL14, which are essential for m6A modification, have been demonstrated to play a crucial role in catalyzing m6A formation. RNA pull-down and RIP assays indicated that pri-miR-3690 interacted with METTL3 and METTL14, as DGCR8 is involved in recognizing and processing pri-miRNAs as well as pri-miR-3690 (Fig. 6C and D). In addition, we found that HNSCC cells expressed more METTL3 and METTL14, and that METTL3 or METTL14 knockdown led to a reduction in the m6A levels of pri-miR-3690 (Fig. 6E and G). Furthermore, depletion of METTL3 or METTL14 reduced the binding of DGCR8 to pri-miR-3690, resulting in a global reduction of mature miR-3690 and concurrent accumulation of unprocessed pri-miR-3690 (Fig. 6H and I). Finally, immunohistochemistry analysis results suggested that the miR-3690 expression level was positively correlated with that of METTL3 and METTL14 in HNSCC tissue microarrays (Fig. 6J). These data implied that METTL3- and METTL14-mediated m6A modifications accelerate pri-miR-3690 maturation by regulating the processing of pri-miR-3690 by DGCR8.

Fig. 6.

m6A methylation accelerates the processing of pri-miR-3690 by DGCR8. (A) Pri-miR-3690 and miR-3690 expression in HNSCC cells treated with 5-AZA-CdR (10 μm) for 24 h were determined using RT-qPCR. (B) m6A modification in pri-miR-3690 was detected by MeRIP-qPCR analysis. (C-D) RNA pulldown and RIP assays indicated that pri-miR-3690 interacted with DGCR8, METTL3, and METTL14. (E-F) METTL3 and METTL14 expression were examined by western blotting. (G) MeRIP-qPCR analysis was used to detect m6A modification in pri-miR-3690 after METTL3 or METTL14 knockdown. (H) The interaction between pri-miR-3690 and DGCR8 was detected by RIP after METTL3 or METTL14 knockdown. (I) Pri-miR-3690 and miR-3690 expression in transfected HNSCC cells were determined using RT-qPCR. (J) METTL3, METTL14, and miR-3690 expression were determined by IHC and in-situ hybridisation in TMAs, and their relationship was assessed by Pearson correlation analysis, the scale bar: 200 μm, 50 μm. (K) Graph illustrating that m6A modification accelerated pri-miR-3690 maturation; miR-3690 upregulation promoted HNSCC cell proliferation, migration, and invasion through direct activation of CKS2 expression and miR-3690/NKD1/Wnt/β-catenin axis. *P < 0.05. **P < 0.01

Discussion

In the present study, miR-3690 was found to be significantly upregulated in HNSCC tissues. High levels of miR-3690 were associated with malignant phenotypes of HNSCC, and miR-3690 may serve as a potential prognostic indicator in patients with HNSCC. The potential mechanism by which miR-3690 promotes HNSCC progression was further elucidated. MiR-3690 could upregulate CKS2 expression by targeting its promoter and increasing BPTF occupancy at the CKS2 promoter. Meanwhile, miR-3690 facilitated Wnt/β-catenin signaling in HNSCC by repressing NKD1 expression through direct targeting of its 3’-UTR. Additionally, METTL3- and METTL14-mediated m6A modifications accelerate pri-miR-3690 maturation by regulating the processing of pri-miR-3690 by DGCR8.

MiRNAs, which are small non-coding RNAs that function as post transcriptional regulators of the target gene, play an important role in the occurrence and progression of a variety of malignancies, including HNSCC. As reported, miR-4295 promoted cell proliferation and metastasis in HNSCC by targeting NPTX1; miR-96-5p served as a novel biomarker for predicting response to radiotherapy and development of local recurrence in patients with HNSCC; cancer-associated fibroblasts facilitated oral cancer progression via exosome-mediated paracrine miR-34a-5p [26–28]. In this study, we found that miR-3690 was overexpressed in HNSCC and that miR-3690 upregulation endowed HNSCC cells with enhanced proliferative, migratory, and invasive abilities in vitro and in vivo. For a major part, miRNAs negatively regulate the expression of target genes primarily by interaction with 3ʹ-UTR of target mRNAs to induce mRNA degradation or translational suppression [29]. However, in addition to this classical inhibitory mechanism, recent studies have suggested that some miRNAs can positively regulate gene expression by targeting their promoter elements, a phenomenon called RNA activation (RNAa) [30]. For example, miRNA-551b-3p can directly bind to a complementary sequence within the STAT3 promoter to activate STAT3 transcription and thus upregulate STAT3 expression [10]; similarly, miR-744 may serve as an oncogenic mediator by activating CCNB1 expression by binding to the CCNB1 promoter to manipulate tumor growth in vivo [30]. In this study, miRNA pulldown DNA-seq was performed to identify the target genes regulated by miR-3690. The results revealed that miR-3690 upregulates CKS2 by enhancing CKS2 transcription by binding to a putative binding site in the CKS2 promoter region. miRNA pulldown DNA-seq results indicated that miR-3690 could also bind to the intergenic or intron regions of target genes, in addition to binding to the promoter regions, and we will continue to investigate the effect of miR-3690 on gene expression by targeting introns, intergenic regions, or other sites of the target genome in the future.

MiRNAs directly interact with proteins to regulate cellular functions. For instance, miRNA-328 binds to hnRNP E2, releasing C/EBPα mRNA from hnRNP E2 and thereby restricting C/EBPα mRNA translation [31]. In Parkinson’s disease, MITF3 and HNF4a were identified as protein-binding partners of miR-4745-5p [32]. However, little is known about the interactions between dysregulated miR-3690 and its protein-binding partners in HNSCC. In the present study, RNA pull-down assays and MS were performed to identify potential miR-3690-binding proteins in HNSCC cells biotin-labeled miR-3690. The findings showed that BPTF was identified as a candidate miR-3690-interacting partner. As the largest subunit of the chromatin-remodeling complex, BPTF is composed of a DNA-binding domain, two zinc-finger structures, and a bromodomain [33]. It regulates DNA accessibility by acting on the chromatin remodeling process and could also serve as a transcription factor to promote the development of cancer by regulating downstream gene expression. For example, BPTF was shown to promote stemness and metastasis of hepatocarcinoma by activating hTERT transcription [18] and to accelerate lung cancer progression by co-regulating COX-2 expression with p50 [34]. In our study, we found that BPTF was enriched in the promoter region of CKS2; while knockdown of BPTF significantly reduced the transcriptional activity of CKS2 enhanced by miR-3690, miR-3690 overexpression increased the occupancy of BPTF and H3K4me3 in the promoter region of CKS2. These results implied that miR-3690 upregulates CKS2 expression by increasing BPTF enrichment in the CKS2 promoter region.

A large number of studies have demonstrated that miRNAs exert their influence by degrading or suppressing translation of target mRNAs through interactions with the 3’-UTR [29]. The mTORC1-regulated STAT3/miR-130b-3p/MBNL1 axis promotes angiogenesis and tumor growth [5], and alcohol-induced dysregulation of miR-30a and miR-934 plays an important role in the pathogenesis and progression of HNSCC [6]. In line with previous reports, we found that NKD1 mRNA transcript had a putative target sequence for miR-3690 in the 3’-UTR and was further verified to be a direct target of miR-3690. NKD1 has been reported to negatively regulate the Wnt/β-catenin pathway [22]. Here, we found that miR-3690 suppresses NKD1 expression, and miR-3690 may activate the Wnt/β-catenin pathway by targeting NKD1.

Recently, increasing evidence has shown that m6A modification, a type of RNA epigenetic modification, plays a pivotal role in promoting cancer progression by influencing RNA fate and function, including mRNA or non-coding RNA stability, mRNA translation, interactions between RNAs and proteins, and miRNA biogenesis. For example, METTL3 has been shown to counteract premature aging through the stabilization of MIS12 mRNA by m6A; m6A-modified LINC00958 increased lipogenesis and was an effective nanotherapeutic target for hepatocellular carcinomas; and METTL3 promoted bladder cancer progression through regulating m6A-dependent pri-miR-221/222 processing [23–25]. In the present study, we observed that m6A was significantly enriched within pri-miR-3690 and that METTL3 and METTL14 were required to catalyze the formation of m6A modifications. The m6A mark has been reported to make pri-miRNAs for recognition and processing by DGCR8, thereby accelerating miRNA maturation [35]. We found that METTL3 or METTL14 knockdown led to a reduction in the m6A levels of pri-miR-3690, as well as a reduction in DGCR8 binding to pri-miR-3690, resulting in a decrease in mature miR-3690 levels and accumulation of unprocessed pri-miR-3690. Therefore, our findings revealed that METTL3- and METTL14-mediated m6A modifications accelerate pri-miR-3690 maturation by promoting the processing of pri-miR-3690 by DGCR8.

In conclusion, miR-3690 is overexpressed in HNSCC, and m6A modification-mediated miR-3690 upregulation promoted the progression of HNSCC through direct activation of CKS2 expression and miR-3690/NKD1/Wnt/β-catenin axis. Therefore, miR-3690 may be a prognostic indicator and potential therapeutic target for HNSCC. We acknowledge several limitations in our study, particularly regarding the imaging data. Specifically, the fluorescence images were not acquired using confocal microscopy, resulting in comparatively lower resolution without three-dimensional depiction or optical slices. These technical constraints were primarily due to equipment limitations in our current experimental setup. We plan to address this in future studies by incorporating more rigorous co-localization experiments using optimized protocols and advanced imaging techniques.

Supplementary Information

Below is the link to the electronic supplementary material.

Abbreviations

HNSCC

head and neck squamous cell carcinoma

miRNAs

MicroRNAs

PMSF

phenylmethanesulfonyl fluoride

qRT-PCR

quantitative reverse transcription PCR

EdU

5-Ethynyl-2′-deoxyuridine

DAPI

diamidino-2-phenylindole

IF

immunofluorescence

DAB

3,3-diaminobenzidine

RIP

RNA immunoprecipitation assay

meRIP

methylated RNA immunoprecipitation

FISH

fluorescence in situ hybridization

MS

mass spectrometry

BPTF

bromodomain PHD Finger Transcription Factor

NURF

nuclear remodeling factor

H3K4me3

histone H3 lysine 4 trimethylation

PHD

plant homeodomain

3'-UTR

3’-untranslated region

KEGG

Kyoto Encyclopedia of Genes and Genomes

NKD1

naked cuticle 1

RNAa

RNA activation

Author contributions

L. Zhou, Y. Guo and C. Xu conceived and designed the project; Y. Zhou, Q. Huang and X. Yuan conducted experiments; Y. Zhou, Q. Huang and X. Yuan collected clinical samples and information; Y. Zhou, Q. Huang, X. Yuan, Y. Guo and C. Xu performed statistical analysis. L. Zhou, Y. Guo, C. Xu and Y. Xu provided technical and material support. All authors approved the manuscript.

Funding

This work was supported by grants from the National Natural Science Foundation of China (81972529, 82372972, and 82403387), Science and Technology Commission of Shanghai Municipality (19411961300), Shanghai Sailing Program (23YF1404700 and 21YF1405600).

Data availability

The data used in this study are available from the corresponding author on reasonable request.

Declarations

Competing interests

The authors declare that they have no competing interests.