Introduction

Bladder cancer (BLCA) is among the most prevalent cancers worldwide, with 549,393 new cases reported in 2018 [1]. The risk of bladder cancer is approximately 1.1% for men and 0.27% for women [2]. BLCA can be divided into two major groups based on tumor stage: non-muscle-invasive bladder cancer (NMIBC) and muscle-invasive bladder cancer (MIBC) [3,4,5]. 20–30% of patients with NMIBC will progress to MIBC, and once the progression is identified, the patient's prognosis decreases [6,7,8]. Therefore, it is necessary to study the pathogenesis of bladder cancer.

MicroRNA (miRNA) is a class of short non-coding RNA molecules with 19 to 25 nucleotides in length, that functions in RNA silencing and post-transcriptional regulation of gene expression [9]. As a result, these mRNA molecules are silenced through the following processes: cleavage of the mRNA strand into two pieces, destabilization of the mRNA through shortening of its poly(A) tail, and less efficient translation of the mRNA into proteins by ribosomes [10, 11]. More and more studies focus on the regulatory pair of miRNA-mRNA, exploring its mechanism in the occurrence and development of diseases [12, 13].

We downloaded 2 miRNA datasets and 6 mRNA datasets from the Gene Expression Omnibus (GEO) database and combined the data from the TCGA database to screen for differential miRNAs (DE-miRNAs) and mRNA (DE-mRNAs) between BLCA and normal tissues. Interactions between DE-miRNAs and DE-mRNAs were determined using Tarbase and miRTarbase databases, where the miRNA-mRNA regulation pairs were validated experimentally. Then, we further validated the DE-miRNAs and DE-mRNAs in 28 pairs of BLCA tissues by qRT-PCR. The correlation analysis between the selected miRNA-mRNAs regulation pair and clinical, survival and tumor-related phenotypes was performed in this study. In summary, the interaction of the miRNA-mRNA regulatory pair had been researched in detail to provide a new idea or strategy for BLCA.

Materials and methods

Data acquisition and processing of miRNA and mRNA expression profiles

We downloaded the miRNA and mRNA microarray expression datasets of BLCA from the Gene Expression Omnibus (GEO) database (http://www.ncbi.nlm.nih.gov/geo/). The TCGA-BLCA miRNA and mRNA sequencing expression profile and related clinicopathological data were downloaded from the GDC data portal of the National Cancer Institute (https://portal.gdc.cancer.gov/). An overview of the workflow steps is shown in Fig. 1. DE-miRNAs and DE-mRNAs were screened by the web analysis tool GEO2R in the GEO database (http://www.ncbi.nlm.nih.gov/geo/geo2r/) and “limma” and “edgeR” R packages.

Fig. 1
figure 1

Flow chart for identifying the miRNA-mRNA regulatory pairs and the comprehensive analysis of regulatory pairs’ role in bladder cancer (BLCA)

Full size image

Identification and function analysis of miRNA-mRNA regulation pairs

TarBase is a reference database specifically designed to index experimentally supported miRNA targets, integrating information on cell type–specific miRNA gene regulation, while hundreds of thousands of miRNA-binding locations have been reported [14]. miRTarBase is a comprehensively annotated and experimentally validated database of miRNA-target interactions. Tarbase and miRTarBase databases were used to construct the miRNA-mRNA regulatory pairs [15]. Pearson correlation analysis of miRNA and mRNA in TCGA-BLCA was performed to filter the miRNA-mRNA regulation pairs. The online tool DAVID (http://david.abcc.ncifcrf.gov/) is a comprehensive tool for researchers and scholars to understand the biological significance behind multiple genes. The DIANA-MirPath is a miRNA pathway analysis web-server, and Hiplot is a comprehensive web platform for scientific data visualization [16]. The DAVID, DiANA-MirPath, and Hiplot were used for Gene Ontology (GO) functional analysis and Kyoto Encyclopedia of Genes Genomes (KEGG) pathways analysis.

Ethical approval and Information of participants

The study was conducted in accordance with the guidelines of the Hospital Ethics Committee and approved by the Institutional Review Boards of the First Affiliated Hospital of Nan**g Medical University (ID: 2016-SRFA-148). Each participant signed informed consent in advance. All the participants of the experiment were recruited from the First Affiliated Hospital of Nan**g Medical University from 2016 to 2017. All samples were obtained from radical cystectomy and did not receive neoadjuvant chemotherapy before surgery. Then, formalin fixation and paraffin embedding (FFPE) specimens of tumor and normal tissues were obtained after operation and their pathological data were recorded. The clinical characteristics of the 28 BLCA patients are shown in Table 1.

Table 1 Clinicopathological and molecular features of BLCA patients
Full size table

Extraction of RNA and quantitative reverse transcription polymerase chain reaction

The TIANGEN RNAprep Pure FFPE kit (Tiangen, Bei**g, China) was used to isolate total RNA from FFPE samples according to the manufacturer’s protocol. The acquired RNA from each sample was lysed in 100 μl RNase-free water and stored at – 80 °C until use. The concentration and purity of RNA samples were measured using the NanoDrop ND-1000 spectrophotometer (NanoDrop, Wilmington, DE, USA). External validation was performed by qRT-PCR using PrimeScript RT Reagent Kit (Takara) and SYBR Premix Ex Taq II (Takara) after adding a poly(A) tail to RNA by Poly(A) Polymerase Kit (Takara). The sequences of PCR primers are listed in Table S1. The expression levels of miRNAs and mRNAs in tissue samples were calculated using the 2–ΔΔCt method (RNU6B[U6] as endogenous reference miRNA and GAPDH as endogenous reference mRNA for sample normalization; ΔCt = Ct miRNA − Ct normalizer; Ct: the threshold cycle) [17].

Evaluation of interactions of miRNA-mRNA regulation pairs and tumor-relative phenotypes

Single sample gene set enrichment analysis (ssGSEA) is an extension of the GSEA method, which allows the definition of an enrichment score representing the absolute degree of enrichment of the gene set in each sample within a given dataset [29]. miR-17-5p promotes tumor proliferation by targeting PTEN and P21 [30]. It has been reported that miR-17-5p inhibits cell growth and promotes apoptosis of cervical cancer cells by targeting TP53INP1 [31]. In addition, miR-17-5p represses migration and invasion by directly targeting KCa1.1 and ERBB3 [32, 33]. miR-17-5p is a metastasis suppressor, and miR-17-5p plays an inhibitory role by targeting ETV1 and AIB1 [34, 35]. In this study, TGFBR2 was demonstrated to be downregulated in bladder cancer tissues. TGFBR2 has been reported to regulate the Hedgehog pathway and cervical cancer cell proliferation and migration by mediating SMAD4 [36]. Downregulation of TGFBR2 promotes the migration and invasion of CRC cells in colorectal cancer [37]. In conclusion, miR-17-5p and TGFBR2 play an essential role in tumor genesis and development.

This study also indicated that the miRNA-mRNA regulation pair had good diagnostic efficacy. We combined miR-17-5p and TGFBR2 as a panel using the logistic regression analysis and the AUC of the panel was 0.929 (95% CI: 0.885–0.972, p < 0.0001) in TCGA-BLCA and 0.767 (95% CI: 0.643–0.891, p = 0.001) in the external validation. The DCA is a method to evaluate prediction models and diagnostic tests that also showed that the miRNA-mRNA regulation pairs have an excellent diagnostic performance in distinguishing BLCA from normal patients. The expression of miR-17-5p and TGFBR2 showed no significant difference in age, gender, and TNM stage. The biomarker showed stable diagnostic performance in identifying BLCA patients regardless of the clinicopathological parameters.

Numerous studies support the critical role of immune infiltration in cancer development. The extent of T cell infiltration in tumors can predict a patient’s response to cancer immunotherapy [38]. M1 macrophages tend to polarize into M2 macrophages at advanced stages of the tumor and thus have cancer-promoting functions [39]. Correlation analysis between the miRNA-mRNA regulation pair and ssGSEA showed that the miRNA-mRNA regulation pair were related to Immunoregulatory interactions between a Lymphoid and a non-Lymphoid cell. Correlation analysis showed that miR-17-5p was negatively correlated with Macrophages M1. The tumor microenvironment plays a crucial role in the occurrence and development of tumors, and immune infiltration is one of the most essential features [40]. MiR-17-5p by targeting TGFBR2 could have an impact on the tumor microenvironment. Therefore, miR-17-5p and TGFBR2 have a necessary relationship with tumor immunity.

Although we carried out a comprehensive analysis and experimental verification of the miRNA-mRNA regulatory pairs involved in BLCA, there are still some deficiencies in this study, such as insufficient sample size and lack of studies on the mechanisms of DE-miRNAs and DE-mRNAs. Therefore, further studies on larger clinical samples and corresponding experiments are needed.

Conclusion

In summary, we have identified a miRNA-mRNA regulatory pair (miR-17-5p and TGFBR2) that may be involved in the pathogenesis of BLCA and played an important role in disease diagnosis, tumor immunity, and other clinical applications.