Expression assays of selected lncRNAs in non-functioning pituitary adenomas.

According to the data accessible in this database until the date of 2022–10-9, 28 and 49 lncRNAs regulate EGFR and STAT3 genes in different cancers, respectively. These lncRNAs are listed in Tables 2 and 3 . Table 2 LncRNAs regulating EGFR gene in different cancers Cancer name LncRNA Pancreatic cancer AFAP1-AS1 Hepatocellular carcinoma AL163636.1 Malignant glioma BCAR4 Lung cancer CAR10 Esophagus squamous cell carcinoma CYTOR Clear cell renal cell carcinoma DGCR5 Non-small cell lung cancer DUXAP9-206 Gastric cancer; Hepatocellular carcinoma; Renal cancer EGFR-AS1 Oral squamous cell carcinoma ELDR Non-small cell lung cancer FAM201A Lung adenocarcinoma GAS5 Non-small cell lung cancer H19 Colorectal cancer HOTAIR Malignant glioma HOXA-AS2 Head and neck squamous cell carcinoma LINC00052 Colorectal cancer LINC00265 Lung adenocarcinoma; Non-small cell lung cancer LINC00460 Hepatocellular carcinoma LINC01225 Tongue cancer Lnc-EGFR Prostate cancer LOXL1-AS1 Hepatocellular carcinoma MALAT1 Ovarian cancer MEG3 Colorectal cancer SCARNA2 Colorectal cancer SLCO4A1-AS1 Glioblastoma SNHG15 Malignant glioma SNHG16 Breast cancer TINCR Pancreatic cancer XIST Table 3 LncRNAs regulating STAT3 gene in different cancers Cancer name LncRNA Colorectal cancer AB073614 Endometriosis AFAP1-AS1 Osteosarcoma AK093407 Hepatocellular carcinoma ARSR Colorectal cancer CASC2 Malignant glioma CASC9 Epithelial ovarian cancer CCAT1 Hepatocellular carcinoma; Triple-negative breast cancer CERNA2 Oral squamous cell carcinoma circPVT1 Nasopharynx carcinoma DANCR Colon cancer FALEC Ovarian cancer FEZF1-AS1 Laryngeal squamous cell carcinoma FOXD2-AS1 Cervical cancer GAS5 Breast cancer; Esophageal cancer; Lung cancer H19 Gastric cancer; Hepatocellular carcinoma HOTAIR Liver cancer HOXA11-AS Cervical cancer LINC00052 Cervical cancer LINC00240 Retinoblastoma LINC00324 Non-small cell lung cancer LINC00346 Hepatocellular carcinoma; Non-small cell lung cancer LINC00589 Kidney clear cell carcinoma LINC00997 Hepatocellular carcinoma LINC01433 Breast cancer Lnc-BM Hepatocellular carcinoma lnc-DILC Renal cell carcinoma LNCSRLR Hepatocellular carcinoma LSINCT5 Hepatoblastoma LUCAT1 Non-small cell lung cancer; Retinoblastoma MALAT1 Esophagus adenocarcinoma MIR22HG Colorectal cancer MNX1-AS1 Breast cancer; Gastric cancer; Hepatocellular carcinoma NEAT1 Ovarian cancer NORAD Endometrial cancer PCGEM1 T-cell acute lymphocytic leukemia PPM1A-AS Hepatoblastoma PVT1 Colorectal cancer RP11-468E2.5 Cervical cancer SNHG12 Retinoblastoma SNHG14 Ovarian cancer THORLNC Hepatocellular carcinoma TINCR Non-small cell lung cancer TNK2-AS1 Head and neck squamous cell carcinoma; Prostate cancer TP53COR1 Hepatocellular carcinoma TPTEP1 Urinary bladder cancer UCA1 Hepatocellular carcinoma; Lung cancer WFDC21P B-cell acute lymphocytic leukemia; Non-small cell lung cancer ZEB1-AS1 Esophagus squamous cell carcinoma; Triple-negative breast cancer ZFAS1 LncRNAs regulating EGFR gene in different cancers LncRNAs regulating STAT3 gene in different cancers According to the available data in this database until the date of 2022–10-9, 119 lncRNAs were found to be examined in pituitary adenomas. Then, with the purpose of excluding common lncRNAs, the Venny 2.1 online software ( https://bioinfogp.cnb.csic.es/tools/venny/ ) was used. For EGFR and STAT3 genes, 8 and 10 common lncRNAs were found, respectively. These lncRNAs were excluded from this study. Then, 20 and 39 novel lncRNAs were used for mentioned genes, respectively (Tables 4 and 5 ). Table 4 Genes used for EGFR Venn analysis and resulted novel lncRNAs LncTarD2.0 LncRNADisease v2.0 Common LncRNAs Novel LncRNAs AFAP1-AS1 MEG3 BCAR4 AFAP1-AS1 AL163636.1 CCAT2 DGCR5 AL163636.1 BCAR4 HOTAIR GAS5 CAR10 CAR10 MALAT1 H19 CYTOR CYTOR GADD45G HOTAIR DUXAP9-206 DGCR5 H19 MALAT1 EGFR-AS1 DUXAP9-206 AC091891.2 MEG3 ELDR EGFR-AS1 AIRN XIST FAM201A ELDR anti-NOS2A HOXA-AS2 FAM201A ATP6V1G2-DDX39B LINC00052 GAS5 ATXN8OS LINC00265 H19 B1 SINE RNA LINC00460 HOTAIR B2 SINE RNA LINC01225 HOXA-AS2 BACE1-AS Lnc-EGFR LINC00052 BCAR4 LOXL1-AS1 LINC00265 BCYRN1 SCARNA2 LINC00460 BDNF-AS1 SLCO4A1-AS1 LINC01225 BOK-AS1 SNHG15 Lnc-EGFR BPESC1 SNHG16 LOXL1-AS1 BX118339 TINCR MALAT1 C15orf2 MEG3 C1QTNF9B-AS1 SCARNA2 CASC2 SLCO4A1-AS1 CBR3-AS1 SNHG15 CDKN2B-AS1 SNHG16 CECR3 TINCR CECR9 XIST CHL1-AS2 CRNDE DAOA-AS1 DAPK1 DGCR5 DISC2 DLEU1 DLEU2 DLG2-AS1 DLX6-AS1 DMPK DNM3OS DSCAM-AS1 EPB41L4A-AS1 ESRG FMR1-AS1 GAS5 GDNF-AS1 GNAS-AS1 HAR1A HAR1B HCP5 HIF1A-AS1 HLA-AS1 HTT-AS HULC HYMAI IFNG-AS1 IGF2-AS IPW KCNQ1DN KCNQ1OT1 LDMAR LINC00032 LINC00271 LINC00312 LINC00538 LINC00901 MAP3K14 MESTIT1 MIAT MIR100HG MIR155HG MIR17HG MIR31HG MKRN3-AS1 MYCNOS NAMA NDM29 NEAT1 NRON PCA3 PCAT1 PCGEM1 PDZRN3-AS1 PICSAR PINC PINK1-AS PISRT1 PPP3CB PRINS PSORS1C3 PTCSC1 PVT1 RMST RN7SK RN7SL1 RRP1B SCAANT1 SNHG11 SNHG3 SNHG4 SNHG5 SOX2-OT SPRY4-IT1 SRA1 TCL6 TDRG1 TERC TRAF3IP2-AS1 TUG1 TUSC7 Ube3a-as uc021oqb.2 uc061gkt.1 UCA1 WRAP53 WT1-AS XIST ZFAS1 ZFAT-AS1 A130040M12Rik Table 5 Genes used for STAT3 Venn analysis and resulted novel lncRNAs LncTarD2.0 LncRNADisease v2.0 Common LncRNAs Novel LncRNAs AB073614 MEG3 CASC2 AB073614 AFAP1-AS1 CCAT2 GAS5 AFAP1-AS1 AK093407 HOTAIR H19 AK093407 ARSR MALAT1 HOTAIR ARSR CASC2 GADD45G MALAT1 CASC9 CASC9 H19 NEAT1 CCAT1 CCAT1 AC091891.2 PCGEM1 CERNA2 CERNA2 AIRN PVT1 circPVT1 circPVT1 anti-NOS2A UCA1 DANCR DANCR ATP6V1G2-DDX39B ZFAS1 FALEC FALEC ATXN8OS FEZF1-AS1 FEZF1-AS1 B1 SINE RNA FOXD2-AS1 FOXD2-AS1 B2 SINE RNA HOXA11-AS GAS5 BACE1-AS LINC00052 H19 BCAR4 LINC00240 HOTAIR BCYRN1 LINC00324 HOXA11-AS BDNF-AS1 LINC00346 LINC00052 BOK-AS1 LINC00589 LINC00240 BPESC1 LINC00997 LINC00324 BX118339 LINC01433 LINC00346 C15orf2 Lnc-BM LINC00589 C1QTNF9B-AS1 lnc-DILC LINC00997 CASC2 LNCSRLR LINC01433 CBR3-AS1 LSINCT5 Lnc-BM CDKN2B-AS1 LUCAT1 lnc-DILC CECR3 MIR22HG LNCSRLR CECR9 MNX1-AS1 LSINCT5 CHL1-AS2 NORAD LUCAT1 CRNDE PPM1A-AS MALAT1 DAOA-AS1 RP11-468E2.5 MIR22HG DAPK1 SNHG12 MNX1-AS1 DGCR5 SNHG14 NEAT1 DISC2 THORLNC NORAD DLEU1 TINCR PCGEM1 DLEU2 TNK2-AS1 PPM1A-AS DLG2-AS1 TP53COR1 PVT1 DLX6-AS1 TPTEP1 RP11-468E2.5 DMPK WFDC21P SNHG12 DNM3OS ZEB1-AS1 SNHG14 DSCAM-AS1 THORLNC EPB41L4A-AS1 TINCR ESRG TNK2-AS1 FMR1-AS1 TP53COR1 GAS5 TPTEP1 GDNF-AS1 UCA1 GNAS-AS1 WFDC21P HAR1A ZEB1-AS1 HAR1B ZFAS1 HCP5 HIF1A-AS1 HLA-AS1 HTT-AS HULC HYMAI IFNG-AS1 IGF2-AS IPW KCNQ1DN KCNQ1OT1 LDMAR LINC00032 LINC00271 LINC00312 LINC00538 LINC00901 MAP3K14 MESTIT1 MIAT MIR100HG MIR155HG MIR17HG MIR31HG MKRN3-AS1 MYCNOS NAMA NDM29 NEAT1 NRON PCA3 PCAT1 PCGEM1 PDZRN3-AS1 PICSAR PINC PINK1-AS PISRT1 PPP3CB PRINS PSORS1C3 PTCSC1 PVT1 RMST RN7SK RN7SL1 RRP1B SCAANT1 SNHG11 SNHG3 SNHG4 SNHG5 SOX2-OT SPRY4-IT1 SRA1 TCL6 TDRG1 TERC TRAF3IP2-AS1 TUG1 TUSC7 Ube3a-as uc021oqb.2 uc061gkt.1 UCA1 WRAP53 WT1-AS XIST ZFAS1 ZFAT-AS1 A130040M12Rik Genes used for EGFR Venn analysis and resulted novel lncRNAs Genes used for STAT3 Venn analysis and resulted novel lncRNAs GEPIA2 showed that 7 and 13 lncRNAs were significantly correlated to EGFR and STAT3 genes, respectively, in normal pituitary tissue (Table  6 ). Among these lncRNAs, EGFRAS1 (for EGFR gene); and SNHG12, FALEC and LINC00240 (for STAT3 gene) were selected for experimental investigation in the present study (Fig.  1 ). Table 6 LncRNAs significantly correlated to EGFR and STAT3 genes STAT3 -related lncRNAs EGFR -related lncRNAs FALEC CAR10 FEZF1-AS1 EGFR-AS1 LINC00240 LINC00265 LINC00324 LINC00460 LINC00346 SLCO4A1-AS1 LINC01433 SNHG15 LUCAT1 TINCR MIR22HG SNHG12 SNHG14 TINCR TPTEP1 ZEB1-AS1 Fig. 1 Correlation analysis results of ( A ) LINC00240 , ( B ) FALEC , ( C ) SNHG12 , and ( D ) EGFR-AS1 LncRNAs significantly correlated to EGFR and STAT3 genes Correlation analysis results of ( A ) LINC00240 , ( B ) FALEC , ( C ) SNHG12 , and ( D ) EGFR-AS1 Table 7 shows the characteristics of selected genes. Table 7 Characteristics of selected genes Name/Gene ID Accession number Location Gene type EGFR NM_001346897.2 , NM_001346898.2 , NM_001346899.2 , NM_001346900.2 , NM_001346941.2 , NM_005228.5 , NM_201282.2 , NM_201283.2 , NM_201284.2 7p11.2 Protein coding EGFR-AS1 NR_047551.1 7p11.2 ncRNA FALEC NR_051960.2 , NR_186305.1 , NR_186306.1 , NR_186307.1 1q21.2 ncRNA LINC00240 NR_026775.2 6p22.2 ncRNA SNHG12 NR_024127.2 , NR_146381.1 , NR_146382.1 , NR_146383.1 , NR_146384.1 , NR_146385.1 , NR_146386.1 , NR_146387.1 1p35.3 ncRNA Characteristics of selected genes Table 8 shows an overview of included samples. Table 8 Overview of included NFPA samples Characteristic Values Age (years) (Mean ± SD) 51.31 ± 12.97 Gender  Male 31  Female 10 Tumor type (Number)  Microadenoma 2  Macroadenoma 32  Giant 7 Invasiveness (Number)  Non-invasive 34  Invasive 6 Hardy classification (Number)  Grade 1 2  Grade 2 0  Grade 3 35  Grade 4 4 Knosp classification (Number)  Grade 1 14  Grade 2 13  Grade 3a 8  Grade 3b 4  Grade 3a*b 1 Diseases Duration (Day) (Mean ± SD) 423.7 ± 565.5 Tumor volume (mm 3 ) (Mean ± SD) 756.5 ± 474.4 CSF leak (Number)  No 20  Low flow 10  High flow 3 Sphenoid type (Number)  Sellar 38  Presellar 3 Overview of included NFPA samples We detected significant difference in the expression of all mentioned genes except for FALEC between NFPA samples and their corresponding NTATs (Fig.  2 ). Fig. 2 The relative expression levels of five studied genes in non-functional pituitary adenoma (NFPA) tissue types relative to the normal tissues adjacent to the tumors (NTATs).–delta Ct data was plotted as individual values (including the mean with 95% CI). Data was analyzed using the Wilcoxon rank-sum test or paired t test. Asterisks indicate significant difference between two mentioned groups (***P value < 0.001, ****P value < 0.0001, ns; non-significant) The relative expression levels of five studied genes in non-functional pituitary adenoma (NFPA) tissue types relative to the normal tissues adjacent to the tumors (NTATs).–delta Ct data was plotted as individual values (including the mean with 95% CI). Data was analyzed using the Wilcoxon rank-sum test or paired t test. Asterisks indicate significant difference between two mentioned groups (***P value < 0.001, ****P value < 0.0001, ns; non-significant) All studied genes were down-regulated in NFPA samples compared with NTATs, except for FALEC whose expression was not different between these two sets of samples (Table  9 ). EGFR was the most significantly down-regulated gene in NFPAs (Expression ratio (95% CI) = 0.009 (0.002–0.04), P value < 0.0001). Table 9 Expression of genes in NFPA tissues compared with NTATs. The expression ratio of each gene is shown as mean and 95% confidence intervals and SEM Studied genes Expression ratio (95% CI) SEM P value EGFR 0.009 (0.002–0.04) 1.04  < 0.0001 EGFR-AS1 0.17 (0.07–0.44) 0.66 0.0011 FALEC 1.5 (0.4–5.65) 0.85 0.73 LINC00240 0.2 (0.05–0.73) 0.85 0.022 SNHG12 0.04 (0.015–0.1) 0.66  < 0.0001 Expression of genes in NFPA tissues compared with NTATs. The expression ratio of each gene is shown as mean and 95% confidence intervals and SEM ROC curve analyses proposed that the expression levels of SNHG12 , EGFR , EGFR-AS1 and LINC00240 can be used to distinguish NFPAs from NTATs with AUC values of 0.88, 0.83, 0.7 and 0.66, respectively (Fig.  3 and Table  10 ). Fig. 3 ROC curves of four differentially expressed genes for discrimination of NFPA tumors from the normal tissues adjacent to the tumors. AUC indicates area under the ROC curve Table 10 ROC curve analyses for four differentially expressed genes for discrimination of NFPA tumors genes from the normal tissues adjacent to the tumors SNHG12 EGFR LINC00240 EGFR-AS1 AUC ± SD Sensitivity Specificity P Value AUC ± SD Sensitivity Specificity P Value AUC ± SD Sensitivity Specificity P Value AUC ± SD Sensitivity Specificity P Value 0.88 ± 0.04 0.87 0.90  < 0.0001 0.83  ±  0.04 0.87 0.78  < 0.0001 0.66 ± 0.06 0.8 0.53 0.011 0.7 ± 0.05 0.51 0.85 0.001 ROC curves of four differentially expressed genes for discrimination of NFPA tumors from the normal tissues adjacent to the tumors. AUC indicates area under the ROC curve ROC curve analyses for four differentially expressed genes for discrimination of NFPA tumors genes from the normal tissues adjacent to the tumors Spearman's correlation analyses showed significant correlations between FALEC and EGFR-AS1 in both types of tissues, and between FALEC and EGFR in NFPAs. Moreover, expression of LINC00240 was correlated with EGFR-AS1 , FALEC and SNHG12 in NFPAs (Table  11 ). Table 11 Spearman’s correlations between five studied genes among NFPAs ( N  = 41) and NTATs ( N  = 41) EGFRAS1 FALEC LINC00240 SNHG12 NTAT NFPA NTAT NFPA NTAT NFPA NTAT NFPA EGFR − 0.15 0.14 0.08 0.37* − 0.2 0.14 0.24 0.01 EGFRAS1 0.47** 0.54** 0.04 0.53** 0.01 0.23 FALEC − 0.21 0.46** − 0.13 0.22 LINC00240 0.05 0.42** * P  < 0.05 ** P  < 0.01 Spearman’s correlations between five studied genes among NFPAs ( N  = 41) and NTATs ( N  = 41) * P  < 0.05 ** P  < 0.01 We detected no association between clinicopathological data and expression of mentioned genes (Table  12 ). Table 12 Expression levels of genes in NFPA patients with different clinicopathologic factors as analyzed by Mann–Whitney and Kruskal–Wallis one-way ANOVA tests Parameters Subclasses Number (%) Relative expression level EGFR (mean ± SD) P Relative expression level EGFRAS1 (mean ± SD) P Relative expression level FALEC (mean ± SD) P Relative expression level of LINC00240 (mean ± SD) P Relative expression level of SNHG12 (mean ± SD) P Age 22–48 19 − 14.2 ± 3.96 0.27 − 2.06 ± 3.02 0.54 − 7.77 ± 2.7 0.85 − 14.5 ± 4.01 0.58 − 9.5 ± 4.27 0.46 49–77 22 − 12.6 ± 4.8 − 2.52 ± 4.77 − 8.04 ± 3.9 − 13.8 ± 4.04 − 9.8 ± 3.62 Gender Female 10 − 11.9 ± 4.8 0.18 − 3.3 ± 4.21 0.41 − 7.78 ± 4.8 0.25 − 14.6 ± 4.25 0.67 − 7.95 ± 2.65 0.07 Male 31 − 13.8 ± 4.37 − 1.99 ± 3.96 − 7.96 ± 2.9 − 14 ± 3.96 − 10.3 ± 4.08 Disease duration  < 1y 22 − 12.9 ± 4.66 0.53 − 2.15 ± 4.56 0.81 − 8.35 ± 3.83 0.55 − 14.1 ± 3.64 0.73 − 9.36 ± 3.9 0.46  ≥ 1 y 19 − 13.8 ± 4.36 − 2.49 ± 3.38 − 7.42 ± 2.82 − 14.2 ± 4.46 − 10.16 ± 3.93 Tumor volume (mm 3 )   800 15 − 13.8 ± 3.4 − 1.14 ± 4.26 − 7.8 ± 3.27 − 14.2 ± 3.36 − 8.3 ± 3.85 CSF leak No 20 − 12.5 ± 4.18 0.18 − 1.58 ± 3.76 0.36 − 7.38 ± 3.27 0.7 − 13.26 ± 3.6 0.21 − 9.7 ± 3.6 0.77 Low flow 10 − 12.4 ± 4.5 − 2.79 ± 3.67 − 7.96 ± 2.24 − 14.2 ± 4.6 − 9.1 ± 4.34 High flow 11 − 15.7 ± 4.6 − 3.19 ± 4.82 − 8.86 ± 4.43 − 15.7 ± 3.94 − 10.27 ± 4.24 Knosp classification 1 14 − 14.04 ± 5.4 0.8 − 3.8 ± 4 0.15 − 9.5 ± 4.1 0.08 − 14.9 ± 3.23 0.68 − 10.6 ± 2.8 0.24 2 13 − 13.5 ± 3.7 − 2.26 ± 3.86 − 7.5 ± 2.44 − 13.5 ± 4.6 − 8.4 ± 3.6 3 14 − 12.5 ± 4.4 − 0.78 ± 3.83 − 6.7 ± 2.9 − 13.9 ± 4.2 − 10 ± 4.8 Invasiveness Invasive 7 − 12.9 ± 4.32 0.29 − 1.98 ± 3.8 0.16 − 7.67 ± 2.9 0.48 − 13.9 ± 4.1 0.48 − 9.8 ± 4.1 0.78 Non invasive 34 − 15.5 ± 5.06 − 3.9 ± 4.7 − 9.4 ± 5.2 − 15.2 ± 3.5 − 9.3 ± 2.6 Expression levels of genes in NFPA patients with different clinicopathologic factors as analyzed by Mann–Whitney and Kruskal–Wallis one-way ANOVA tests

The bioinformatics approach used in this study was described in detail in our previous study [ 8 ]. Briefly, GSE63357 dataset [ 12 ] was downloaded from The Gene Expression Omnibus database (GEO, http://www.ncbi.nlm.nih.gov/geo/ ) [ 13 ]. GEO2R web tool ( http://www.ncbi.nlm.nih.gov/geo/geo2r/ ) was used to identify DEGs [ 14 ]. Using STRING online tool ( https://string-db.org/ ), PPI network was drawn for 1500 most significant DEGs [ 15 ]. The Cytoscape software ( https://cytoscape.org/ , version 3.8.0), the CytoHubba plugin, and the Degree algorithm were used to identify top ten hub genes [ 16 ]. The Enrichr database ( https://maayanlab.cloud/Enrichr/ ) was used for functional and pathway enrichment analyses [ 17 ]. KEGG pathways analysis showed that the HIF1 signaling pathway, among other signaling pathways, has the most significant P value. It was also shown that four hub genes, namely STAT3, MAPK1, GAPDH, EGFR, are related to this pathway. In our previous study, two genes, STAT3 and MAPK1 were experimentally investigated, and in this study, another upstream gene of this pathway, i.e. EGFR is further investigated. Next, as mentioned in the previous study, LncTarD 2.0 ( https://lnctard.bio-database.com/ ) was used to identify LncRNAs potentially regulating each of these genes in various tumors [ 18 ]. Then, each of the lncRNAs was checked in the LncRNADisease v2.0 database ( http://www.rnanut.net/lncrnadisease/ ) in terms of novelty in the field of pituitary gland tumor research [ 19 ]. Finally, correlation analysis and validation of lncRNAs were performed using the GTEx data available in GEPIA2 database ( http://gepia2.cancer-pku.cn/ ) [ 20 ]. In this way, lncRNAs EGFRAS1 (for EGFR gene), and SNHG12 , FALEC and LINC00240 (for STAT3 gene) were selected for experimental studies. Paired tumor and normal tissues adjacent to the tumors (NTATs) from 41 NFPA patients entered the current study. Patients underwent surgery between 2021 and 2022 in hospitals affiliated to Shahid Beheshti University of Medical Sciences. Surgery was performed under a microscope. The patients did not take chemoradiotherapy before surgical resection. All patients signed informed consent from. We got the approval of the study from the ethical committee of Shahid Beheshti University of Medical Sciences. EGFR , EGFR-AS1 , SNHG12 , FALEC , and LINC00240 were selected for expression assays. RNA was isolated using the RNJia extraction kit (Roje Technologies Company, Iran). The absorbance of a diluted RNA sample is measured at 260 and 280 nm. A260/280 ratio between 1.8 and 2.0 was acceptable. Then, 3 μg of extracted total RNA was used for cDNA synthesis. The AddScript Kit (ADDBIO Company, South Korea) was used for this purpose. Then, qRT-PCR was performed by using of RealQ Plus 2 × Master Mix Green with high ROX (AMPLIQON, Denmark), according to the method described in our recently published paper [ 8 ]. Primers were purchased from the METABION Company (Germany). In our previous works, we found that expression of B2M is stable among pituitary samples, thus we chose this gene as an internal control [ 8 ]. Details are shown in Table  1 . Table 1 Primers used for expression assays Name Type Sequence (5'⇾3') Primers length (bp) PCR products size (bp) Annealing temperatures (℃) EGFR -F mRNA AGGCACGAGTAACAAGCTCAC 21 177 62 EGFR -R ATGAGGACATAACCAGCCACC 21 EGFRAS1 -F LncRNA CCATCACGTAGGCTTCCTGG 20 108 62.5 EGFRAS1 -R GCATTCATGCGTCTTCACCTG 21 SNHG12 -F LncRNA TCTGGTGATCGAGGACTTCC 20 96 62.5 SNHG12 -R ACCTCCTCAGTATCACACACT 21 FALEC -F LncRNA GCAAGCGGAGACTTGTCTTTA 21 192 64 FALEC -R TCTTCCCTCTGTGAAACCTGC 21 LINC00240 -F LncRNA CAACCTCTCCTCTGGATGCTC 21 142 62 LINC00240 -R GTAGTTGAGGGTTGGCAAGGA 21 B2M -F mRNA AGATGAGTATGCCTGCCGTG 20 105 62 B2M -R GCGGCATCTTCAAACCTCCA 20 Primers used for expression assays SPSS version 22.0 (SPSS Inc., Chicago, IL) was used for analyses. Figures and ROC curves were created using GraphPad Prism version 9.0 (California, USA). Expression levels of EGFR and four lncRNA genes including EGFRAS1 , FALEC , LINC00240 and SNHG12 were compared between pituitary adenomas and NTATs. Expression levels were calculated using the Efficiency adjusted Ct values. Distribution of the values was examined using the Shapiro–wilk test. Wilcoxon matched-pairs signed rank test showed differentially expressed genes between the adenoma and NTATs. – delta Ct Data in the figures were plotted as individual values (showing mean and 95% CI values). The correlation between expressions of genes was tested using Spearman correlation coefficient. Mann–Whitney test and Kruskal–Wallis one-way ANOVA were used for comparison of expression levels between subgroups. Chi-square test was used to report the association between reported factors and expressions of genes. P value < 0.05 was considered as significant.

The current study used a combined bioinformatics and expression assays strategy to find dysregulated genes in the NFPAs. The lncRNAs found from the bioinformatics step were EGFR-AS1 for the EGFR gene, and LINC00240 , FALEC and SNHG12 for the STAT3 gene. We recently investigated expression of STAT3 in this set of samples and detected no meaningful difference in its expression between NFPAs and NTATs [ 8 ]. All other studied genes were down-regulated in NFPA samples compared with NTATs, except for FALEC whose expression was not different between these two sets of samples. Notably, EGFR was the most significantly down-regulated gene in NFPAs. A previous study showed over-expression of EGFR in more than half of the pituitary corticotroph adenomas using immunohistochemistry and Western blot methods [ 21 ]. They also reported activation of phosphorylated Erk (p-Erk) in EGFR-overexpressing adenomas [ 21 ]. Finally, authors detected correlation between expression level of EGFR and recurrence-free interval [ 21 ]. The difference between the results of our study and mentioned study might be attributed to the methods used for expression assay (real time PCR vs. immunohistochemistry and Western blot methods) and type of samples (NFPAs vs. corticotroph adenomas). Similarly, overexpression of EGFR was shown to enhance the growth of explanted ACTH-producing adenomas and elevate serum corticosterone level [ 22 ]. Another study also showed higher rate of pEGFR T693 positivity in recurrent NFPAs as compared to non-recurrent NFPAs [ 23 ]. Thus, both expression and phosphorylation of EGFR should be assessed in different types of pituitary adenomas, including NFPAs to find the importance of this nuclear factor in the pathogenesis of pituitary tumors. EGFR-AS1 was demonstrated to be over-expressed in different types of tumors in correlation with numerous clinical characteristics [ 24 ]. This lncRNA contributes to the regulation of many cellular activities, such as proliferation, invasiveness, migration, chemosensitivity, and stem cell properties [ 24 ]. However, the observed down-regulation of EGFR-AS1 in NFPAs suggests a distinct role for this lncRNA in the pathogenesis of this non-malignant condition. ROC curve analysis showed that the expressions of SNHG12 , EGFR , LINC00240 and EGFR-AS1 can be used to distinguish NFPAs from NTATs with AUC values of 0.87, 0.85, 0.74 and 0.69, respectively. Thus, SNHG12 and EGFR are suggested as potential biomarkers for NFPA. Spearman's correlation analyses showed significant correlations between FALEC and EGFR-AS1 in both types of tissues, and between FALEC and EGFR in NFPAs. Moreover, expression of LINC00240 was correlated with EGFR-AS1 , FALEC and SNHG12 in NFPAs. In fact, we did not detect the expected correlation between EGFR and EGFR-AS1 in either type of tissues. Among the assessed clinical factors, expression of FALEC tended to be associated with Knosp classification in a way that higher expression of this lncRNA was reported in higher grades. However, the association was not statistically significant. FALEC is located in focal amplicon on chromosome 1q21.2, and was shown to be abnormally expressed in several malignances, including ovarian and prostate cancers [ 25 , 26 ]. However, we did not detect up-regulation of FALEC in NFPAs. Additionally, SNHG12 levels tended to be higher in female samples with a non-significant P value. Taken together, EGFR and STAT3 -related lncRNAs might be involved in the pathogenesis of NFPA and can be suggested as potential biomarkers for this condition. Our study had some limitations. First, tissues that were considered as NTAT cannot be considered as a certainly "normal" pituitary tissue and might have aberrations in the tissue structure or microenvironment. Moreover, the in vivo effects of the observed differences in the expression of mentioned genes are not clear. Additionally, since most of samples were taken from a single center, samples do not accurately represent the total population of patients. Finally, since we did not use the most stable reference genes in the pituitary adenoma [ 27 ], our experimental data was not optimal.

Pituitary adenomas are generally benign neoplasms that comprise around 10–20% of intracranial tumors [ 1 ]. Non-functioning pituitary adenomas (NFPAs) are a group of these neoplasms originated from the adenohypophyse and do not show clinical evidence of hormonal oversecretion. Thus, NFPAs are usually diagnosed based on the “mass effects” symptoms [ 2 ]. According to the 2017 WHO classification [ 3 ], pituitary adenomas are categorized based on the secreted pituitary hormones and transcription factor profile. In fact, the new categorization is founded on IHC and adenohypophyseal cell lineage description of the pituitary adenomas [ 4 ]. The advent of transcriptome sequencing has changed the paradigm of cancer research. This technique has also been applied in pituitary adenoma revealing reveals the impact of copy number variations on expressions profile and patients' outcome [ 5 ]. Parallel with this high throughput technique, candidate gene expression analyses have shown dysregulation of a number of genes in different subtypes of pituitary adenomas [ 6 , 7 ]. Selection of candidate genes for expression analyses is an important step, which has been facilitated through bioinformatics analyses [ 8 ]. In our recently published article, we developed a pipeline for detection of pairs of long non-coding RNAs (lncRNAs)/target mRNAs with possible roles in the tumorigenesis of pituitary adenoma [ 8 ]. LncRNAs, which have > 200 nucleotides, contribute to several processes of gene regulation, including nuclear and cytoplasmic transferring, chromosomes dosage compensation, and transcript splicing and translation [ 9 ]. Previous reports have shown involvement of lncRNAs in the progression of pituitary adenomas [ 10 ] and their recurrence [ 11 ]. Therefore, discovering the expression patterns of lncRNAs in pituitary adenoma might present novel biomarkers for diagnostic approaches in these lesions. In our recent study, using bioinformatics and experimental assays, we showed that the HIF1 pathway is involved in the development and progression of pituitary gland tumors [ 8 ]. Also, using bioinformatics analysis, we determined the hub genes related to this signaling pathway and among them, we analyzed two hub genes named MAPK1 and STAT3 [ 8 ]. In this study, another hub gene involved in upstream of this pathway, called EGFR is investigated. In this study, in order to detect novel biomarkers, using the same bioinformatics approach mentioned in the previous study, a set of regulatory lncRNAs for the two other hub genes, i.e. STAT3 and EGFR , were selected and subjected to experimental investigation. These lncRNAs are EGFR-AS1 for the EGFR gene, and LINC00240 , FALEC and SNHG12 for the STAT3 gene.