Background
Head and neck carcinomas are the sixth most common cancers by incidence among all types of malignant tumors, while the laryngeal cancer is the second most common cancer by incidence in the head and neck region. In 2017, approximately 13,360 new LSCC patients were diagnosed in the United States [
1,
2], while in China, this number amounted to approximately 26,400 new LSCC patients in 2015. Although a variety of treatment methods have improved, the 5-year survival rate of LSCC patients has not significantly improved over recent decades [
3,
4]. Therefore, it is urgent to find effective molecular markers and reveal the mechanisms underlying LSCC tumorigenesis or development so as to explore more effective strategies for early diagnosis, prognostic evaluation, and treatment.
N
6-methyladenosine (m
6A) is the most abundant RNA modification, which was the first revealed in mRNA and has also been recently found in precursor mRNAs, circRNAs, and lncRNAs [
5‐
7]. m6A can be catalyzed by various enzymes, including methyltransferases, demethylases, and effector proteins. Furthermore, m6A modification can control RNA fate by affecting post-transcriptional regulation, such as RNA stability, splicing, and translation efficiency, which influences multiple biological processes, including cell differentiation, tissue development, as well as spermatogenesis, and RNA–protein interactions [
8,
9]. Recent researches have demonstrated that m6A RNA methylation has an essential role in various human diseases such as hypertension [
10], cardiac hypertrophy [
11], viral infection [
12], diabetes [
13] and cancers [
14]. For instance, Chen et al found that WTAP’s participation in m6A methylation has a vital role in the occurrence of hepatocellular carcinoma [
15]. Furthermore, Wang et al reported that high METTL3 expression promotes tumor angiogenesis and glycolysis in gastric cancer [
16]. Bai et al determined that YTHDF1 exerts a crucial oncogenic role in CRC by promoting Wnt/β-catenin pathway [
17]. However, the RNA m6A expression patterns and their relevant mechanisms in LSCC remain largely unknown.
In this study, the overexpression of m6A methyltransferase RBM15 in LSCC tissues was first authenticated. The RBM15 level was associated with the progression of LSCC and the prognosis of LSCC patients. We further revealed that RBM15 exerts an oncogenic role in LSCC both in vitro and in vivo. From the mechanism, RBM15 can regulate the m6A level of TMBIM6 mRNA, which depends on the m6A reader IGF2BP3. Taken together, these results indicated that the RBM15/IGF2BP3/TMBIM6 axis might be a novel and promising therapeutic target for LSCC.
Materials and methods
Tissue samples
All 164 pairs of LSCC and adjacent non-tumour tissues were collected from the Department of Otorhinolaryngology, the Second Affiliated Hospital of Harbin Medical University. The selection criteria included patients diagnosed with laryngeal cancer before surgery and those who have never received any treatment from 2013 to 2016. The expression levels of m6A were performed in 34 patients, a total of 8 matched samples of LSCC tissues, and the corresponding adjacent non-tumor tissues were tested by microarray and proteomics detection. Moreover, we additionally gathered 34 pairs of LSCC tissues to determine the expression of RBM15, TMBIM6, and IGF2BP3 by qRT-PCR. LSCC tissues from 122 patients were used for IHC assay.
The Harbin Medical University Ethics Committee has endorsed this study.
Cell culture and transfection
Human LSCC cells (AMC-HN-8 cells, TU-212 cells, and TU-177 cells) and NHBEC (normal human bronchial epithelial cell) cells were maintained in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% fetal bovine serum in a humidified incubator at 37 °C with 5% CO2. The cells were cultured in 24-well plates (5 × 104 cells/well) overnight. RBM15, TMBIM6, and IGF2BP3 knockdown and overexpression viruses and their respective control vectors were provided by Genechem (Shanghai, China). Transfection protocol followed the transfection reagent instructions.
m6A RNA methylation quantification
The m6A quantification was carried out utilizing the Abcam m6A RNA Methylation Quantification Kit (Abcam ab185912). The relevant solutions were added to each well according to the manufacturer’s instructions. The solution was mixed by gentle tilting from side to side, after which m6A RNA capture was performed. It was ensured that any residual wash buffer in the wells was thoroughly removed at each wash step. Signals were detected at the end when the color of the positive control wells changed to medium blue. A 100 μL Stop Solution was then added to each well to stop the enzyme reaction. After adding the Stop Solution, the color of the compound solution changed to yellow. The absorbance was read at 450 nm on the microplate reader.
qRT-PCR
Total RNA from LSCC tissues and cell lines was extracted using Trizol reagent (Invitrogen, Carlsbad, CA, USA) according to the manufacturer’s instructions. The process of qRT-PCR was as previously reported [
18]. Data were calculated using the 2
-ΔΔCt method. Sequences of the primers are listed in Additional file
1: Table S1.
Microarray analysis
Human m6A epitranscriptomic microarray and mRNA microarray analysis were from Arraystar Arraystar Company (Rockville, MD, USA). Briefly, the total RNAs were immunoprecipitated with anti-N6-methyadenosine (m6A) antibody. The elution from the immunoprecipitation magnetic beads was called “IP”. The recovered supernatant was called “Sup”, and Labels “IP” and “Sup” RNA were used for Cy5 and Cy3, respectively. After merging, it was hybridized to Arraystar Human m6A Epitranscriptomic Microarray (8 × 60 K, Arraystar). Finally, an Agilent scanner G2505C was used to scan the array.
Proteomics
The protein was extracted from the sample tissue. A 20 μg of protein was taken from each sample, which was mixed with 6 X loading buffer, boiled in a water bath for 5 min. Samples were then analyzed using a 12% SDS-PAGE electrophoresis (250 V, 40 min) and FASP enzymolysis. A 100 μg peptide was taken for each sample and was labeled according to the Thermo Company’s TMT labeling kit instructions. Data analysis was performed after mass spectral analysis and identification. An appropriate protein sequence database was selected, which is the foundation and critical step for the qualitative analysis of protein in mass spectrometry data.
MeRIP-qPCR
A 1–3 μg total RNA and m6A spike-in control mixture was added to 300 μL 1IP buffer (50 mM Tris-HCl, pH 7.4, 150 mM NaCl, 0.1% NP40, 40 U/μL RNase Inhibitor) containing 2 μg anti-m6A rabbit polyclonal antibody (Synaptic Systems). A 20 L Dynabeads™ M-280 Sheep Anti-Rabbit IgG suspension per sample was blocked with freshly prepared 0.5% BSA at 4 °C for 2 h, washed three times with 300 μL 1IP buffer, and re-suspended in the total RNA-antibody mixture prepared above. The RNA binding to the m6A-antibody beads was carried out with head-over-tail rotation at 4 °C for 2 h. The enriched RNA was eluted with 200 μL Elution buffer at 50 °C for 1 h, after which the co-precipitated RNA was isolated and qRT-PCR detection was performed on the RNA as the previously described. Sequences of the primers are listed in Additional file
1: Table S1.
Western blot
After the total protein was extracted, the protein concentration was measured with the BCA protein assay kit (Beyotime Biotechnology, China). Protein samples were electrophoresed, transferred to PVDF membranes, and blocked with 5% nonfat milk at RT for 1 h. Then the PVDF membrane was incubated with the primary antibody solution anti-RBM15 antibody (ab244374; Abcam, Shanghai, China, 1:1000), anti-TMBIM6 antibody (ab18852; Abcam, Shanghai, China, 1:1000) and anti-IGF2BP3 antibody (ab177477; Abcam, Shanghai, China, 1:1000) overnight at 4 °C. Monoclonal anti-rabbit IgG (RS0002; Immunoway, Plano, USA, 1:5000) or monoclonal anti-mouse IgG (3420; Abcam, Shanghai, China, 1:5000) were incubated with the PVDF membrane for a secondary step 1 h at room temperature. Finally, chemiluminescence (ECL) detection reagents (Beyotime Biotechnology, China) were prepared under dark conditions to detect blots.
Luciferase reporter assay
AMC-HN-8 and TU-212 cells in 24-well plates were transfected with a luciferase reporter and indicated expression constructs. The assays were carried out as previously reported [
18].
Immunohistochemistry
The slice number was marked and placed on the grill for 30 min. The paraffin slides were deparaffinized and rehydrated, and then the antigen was retrieved in a pressure cooker. The protein was blocked with 10% goat serum, and the slides were incubated with anti-RBM15 primary antibody (10587–1-AP; Proteintech, Wuhan, China, 1:500) at 4 °C overnight. The expression evaluation was as previously described [
19].
CCK-8 assay
LSCC cells were cultured overnight on 96-well plate (100 μL per well contains 2000 cells). The culture medium and CCK-8 solution were used as a blank control group. On the next day, 10 μL CCK-8 solution was added to each well. The absorbance at 450 nm was measured at 0, 24, 48 and 72 h.
Wound healing assay
A totoal of 5 × 105 cells (per well) were seeded into a 6-well culture plate. After 48 h of incubation, the cells grew to about 90% confluence, and the middle of every well was scraped with a 200 μl sterile pipette tip. The plate was then washed three times with PBS to remove cell debris, an cultured in fresh DMEM containing 2% FBS. The wound was imaged at two-time points after 0 h and 20 h. Image-Pro Plus 6.0 was used to assess the gap distance quantitatively.
Transwell assay
For cell invasion/migration analysis, LSCC cells were starved for 24 h. The bottom of the transwell filter with a pore size of 8 μm was coated with a 1:8 dilution of Matrigel (ignoring this step in cell migration assay). The single-cell suspension (1 × 105; 200 ul) diluted in serum-free medium was added to the upper chamber, and a medium (600 ul) containing 15% FBS was added to the lower chamber. The cells were allowed to migrate for 24 h at 37 °C. The cells were fixed in 3.7% formaldehyde for 15 min, washed with PBS, and stained with crystal violet for 10 min, after which the upper chamber non-invasive/migrating cells were wiped. The number of cells in three replicate experiments was evaluated to quantify cell invasion or migration.
In vivo experiment
Balb/c male nude mice, 5–6 weeks old, weighing 20–25 g, were obtained from Vital River Laboratories (Beijing, China). All the animals were housed in an environment with a temperature of 22 ± 1 °C, relative humidity of 50 ± 1%, and a light/dark cycle of 12/12 h. All animal studies (including the mice euthanasia procedure) were done in compliance with the regulations and guidelines of Harbin Medical University institutional animal care and conducted according to the AAALAC and the IACUC guidelines.
For tumor growth studies, whether in vivo RBM15 knockdown/overexpression experiments or in vivo rescue experiments, each group included six mice. Each mouse was injected with 100 μl of lentivirus-transfected tumor cells. The simplified volume of the spheroid (length × width 2 × 0.5) was used to determine the tumor volume. Six weeks after inoculation, xenografts were excised and evaluated for volume.
TUNEL assay
After being deparaffinized twice with dimethylbenzene, alcohol gradient hydration was performed. The reagents from the TUNEL kit (Contains 5 μL of equilibrium solution and 45 μL of TdT enzyme) were added to all slides at 37 °C for 1 h (50 μL/slide). The anti-digoxin antibody was dropped on all the slides, and then placed in a wet box, and infiltrated at 37 °C for 20 min. DAB developer was added to the slide to react for 20 min. Hematoxylin was slightly counterstained, differentiated in a hydrochloric acid alcohol solution for 2 s, and rinsed with distilled water for 10 min. The tissues were then dehydrated and transparently treated. Finally, the apoptosis of the two groups was observed under the microscope.
Transmission electron microscopy
The tumor was put in a refrigerator at 4 °C, fixed with glutaraldehyde for 24 h; the pH of the solution was adjusted to 7.3 with 0.1 mmol/L dimethyl arsenate buffer. The tissue was then thoroughly washed, and 1% tetraoxide Osmium was fixed in draught cupboard for 2 h at room temperature. After progressive dehydration, the tissues were embedded with the SPI-Pon812 embedding agent. The specimen was heated on an alcohol burner to detach it from the slide and was cut to a thickness of about 70 nm under the microtome. The tissue sections were stained with double electron staining of uranium acetate and lead citrate and rinsed three times, after which they were air-dried at room temperature. The ultrastructure of the transplanted tumor cells was observed under a transmission electron microscope.
RNase Mazf
Cellular RNAs were digested at the unmethylated ACA site using bacterial single-stranded RNase MazF. Sites with m6A methylation remain uncleaved. Based on the ability of MazF to discriminate between 5′-ACA-3′ and 5′-(m6A)CA-3′, we determined the m6A methylation modification site on TMBIM6 mRNA. When performing qRT-PCR, the amount of each sample added was 1 μl, and MazF- was used as a control. MazF-correction formula: % MazF- - = (2 –CtMazF +)/ (2 –CtMazF -)100%.
The KMplot program (
http://kmplot.com/analysis/) and the GEPIA (
http://gepia.cancer-pku.cn/) database were used to plot the Kaplan-Meier survival curves of RBM15, TMBIM6, and IGF2BP3. The plot of the Pearson Correlation Coefficient (RBM15/TMBIM6 and TMBIM/IGF2BP3) and the differential expression boxplot in HNSC were based on the GEPIA datasets.
Hoechst staining assay
The apoptosis assay was performed 48 h after transfection. The clean coverslip was immersed in 70% ethanol for 5 min, washed with sterile PBS three times, and then washed with a cell culture solution again. The coverslip was placed in a six-well plate; transfected cell suspension was added and incubated overnight. 0.5 ml of fixative was added and was fixed for 10 min, after which it was removed and washed twice with PBS for 3 min each time. 0.5 ml Hoechst 33258 staining solution was added and stained for 5 min, after which the staining solution was removed, washed twice with PBS, and the liquid was drained. Cell apoptosis was detected at 350 nm by a fluorescence microscope.
RNA stability assays
To measure the stability of TMBIM6 mRNA in LSCC under the influence of knockdown or overexpression of RBM15/IGF2BP3, actinomycin D (a9415; Sigma, USA, 5 μg/mL) was applied to cells. The procedure of isolating total RNA for qPCR analysis was as described previously. Finally, the mRNA expression at the specified time was calculated and normalized using GAPDH.
Statistical analysis
Statistical analyses were performed using SPSS version 17.0 software. The graphics were mainly plotted by GraphPad Prism 7.0. The statistically significant differences were evaluated by the two-tailed Student’s t-test. The OS analysis of LSCC patients was plotted by the Kaplan-Meier method and the log-rank test. The correlations among RBM15, TMBIM6, and IGF2BP3 expression in LSCC patients were calculated by Pearson correlation analysis. A p-value of < 0.05 was considered statistically significant.
Discussion
There are many types of chemical modifications in RNA. Among them, m6A is the most abundant and reversible in mammalian mRNA and non-coding RNA [
20,
21]. The m6A modification has an essential role in various biological processes such as modulation of mRNA stability, pre-mRNA splicing, translation, and DNA damage repair [
22,
23]. A series of researches have shown that m6A modification is related to a variety of human cancers, including hepatocellular carcinoma, gastric cancer [
24,
25], glioblastoma [
26], non-small cell lung cancer [
27], and so on. Still, no study explored the function of m6A RNA methylation in LSCC. In the present study, our data showed that m6A modification was increased in LSCC compared with non-neoplastic tissues, which suggested that m6A may participate in the tumorigenesis.
We first found the differential m6A enrichment in RNAs of LSCC and adjacent non-neoplastic tissues by microarray. Combining proteomics results, we screened RBM15 as the most differentially expressed m6A methyltransferase in “writers”. RBM15 is a member of the SPEN (Split-end) family of proteins, which may bind to RNA by interacting with spliceosome components [
28]. It is primarily localized in the nucleus. A previous study has shown that RBM15 and its paralogue RBM15B bind m6A-methylation complex to regulate XIST lncRNA m6A formation in human cells [
29]. Another recent study showed that RBM15 is involved in the mechanism of mRNA methylation in the developing cortex [
30].
Interestingly, RBM15 may have different roles in different types of diseases. In this study, we further revealed that the expression of RBM15 was increased compared with the non-neoplastic tissues at both mRNA and protein levels by microarray, qRT-PCR, and immunohistochemistry in LSCC. Moreover, our results indicated that the expression level of RBM15 was related to the clinicopathological characteristics and prognosis of LSCC patients. Among patients with HNSC, those with high RBM15 expression had a poor prognosis. Our data showed that inhibition of RBM15 decreased LSCC cell proliferation, migration, invasion, and apoptosis. Overexpression of RBM15 had the opposite effect, which indicates that RBM15 has a non-negligible role in LSCC and a potential value as a biomarker in the future.
We further performed m6A RNA immunoprecipitation (RIP) microarrays to identify candidate genes with high m6A methylation modification and increased mRNA expression levels. Six candidate genes were screened and verified in LSCC cell lines by using a lentiviral vector. Results showed that 6 mRNAs, including TMBIM6, were significantly hypermethylated in LSCC. By using the qRT-PCR, RIP, MazF, luciferase reporter assay, and rescue experiment, we identified TMBIM6 as the target of RBM15 in LSCC. The human TMBIM6 gene, known as Bax Inhibitor-1, is located on 12q13.12. TMBIM6 is overexpressed and has oncogene roles in multiple cancers such as squamous cervical cancer, non-small cell lung, breast, and nasopharyngeal cancers [
31‐
34]. Yet, the effect of TMBIM6 in LSCC remains unclear. The present analysis revealed that TMBIM6, which was upregulated, exerted a cancerogenic role in LSCC. TMBIM6 downregulation markedly inhibited the migration and invasion of LSCC cells.
Moreover, the ablation of TMBIM6 increased cell apoptosis characterized by brightly stained nuclei, nuclear condensation, and fragmentation. More importantly, the overexpression of RBM15 or IGF2BP3 partially neutralized the ablation effect of TMBIM6. Furthermore, MeRIP assay identified that m6A modification was enriched in the TMBIM6 sequence. Spontaneously, RBM15 downregulation could reduce TMBIM6 expression, which suggested TMBIM6 m6A modification was an essential mechanism in the regulation of cell migration, invasion and apoptosis of LSCC.
m6A “readers” are considered as a regulator of mRNA metabolism, while mainly IGF2BPs family is associated with methylated mRNA stability [
35,
36]. Among RBPs, IGF2BP3 is particularly important in tumorigenesis and tumour progression. Moreover, previous studies have addressed the molecular mechanism of IGF2BP3 carcinogenesis [
37,
38]. Recent research showed that IGF2BP3 exerts a vital part in melanoma invasion and metastasis [
39]. Another study suggested that upregulated IGF2BP2 may cause poor prognosis in pancreatic ductal adenocarcinoma [
40]. Our data identified that the expression of IGF2BP3 was positively related to RBM15 and TMBIM6 in LSCC. We thus speculated that IGF2BP3 was involved in the process of the m6A modification of TMBIM6 by RBM15. By using MeRIP-qPCR, MazF, luciferase report assay, and rescue experiment, we revealed that IGF2BP3 regulates the m6A level of TMBIM6 mediated by RBM15. Furthermore, the downregulation of IGF2BP3 or RBM15 resulted in a significant decrease in TMBIM6 level and stability. IGF2BP3 bound to the m6A site in the 3’UTR region of TMBIM6. Moreover, IGF2BP3 controlled the stability of TMBIM6 mRNA. The Pearson Correlation Coefficient obtained from the samples data corroborated the above results, and the information obtained from the TCGA data also corroborated our results. In addition, our research also identified the expression of IGF2BP3 in laryngeal cancer. The results of the collected specimens were consistent with the results in the TCGA database. The expression of IGF2BP3 in laryngeal cancer tissues was significantly higher than that of noncancerous tissues adjacent to cancer and was associated with a poor prognosis.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.