A total of 495 cases containing both gene expression data (HTSeq-Counts) and clinical information from the PRAD database were obtained from The Cancer Genome Atlas (TCGA) for further analysis. HTSeq-Counts data were transformed into transcripts per million reads (TPM). Data of 495 cases were used for survival analysis. Next, the characteristics of patients consists of T stage, N stage, Gleason score of pathologic of surgical specimens, and progression-free interval (PFI) result. Pathological T and N stage were performed according to the extent of tumor invasion and the presence of lymphatic metastasis. This study satisfied the publication requirements stated by TCGA (http://cancergenome.nih.gov/publications-/publicationguidelines).

The human prostate cancer cell line PC-3 originated from bone marrow metastases in a 62-year-old white male patient diagnosed with grade IV prostate cancer. Prostate cancer cells DU145 were established from a brain metastasis of a 69-year-old Caucasian patient with prostate cancer. The cells were obtained from the National Collection of Authenticated Cell Culture at the Chinese Academy of Science (Shanghai, China). PC-3 and DU145 cells were cultured in MEM supplemented with 10% fetal bovine serum (FBS) (Gibco, USA). Cell authentication was validated using STR profiling.

Small interfering RNAs (siRNAs) targeting MAFG-AS1 were obtained from GenePharma (Shanghai, China), and the sequence information targeting MAFG-AS1 was as follows: siMAFG-AS1-1, GGAGTCAGGGCAATTCCAA; siMAFG-AS1-2, GGTAACATAGAGACCCTAT.

Total RNA was extracted from prostate cancer cells using TRIzol reagent (Invitrogen, USA) according to the manufacturer’s protocol. Total RNA was then reverse transcribed to cDNA using random primers using a Revert Aid First Strand cDNA Synthesis kit (Thermo Fisher, USA). RT-qPCR was performed using TB Green Premix Ex Taq (Takara, Germany). GAPDH was used as an internal control. qPCR primers were synthesized from BioSune (Shanghai, China). The primers used were as follows: MAFG-AS1-F: CGGGAGGAAGATAAACGGGG, MAFG-AS1-R: TGACCACGGAACACCTTCAG, GAPDH-F: CTGGGCTACACTGAGCACC, GAPFH-R: AAGTGGTCGTTGAGGGCAATG.

For the cell proliferation assay, a total of 3000 prostate cancer cells were seeded into 96-well plates and incubated in 10% CCK-8 medium for one hour at 0 hour, 24 hours, 48 hours, and 72 hours after seeding. The absorbance was measured at 450 nm with a spectrophotometer. For the colony formation assay, prostate cancer cells were seeded in 6-well plates at a density of 200 cells per well and cultured for 2 weeks. The cells were then fixed and subsequently examined by crystal violet staining.

For the transwell migration assay, 20000 cells were suspended in 200 μl of medium without FBS and were seeded in the upper chamber of transwell inserts (Corning, USA). 500 μl medium with 20% FBS were added into the lower chamber. The cells were then incubated for 18 h. For the transwell invasion assay, 30000 cells were suspended in 200 μl of medium without FBS and were seeded on the upper chamber of transwell inserts coated with Matrigel (BD Biosciences, USA). The cells were then cultured for 24 h. Cells were fixed with 4% PFA and stained with a crystal violet staining solution. Images were captured with a ×20 objective using a Leica DM LB light microscope and the number of cells was counted using ImageJ.

All statistical analyses were conducted using SPSS (22.0). Pearson’s χ2 test was used to determine the correlation between MAFG-AS1 expression and clinicopathological variables. The t-test was used to determine statistically significant differences between the two groups. Kaplan–Meier analysis was performed to compare the survival time differences between the MAFG-AS1 high expression group and low expression group. The log-rank test p < 0.05 suggested the significance of survival time differences. The hazard risk of the individual indicators was estimated using hazard ratios (HRs) with 95% confidence intervals (CIs). All reported P-values were two-sided and P-values of less than 0.05 were considered to be significant. * represents P < 0.05, ** represents P < 0.01, and *** represents P < 0.001.

The data collected from TCGA in October 2019 contained 495 tumor samples with both clinical information and gene expression data (Table 1) and the clinical features of the patients included age, TNM stage, Gleason scores, and PFI events. To evaluate the expression of MAFG-AS1 in prostate cancer, we compared MAFG-AS1 expression in 495 prostate cancer and 50 adjacent normal tissues, and the results suggested that MAFG-AS1 was upregulated in prostate cancer (Figure 1A). Similarly, by comparing 50 pairs of prostate cancer tissues and adjacent normal tissues, we also found that MAFG-AS1 was highly expressed in prostate cancer (Figure 1B). In addition, we collected 18 pairs of prostate cancer tissues and adjacent normal tissues, and the results of RT-qPCR demonstrated that MAFG-AS1 expression was upregulated in prostate cancer (Figure 1C). Next, the results of RT-qPCR indicated that MAFG-AS1 expression in prostate cancer cell lines PC-3 and DU145 was higher than that in the normal prostate epithelial cell line RWPE-1 (Figure 1D). These results demonstrated that MAFG-AS1 expression is overexpressed in prostate cancer.

Logistic regression analysis demonstrated that higher MAFG-AS1 expression, regarded as an independent variable, was correlated with better prognostic characteristics (Table 2). High MAFG-AS1 expression in the PRAD cohort was significantly associated with T classification (OR = 1.952 for T3&T4 vs. T2, P < 0.001), N classification (OR = 2.005 for N1 vs. N0, P = 0.008), and Gleason score (OR = 2.074 for 8&9&10 vs. 6&7, P < 0.001). These results revealed that prostate cancer with high MAFG-AS1 expression is more likely to be in a primitive stage than those with low MAFG-AS1 expression. The expression level of MAFG-AS1 was higher in the T3&T4 stage than in the T2 stage (Figure 2A). Similarly, MAFG-AS1 expression was higher in advanced prostate cancer according to N stage, Gleason scores, and PFI events (Figures 2B–D). Consistently, MAFG-AS1 high expression was significantly correlated with poor prognosis in patients with prostate cancer patients (Figure 2E). Furthermore, univariate Cox regression analysis showed that high MAFG-AS1 expression was correlated with a poor PFI (hazard ratio [HR]: 1.985; 95% confidence interval [CI]: 1.299- 3.035; P < 0.01), and other clinical variables, including advanced T stage, N stage, and Gleason score, remained associated with a poor prognosis (Table 3). Multivariate analysis suggested that high MAFG-AS1 expression was independently associated with poor PFI (HR = 1.78; CI: 1.121- 2.847; P = 0.015).

We then performed a stratified analysis based on the clinical information of prostate cancer patients. KM-plot analysis revealed that high MAFG-AS1 expression was associated with a poor prognosis in T3&T4 stage patients with prostate cancer; however, MAFG-AS1 expression was not associated with prognosis in T2 stage patients (Figures 2F, G). In patients with or without lymph node metastasis, MAFG-AS1 expression could be used as a prognostic marker, indicating poor PFI (Figures 2H, I). Similarly, in patients with high or low Gleason scores, high expression of MAFG-AS1 was correlated with poor prognosis (Figures 2J, K). In addition, high MAFG-AS1 expression was associated with a poor prognosis in prostate cancer patients with PSA levels more than 4 ng/ml or less than 4 ng/ml (Figures 2L, M).

To critically evaluate the prognostic value of MAFG-AS1 in prostate cancer, we performed a tissue microarray (Figure 3A; Figure S1), and the clinical information of TMA patient cohort was shown in Table 4. The ISH results demonstrated that MAFG-AS1 expression was higher in the T3&T4 stages than in the T2 stage (Figure 3B). Similarly, MAFG-AS1 was upregulated in the N1 stage compared to the N0 stage (Figure 3C). Furthermore, prostate cancer with a higher Gleason score showed higher MAFG-AS1 expression (Figure 3D). KM-plot analysis suggested that high expression of MAFG-AS1 indicated a poor overall survival (OS) in the TMA patient cohort (Figure 3E). These results demonstrate that MAFG-AS1 is correlated with prostate cancer clinical features, and high expression of MAFG-AS1 is associated with a poor prognosis in prostate cancer.

Next, we performed a functional gene enrichment analysis of MAFG-AS1. We searched for the top 500 genes related to MAFG-AS1 expression according to the TCGA-PRAD database. The results showed that these genes were enriched in ribosome biogenesis, ribonucleoprotein complex subunit organization, ribonucleoprotein complex assembly, and rRNA metabolic process in biological process (BP). Results also showed that these genes were enriched in ribosome and ribosomal subunit in cellular component (CC), and enriched in structural constituent of ribosome and ribonucleoprotein complex binding in molecular function (MF) (Figures 4A–C). Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis revealed that MAFG-AS1 related genes were enriched in the ribosome and DNA replication pathways (Figure 4D).

Since functional enrichment analysis revealed that MAFG-AS1 may be involved in ribosome biogenesis, we investigated the correlation between MAFG-AS1 and ribosome-related genes. Among the top 500 genes most related to MAFG-AS1 expression, a number of genes were the components of ribosomes or involved in ribosome biosynthesis. For example, MAFG-AS1 expression was positively correlated with 60S ribosomal proteins (RPL6, RPL7, RPL7A, RPL8, etc.) and 40S ribosomal proteins (RPS2, RPS3, RPS3A, RPS7, etc.) (Figure 4E). Collectively, these results suggest that MAFG-AS1 may be involved in ribosome biogenesis to regulate prostate cancer tumorigenicity.

To further elucidate the role of MAFG-AS1 expression in prostate cancer, we selected two prostate cancer cell lines (PC-3 and DU145) for subsequent research. RT-qPCR results showed that MAFG-AS1 expression was effectively downregulated in PC-3 and DU145 cells transfected with si-MAFG-AS1 (Figure 5A). The results of the CCK-8 assay demonstrated that MAFG-AS1 knockdown significantly inhibited prostate cancer cell viability (Figures 5B, C). The colony formation assay demonstrated that downregulation of MAFG-AS1 decreased the colony formation rate in prostate cancer cells (Figure 5D). For further studies, we conducted a transwell assay to clarify the role of MAFG-AS1 in prostate cancer migration and invasion. The results suggested that migration and invasion abilities were prominently impaired in MAFG-AS1 knockdown prostate cancer cells (Figures 5E, F). Taken together, we demonstrated that MAFG-AS1 knockdown significantly inhibited prostate cancer cell proliferation, migration, and invasion.