Epigenetic Role of Noncoding RNAs in the Recurrence of Pituitary Adenoma after Surgical Resection

Background: Due to the lack of related symptoms and the hypersecretion of serum hormones, it is difficult to monitor and predict the postoperative recurrence of non-functioning pituitary adenomas (NFPAs). Long noncoding RNAs (lncRNAs) and protein-coding genes (PCGs) play critical roles in the development and progression of many tumors. However, the complex network of RNA interactions related to the mechanism and function of the postoperative recurrence of NFPA is still unclear. Methods: In the present study, 100 patients with NFPA were investigated by high-throughput sequencing and follow up. Among them, 16 NFPA patients experienced recurrence at different intervals of more than five years or less than one year. Results: By performing differential expression analysis of the fast recurrence and slow recurrence groups (t test P < 0.05), we obtained a set of differentially expressed PCGs and lncRNAs. We then identified protein-protein interaction (PPI) coregulatory networks and lncRNA-mRNA coexpression networks. In addition, we further screened the hub lncRNA-mRNA modules related to NFPA recurrence. These modules identified transcriptome expression markers for NFPA regression (log-rank test P <0.05). Finally, we evaluated the ability of the hub and module genes to predict recurrence and progression-free survival (PFS) in NFPA patients. To confirm the credibility of the bioinformatic analysis, NOL6 was randomly selected from the prognosis genes for validation by quantitative real-time polymerase chain reaction (qRT-PCR) in another set of NFPA samples (n=9). Conclusion: These results may be helpful for evaluating the slow and rapid recurrence of NFPA after surgery, may help us explore the mechanisms of NFPA recurrence and may also serve as future effective biomarkers and therapeutic targets.

that approximately 12-58% of NFPA patients with macroadenoma may experience regrowth within five years [10][11][12][13]. Radiotherapy is often recommended for patients with tumour residue, but its longterm complications, such as visual defects and hypopituitarism, are still of concern [14,15]. Therefore, surgery is still the best option for patients with tumour recurrence. Serum hormone monitoring is used as a detection approach in functional pituitary adenoma; however, the absence of an effective evaluation approach for NFPA results in the failure of early intervention. Research concerning the molecular mechanisms of tumour recurrence and effective prognosis prediction methods is of great significance. Increasing evidence shows that protein-coding genes (PCGs) are involved in the activation of pathways or key proteins and play vital roles in the biological processes of pituitary adenomas. Studies by Uraki S et al. [16] show that reducing the expression of MSH6 and MSH2 can directly promote the growth of pituitary tumour's through the ATR-Chk1 pathway [16]. Ruiqing Long et al. [17] suggested that COL6A6 interacted with P4HA3 to inhibit pituitary adenoma cell growth and invasion by inhibiting the PI3K-Akt pathway [17]. It has been reported that the low expression of TGF-β RII may be related to the development and invasion of NFPAs [18]. The research of Zhu H et al. [19] confirmed that the expression of TGF-β1 and WIF1 in recurrent tumour's is higher than that in primary tumour's, suggesting that these PCGs may be related to cell proliferation and recurrence [19]. Compared to noninvasive NFPAs, the expression levels of WIF1 and sFRP4 were reduced in invasive NFPAs, and WIF1 may be a potential biomarker for the aggressiveness of NFPAs [20].
Long noncoding RNAs (lncRNAs) play an important role in regulating gene expression through epigenetic or posttranscriptional mechanisms. However, lncRNAs cannot encode protein; they are a type of RNA molecule with a transcript longer than 200 nucleotides [21][22][23]. The differential expression and dysregulation of lncRNAs is believed to be involved in carcinogenesis and cancer progression, recurrence, and metastasis [21]. However, the role of lncRNAs in NFPA recurrence and the regulation of cellular processes remain unknown. Studies have shown that LINC00858 plays a tumourpromoting role in colon cancer by upregulating HNF4α and downregulating WNK2 [24]. Xu H et al. [25] showed that the over expressions of the lncRNA PAXIP1-AS1 can upregulate KIF14, thereby enhancing human umbilical vein endothelial cell migration, invasion, and angiogenesis in gliomas [25]. Moreover, many studies found that identifying novel lncRNA-mRNA networks by microarray analyses could contribute to exploring the potential molecular mechanisms and prognosis of tumour's [26][27][28]. The above studies indicated that the dysregulation of lncRNAs and lncRNA-mRNA interactions may affect the prognosis of NFPAs.
In this study, we obtained 299 differentially expressed PCGs (228 unregulated and 71 down regulated PCGs) and 214 differentially expressed lncRNAs (120 unregulated and 94 down regulated lncRNAs) by performing differential expression analysis of the fast recurrence and slow recurrence groups (P < 0.05). We also identified protein-protein interaction (PPI) networks and coregulatory networks between lncRNAs and mRNAs. We further screened the hub lncRNA-mRNA modules related to NFPA recurrence and assessed the enrichment of the differentially expressed genes (DEGs) in different pathways by gene set enrichment analysis (GSEA). In addition, we evaluated the ability of the hub and module genes (NOL6, CDK15, MOV10, SAMM50, COL24A1, EPHX1, and DCP1A) to predict recurrence and progression-free survival (PFS) in NFPA patients. These results may help us explore the mechanisms of NFPA recurrence and may also serve as future effective biomarkers and therapeutic targets.  Table 1. All tumour samples were immediately placed into a sample tube, frozen in liquid nitrogen and stored. Among them, 16 NFPA patients experienced recurrence at different intervals of more than five years (n=8) and less than one year (n=8). In addition, tissues from another 5 NFPAs that recurred within one year and 4 NFPAs that recurred after more than five years were collected as validated

Identification of Differentially Expressed lncRNAs and mRNAs
Differential gene expression analysis was performed within one year after the initial postoperative NFPA (n=6) and five years later (n=6), and significance analysis of microarrays (SAM) was performed to identify the differentially expressed PCGs and lncRNAs (DEGLs) between the two groups [29]. We first downloaded the Biobase, multtest and siggenes packages from Bioconductor (http://www.bioconductor.org/). Subsequently, the available data were analyzed by the R program (www.r-project.org), and DEGLs with fold changes of > 2 and <-2 and P values of <0.05 were selected for further research.

Construction of a PPI Network and lncRNA-mRNA Coexpression Network
Cytoscape software was used to construct, visualize, and analyze the PPI network [30].

Validation and Efficacy evaluation of the Hub Genes by Survival Analysis
Among the hub genes, genes of interest that have not been studied in NFPA were further validated in two groups (recurrence

American Journal of Biomedical Science & Research
Copy@ Young Zoon Kim in less than one year and recurrence after more than five years).
The PFS analysis of the hub genes and module genes was performed using Kaplan-Meier curves in the R program. A P value < 0.05 was considered statistically significant.

Validation of Gene Expression by Quantitative Real-Time Polymerase Chain Reaction (qRT-PCR)
We used qRT-PCR with another set of NFPA samples to verify the credibility of the bioinformatics analysis. Total RNA of validated samples was extracted and purified as described above. Reverse

Identification of Degls between the Fast Recurrence and Slow Recurrence Groups
Through microarray sequencing of 100 NFPA samples, we

Dysregulated lncRNA-mRNA Interaction Network Establishment and Module Analysis
Based on the Pearson test, we constructed a differential expression network of lncRNAs and mRNAs, selecting genes with P < 0.05 and Pearson coefficient absolute value > 0.9/< 0.9 (lncRNA/mRNA quantity = 78/104, Figure 3A), and transferred this network to the differential PCG PPI parent network (see Method).
Afterwards, we obtained the lncRNA-mRNA interaction network by combining these two networks. The lncRNA-mRNA network we constructed for the DEGLs contains a total of 4,490 nodes and 6,933 interactions. Figure 3B shows that the degrees of the nodes follow a power-law distribution, further illustrating that the network is similar to most biological networks, and the network is scale-free. We also calculated the average path length of the network, which shows that the characteristic path length of the network is much longer than the path length of the random network (1000 times longer than the random network, P < 0.001, Figure 3C), which implies that the network had reduced global efficiency.

Evaluation of the Hub and Module Genes for Predicting the Recurrence and PFS of Patients
Then, the predictive ability of the module genes for the recurrence process was evaluated. Kaplan-Meier analysis of the central or module genes NOL6, CDK15, MOV10, SAMM50, COL24A1, EPHX1, and DCP1A showed that the patients could be divided into a high-risk group (n = 37) and a low-risk group (n = 36) according to the median value of gene expression as the cutoff value. Compared with low-risk patients, the PFS time of high-risk patients was significantly shorter (log-rank test P < 0.05, Figure 6A-6G).    We performed qRT-PCR to confirm the reliability of the expression profiles generated using the microarray and DEGs analysis. From among the prognostic hub and module genes above mentioned, NOL6 was randomly selected for verification ( Figure   6H, P < 0.05). As expected, the qRT-PCR result basically matched the microarray analyses. These results indicate that the bioinformatics analysis of the microarray data reliably identified critical candidate genes involved in NFPA recurrence.  [6,36,37]. Transsphenoidal surgery is the recommended first-line treatment [38]. However, unlike functioning pituitary adenoma, it is difficult to monitor the tumour recurrence of NFPA through specific serum hormone alterations. When patients are re-examined because of optic nerve compression symptoms, the tumour may have grown into a large volume, which brings many obstacles to total resection and postoperative recovery. Therefore, we aimed to develop a new predictive signature that could identify early recurrence and be used as a prognostic prediction model. The main purpose of the study was to divide patients into high-risk or low-risk groups so that the most effective and timely treatment can be performed for NFPA patients.
Numerous studies have focused on the factors of tumour recurrence of NFPA to improve the prognosis of postoperative patients. Age is recognized as an important independent factor influencing the prognosis of NFPA patients, and a younger age indicates a greater chance of tumour recurrence [10,39]. However, the prognostic value of age is not as effective as the PCG and lncRNA signature in our study. Ki-67 is another commonly used pathological prognostic evaluation index [40], but a single indicator used in prognostic assessment has certain limitations in accurately evaluating the prognosis of each patient. A previous study tried to establish a statistical model that combines clinical features (age and tumour volume) and molecular markers (p16, WIF1 and TGF-β) to evaluate the recurrence probability of postoperative NFPA patients [41]. In our study, the inclusion of clinical features did not show a better efficacy. Moreover, compared with a previous study, we added a time concept to the prognostic assessment and independently assessed the prognosis of patients at different time points.
In recent years, lncRNAs have been reported in various tumour's, serving as promising new molecular markers for tumour biological behavior, tumour diagnosis and prognostic evaluation [42,43]. The lncRNA H19 was decreased in pituitary adenomas, and its overexpression could markedly inhibit the growth of pituitary tumour cells and be used as a drug resistance marker [44]. Xing et al. [45] identified mRNAs and lncRNAs differentially expressed in clinically NFPA and normal pituitary and constructed an mRNA-lncRNA coexpression network [45]. However, their research failed to illustrate the regulatory mechanisms of the key genes or lncRNAs and their influence on patient prognosis. In this study, we focused on identifying molecular markers of NFPA recurrence.
First, we obtained the DEGLs based on NFPA recurrence in less than one year and more than five years. According to GSEA, these DEGs were enriched in the regulation of cell death and cell adhesion. Our current results are consistent with those of previous studies, which have shown that intercellular adhesion and adhesion molecules are crucial steps for tumour recurrence and proliferation [46,47]. Then, we obtained 8 modules by cluster analysis using the PPI network based on the DEGLs. GO and KEGG enrichment analyses illustrated that these module genes were mainly involved in different GO functions and pathways. For Module 1, the related GO functions were T cell migration and chemotaxis, which implies that the process of recurrence may be correlated with the immunerelated tumour microenvironment. Similar to our study, Marques, P. et al. found that a low CD8:CD4 ratio is associated with a higher proliferative index (Ki-67) in pituitary adenoma [48]. In addition, KEGG analysis of other modules found that these genes were involved in the cell cycle, TNF signaling pathway, VEGF signaling pathway, TGF-beta signalling pathway and so on. These pathways might participate and regulate the proliferation and recurrence processes that occur in NFPAs.
Third, in our analysis, we obtained hub genes and module lncRNAs with significant differential expression (CCR1, CCL3, CCL4, binding protein that is abnormally expressed in a variety of cancers [49]. Liu X et al. [50] through meta-analysis, indicated that ANXA2 overexpression might be related to poor outcomes in patients with malignant tumors [50], which is consistent with our findings. In addition, we found lncRNAs that could be used as a prognostic signature. However, the functions and regulatory mechanisms of lncRNAs in NFPA have not yet been reported. Finally, we also assessed the predictive ability of the module genes (such as NOL6, CDK15, MOV10, SAMM50, COL24A1, EPHX1, and DCP1A) for the recurrence process. In addition, we validated the expression level of NOL6, which was randomly selected from among the hub and module genes, by qRT-PCR. The results confirmed the accuracy of our analysis.
NOL6 (nucleolar protein 6) encodes a nucleolar RNAassociated protein that is associated with the early stage of ribosome biosynthesis [51]. Dong D et al. [52] found that NOL6 is highly expressed in human prostate cancer, and knockdown of NOL6 inhibits the proliferation and mitosis and increases the cell apoptosis of human prostatic carcinoma cells (PC-3) [52]. Here, NOL6 was found to be upregulated in NFPAs recurrent within one year compared with those recurring after more than five years, suggesting that NOL6 could be a critical gene in prognostic  [53]. Nakano M et al. [54] found that the mRNA and protein levels of MOV10 in cancer cells were higher than those in normal cells [54]. In addition, MOV10 has been revealed to promote the angiogenesis of glioma by binding circ-DICER1 [55]. These studies indicated that MOV10 could be critical in tumorigenesis. DCP1A is a protein-coding gene for mRNA-decapping enzyme 1a, and several studies have revealed that DCP1A is upregulated in tumour tissues such as malignant melanoma, colorectal carcinoma and gastric cancer [56][57][58]. In addition, Tang, Y.'s study found that the high expression of DCP1A in colorectal carcinoma is correlated with poor prognosis [57], which is consistent with our results, indicating that the other PCGs and lncRNAs in our results could also be prognostic indicators for NFPA.
There are a few limitations of this study that need to be acknowledged. First, the molecular mechanisms of these PCGs and lncRNAs in NFPA are still unclear, and further studies might provide important information to understand their functional roles.
Second, there are few available sequencing data about NFPA, so we were unable to verify our results in an independent validation set.
Finally, the application of our signature in clinical practice should be tested prospectively. Despite these limitations of the current study, through our analysis, we verified a certain correlation between PCG and lncRNA signatures and regression. These results indicate that it is a potentially powerful prognostic marker of NFPA.

Conclusion
This is the first study to integrate PCGs and lncRNAs to predict tumor recurrence in patients with NFPA. Our study may provide a new aspect of prognostic evaluation and help patients benefit from early intervention.