Melittin inhibits proliferation, migration and invasion of bladder cancer cells by regulating key genes based on bioinformatics and experimental assays

Abstract The antitumour effect of melittin (MEL) has recently attracted considerable attention. Nonetheless, information regarding the functional role of MEL in bladder cancer (BC) is currently limited. Herein, we investigated the effect of MEL on critical module genes identified in BC. In total, 2015 and 4679 differentially expressed genes (DEGs) associated with BC were identified from the GSE31189 set and The Cancer Genome Atlas database, respectively. GSE‐identified DEGs were mapped and analysed using Gene Ontology and Kyoto Encyclopaedia of Genes and Genomes analyses to determine BC‐involved crucial genes and signal pathways. Coupled with protein–protein interaction network and Molecular Complex Detection analyses, Modules 2 and 4 were highlighted in the progression of BC. In in‐vitro experiments, MEL inhibited the proliferation, migration, and invasion of UM‐UC‐3 and 5637 cells. The expression of NRAS, PAK2, EGFR and PAK1 in Module 4—enriched in the MAPK signalling pathway—was significantly reduced after treatment with MEL at concentrations of 4 or 6 μg/mL. Finally, quantitative reverse transcription‐polymerase chain reaction and Western blotting analyses revealed MEL inhibited the expression of genes at the mRNA (ERK1/2, ERK5, JNK and MEK5), protein (ERK5, MEK5, JNK and ERK1/2) and phosphorylation (p‐ERK1/2, p‐JNK, and p‐38) levels. This novel evidence indicates MEL exerts effects on the ERK5‐MAK pathway—a branch of MAPK signalling pathway. Collectively, these findings provide a theoretical basis for MEL application in BC treatment.

revolutionizing molecular analysis platforms and an increased interest in BC among physicians and scientists. 2  Melittin has been proposed as a promising agent for anticancer therapy. Reportedly, MEL is prominently involved in the regulation of multifarious cancers, exemplified by hepatocellular carcinoma, 5,6 breast cancer, 7 lung cancer, 8 leukaemia, 9 ovarian carcinoma, 10 gastric carcinoma, 11 and prostate cancer. 12  In recent years, microarrays employing high-throughput platforms have emerged as a promising and effective approach for the detection of important genetic or epigenetic alterations in carcinogenesis and identification of promising biomarkers for the diagnosis and prognosis of cancers. 17,18 In fact, many gene expression profiles involved in various aspects cancers have been documented. [19][20][21] However, additional critical genes and pathways associated with BC remain to be investigated.
The objective of the present study was to evaluate the ef-

| Microarray Data
The expression profile of GSE31189 was acquired from the GEO database (https ://www.ncbi.nlm.nih.gov/geo/query/ acc.cgi) based on the following two customized criteria: sample volume >30 and BC cells. The platform used for this analysis was the GPL570 (HG-

| DEG screening
The R software (version 3.3.0; https ://www.r-proje ct.org/) was applied for data mining and analysis. In detail, the Limma package 24 F I G U R E 1 The pipeline designed for this work. DEGs, differentially expressed genes; GO, gene ontology; KEGG, Kyoto Encyclopaedia of Genes and Genomes; PPI, protein-protein interaction; MCODE, Molecular Complex Detection; BC, bladder cancer; GEO, Gene Expression Omnibus; TCGA, The Cancer Genome Atlas; MEL, melittin was available for DEG screening in the GEO data files with cut-off criteria of absolute log 2 fold change (FC) >1 and P < .05. The Deseq2 package 25 in Bioconductor was applied for TCGA data files using the aforementioned criteria. False discovery rate analysis as the multiple testing was applied in the statistical analysis of DEGS screening.

| Functional and pathway enrichment analyses
Featured by a community-based bioinformatics resource, GO (http:// www.geneo ntolo gy.org) supplies information with respect to gene product function, adopting ontologies to convey biological knowledge. 26 Similarly, KEGG (http://www.genome.jp/kegg/) is a bioinformatics resource including genomic, enzymatic pathways, chemicals and network information. It is useful in inspecting the functions and utilities of cells and organisms from both high-level and genomic perspectives. 27 GO enrichment and KEGG pathway analyses were performed to gain insight into the potential biological functions of DEGs in BC. The Database for Annotation, Visualization and Integrated Discovery (DAVID; http://david.ncifc rf.gov/) was applied for these analyses with a threshold of P < .05. Significantly enriched GO terms and KEGG pathways were screened using the P < .05 criterion and finally visualized through the R package ggplot2. 28

| Construction of the PPI network
The freely accessible, online, Search Tool for the Retrieval of Interacting Genes (STRING) database (http://string-db.org) includes established and predicted PPI. 29 All DEGs were mapped onto the web-based tool STRING to generate a PPI network, aiming to reveal the functions of proteins at the molecular level. Subsequently, the distribution characteristics of the DEGs in the PPI network were visualized using the Cytoscape software (http://www.cytos cape. org/). 30 The nodes and edges in the PPI network independently represented proteins and their interactions.

| MCODE analysis
The MCODE (https ://omict ools.com/molec ular-compl ex-detec tion-tool) utilizes a theoretical clustering algorithm for the detection of densely connected regions in large PPI networks, potentially representing molecular complexes. 31 In this study, the MCODE was administered to mine the core protein complex in the constructed PPI network. Thereupon, significant module genes were annotated using the UniPro database (https ://www. unipr ot.org/). 32 F I G U R E 2 Volcano plot of detectable DEG profiles in BC. (A) Presentation of the volcano plot of DEGs identified from the GEO database; (B) Presentation of the volcano plot of DEGs identified from the TCGA. Red plots stand for up-regulated genes, whereas blue plots indicate down-regulated genes with the following criteria: P < .05 and absolute log 2 FC > 1. Grey plots indicate non-significantly expressed genes. The abscissa presents the value of fold change in gene expression between BC group and control. The ordinate shows the − log10 of the adjusted P value for each gene, symbolizing the strength of the association. DEGs, differentially expressed genes; BC, bladder cancer; GEO, Gene Expression Omnibus; TCGA, The Cancer Genome Atlas; FC, fold change

| Cell culture and treatment
Human BC cell lines (UM-UC-3 and 5637) were obtained from the American Type Culture Collection. Cells were cultured in Dulbecco's modified Eagle's medium (Gibco BRL), supplemented with 10% foetal bovine serum (Gibco BRL) and penicillin/streptomycin at a concentration of 100 U/mL. The cultured cells were maintained at 37°C in a humidified atmosphere with 5% (v/v) CO 2 .

| Quantitative reverse transcription-polymerase chain reaction (qRT-PCR)
Total RNA was extracted from the UM-UC-3 and 5637 cell lines using the TRIzol reagent (Life Technologies) according to the instructions provided by the manufacturer. Total RNA (1 μg) was used as template to synthesize complementary DNA (cDNA) using a PrimeScript RT Reagent Kit with cDNA Eraser (Takara Biotechnology). Subsequently, qRT-PCR was performed using the SYBR Premix Ex Taq (Takara Bio Inc). The primers utilized in this analysis are described in Table S1. All qRT-PCR assays were performed on an ABI 7900 system (Applied Biosystems). β-actin was used as an internal control to normalize the expression levels of genes. 2 −ΔΔC t method was applied to calculate the relative levels of gene expression.

| Cell proliferation assay
The Cell Counting Kit-8 (CCK-8) assay (Dojindo Molecular Technologies, Inc) was used according to the protocol provided by the manufacturer to assess cell proliferation after treatment with MEL at various concentrations (ie 0, 2, 4 and 6 μg/mL). UM-UC-3 and 5637 cells were seeded in 96-well plates and cultured at 37°C for 18 hours prior to treatment with MEL. After treatment (24 hours), the culture medium was replaced with Dulbecco's modified Eagle's medium containing 10% CCK-8 solution. Each experiment was performed in triplicate. At 1-4 days, the optical density (OD) was measured at a wavelength of 450 nm using a Multiskan FC (ThermoFisher Scientific, Inc).

| Colony formation assay
Cells were seeded in 60-mm plates (0.5 × 10 3 cells/plate), cultured for 7 days, fixed with 10% formaldehyde for 5 minutes and stained with 1% crystal violet for 30 s prior to counting the number of colonies.

| Cell migration assay
In the present study, both the scratch wound-healing and transwell assays were employed to detect cell migration. The scratch woundhealing assay was performed as previously reported. 16 The initial gap length at 0 hours and residual gap length at 24 hours after wounding were calculated from photomicrographs. The transwell assay was performed in 24-well Boyden chambers (Corning Inc) pre-coated with the absence of Matrigel (BD Biosciences), as previously described. 33 Cells were counted from at least four randomly selected microscopic fields.

| Cell invasion assay
The in-vitro invasion assay was performed in 24-well Boyden chambers (Corning Inc) pre-coated with the presence of Matrigel (BD Biosciences), as previously described. 33 Cells were counted from at least four randomly selected microscopic fields.

| Statistical analysis
Data are expressed as mean ± standard deviation. Each experiment was performed in triplicate. The t test was applied to assess differences between two groups. One-way and two-way analysis of variance were employed to analyse multiple groups (>2) and two-factor groups, respectively. A P < .05 denoted statistical significance. All statistical analyses were performed using the SPSS software (SPSS Inc).

| Identification of DEGs in BC independently derived from the GEO and TCGA databases
Data related to BC were extracted from the GEO and TCGA databases for the identification of DEGs. Initially, one expression profile

| GO Term and KEGG pathway enrichment analysis for GEO-derived DEGs
Gene Ontology categories are classified into three groups, namely Subsequently, KEGG pathway analysis was performed to further evaluate the biological significance of the DEGs using P < .05 as a criterion. As presented in Figure 3B and

| Construction of the PPI network and MCODE analyses
The GEO-identified DEGs were mapped onto the STRING database to assess their interactive relationship, visualized through the Cytoscape software and analysed using the MCODE algorithm (a plug-in of Cytoscape). Subsequently, the PPI network was constructed using the criterion of interacting pairs with a combined score >0.9 and visualized through the Cytoscape software ( Figure   S1). The five modules listed in Table 2 were generated using the custom parameters of 0.2 Node Score Cut-off and 3K-Core, analysed through the MCODE. The visualized results are presented in Figure 4.

| Verification of the pivotal module genes identified from the GSE31189 set in BC tissues from the TCGA database
We generated the overlap of DEGs independently identified from the GEO and TCGA databases to further identify the expression of crucial module genes in BC tissues. These two databases had 246 DEGs in common ( Figure S2). In addition, the majority of crucial genes in Modules 2 and 4-apart from ATP6V1E2, ATP6V0C, ATP6V0A2 and PAK1-exhibited similar differential expression in BC tissues obtained from the TCGA database, as listed in Table 4.
Hence, most pivotal module genes associated with the progression of BC were also verified in BC tissues from TCGA database.

| MEL inhibits cell proliferation, migration and invasion in UM-UC-3 and 5637 cells
Prior to functional experiments, the half maximal inhibitory concentration (IC 50 ) of MEL in BC cells was evaluated. We analysed the (2, 4, or 6 μg/mL) compared with the control (Figure 6E,F, P < .05, P < .01). Collectively, these results demonstrated that MEL inhibited cell proliferation, migration and invasion in UM-UC-3 and 5637 cells.

| MEL constrains the expression of pivotal module genes in UM-UC-3 and 5637 cells
We investigated the effect of MEL on the expression of pivotal module genes (Modules 4 and 2) identified in UM-UC-3 and 5637 cells.
There was no significant difference found in the expression of most genes in Module 4 between the MEL-treated group (2 μg/mL) and the control (0 μg/mL; Figure 7A,B). In the aforementioned results, we revealed that NRAS, PAK2, EGFR and PAK1 were associated with the MAPK signalling pathway (Tables 1 and 2). These four genes were significantly affected by treatment with MEL at a concentration of 4 or 6 μg/mL in both cell lines ( Figure 7A,B, P < .05, P < .01).
Similarly, the genes in Module 2 associated with V-ATPase-apart from ATP6V1F-were down-regulated in the MEL-treated group (4 or 6 μg/mL) compared with the control (0 μg/mL; Figure 8A,B, P < .05, P < .01). Collectively, these results indicated that MEL may act through inhibition of the MAPK pathway or V-ATPase in BC cells.

| MEL exerts an effect on the MAPK pathway in UM-UC-3 and 5637 cells
It has been reported that MEL suppresses the development and

| D ISCUSS I ON
Reportedly, MEL possesses multifarious biological properties (ie antifungal, antiviral, antibacterial, antiparasitic and antitumour). 37 During the previous two decades, MEL has attracted considerable attention owing to its potential use in cancer therapy. 38

| CON CLUS IONS
In summary, this research explored the effect of MEL had on the criti-

ACK N OWLED G EM ENT
This research was funded by the National Natural Science Foundation of China (nos. 81673738 and 81703875).

CO N FLI C T O F I NTE R E S T
The authors confirm that there are no conflicts of interest in this work.

AUTH O R CO NTR I B UTI O N S
JY conceived the study, performed the experiments and made writing-original draft preparation. ZZ and SL provided insight and participated in data analysis. SL was responsible for data curation. ZZ and BL contributed to funding acquisition. XW helped with the writing/ review of the manuscript and provided intellectual input. BL and XW contributed to supervision and project administration.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available from the corresponding author upon reasonable request.