Integrative Analysis of Pinin as A Prognostic Factor in Digestive Tract Cancers

Background: Pinin (PNN) was originally identied for acting an essential role in epithelial cell-cell adhesion. We aim to illuminate the expression prole, mutation feature, methylation status of PNN and its prognostic value in digestive tract cancers. Methods: Expression and methylation data of PNN, as well as clinical information on esophagus cancer (ESCA), gastric adenocarcinoma (STAD), colon adenocarcinoma (COAD) and rectal adenocarcinoma (READ) were acquired from The Cancer Genome Atlas database. The value of PNN expression, mutation feature and methylation status in prognosis were assessed. GO enrichment and Gene set enrichment analysis were performed to investigate the enriched biological functions and pathways of PNN in COAD. Tumor Immune Estimation Resource and CIBERSORT were applied for evaluating the effects of PNN on tumor immune inltrating cells. Results: PNN was signicantly overexpressed in digestive tract cancers and was remarkably related to tumor stage. Highly expressed of PNN was positively related to poor free survival (PFS) and overall survival (OS) in COAD. Additionally, 10 CpG sites methylation in PNN had signicant effects on survival. PNN expression was conrmed as an independent prognostic factor for predicting the OS in COAD. GO and GSEA analyses revealed that PNN participates in multiple biological processes underlying carcinogenicity in COAD. PNN was signicantly associated with TIICs. Moreover, a promising prognostic nomogram incorporating the PNN expression and clinicopathological characteristics was established for predicting OS probability in COAD. Conclusions: Our comprehensive bioinformatics study demonstrated that PNN was highly expressed in digestive tract cancers, which could act as an independent prognostic factor for COAD.

unfavorable disease-free survival (DFS). The authors concluded that PNN can act as a predictive biomarker in those patients. Therefore, the PNN expression and its physiological function in cancers are still inconsistent.
Digestive tract cancer is a common malignancy worldwide and accounts for a large proportion of cancer occurrence and death (6). Although some helpful diagnostic biomarkers have been identi ed, more reliable and effective biomarkers are expected in cancer management. However, few studies have systematically explored the PNN pro le in digestive tract cancers up to now. The present study aims to illuminate the expression pro le, mutation feature and methylation status of PNN, as well as its prognostic value in four digestive tract cancers, including esophagus cancer (ESCA), gastric adenocarcinoma (STAD), colon adenocarcinoma (COAD) and rectal adenocarcinoma (READ) by analyzing The Cancer Genome Atlas (TCGA) and several acknowledged open-access bioinformatics databases.

Expression features of PNN in digestive tract cancers
The target data included in the gene transcripts acquired in this study were as follows: 162 ESCA and 11 adjacent normal samples, 375 STAD and 32 adjacent normal samples, 480 COAD and 41 adjacent normal samples, 167 READ and 10 adjacent normal samples. The PNN mRNA expression were observed signi cantly upregulated in four types of digestive tract cancers, including ESCA, STAD, COAD and READ, compared with corresponding normal tissues ( Figure 1A). Correlations between PNN mRNA expression and tumor clinical stages in ESCA, STAD, COAD, READ were also analyzed respectively. In general, the expression of PNN were positively associated with tumor stages in digestive tract cancers ( Figure 1B). Patients with advanced clinical stage tended to have higher expression of PNN and the highest expression was observed in stage III. However, we noticed that the expression of PNN in stage IV were lower than in stage III, it may be due to the small sample number in stage IV.

Relationships between DNA methylation and PNN expression
Studies have suggested that gene promoter region methylation is a signi cant cause affecting gene expression, contributing to the progression of human cancer. Current research assessed the effect of promoter DNA methylation on PNN expression in four types of digestive tract cancers using Pearson correlation. We found a signi cant negative correlation between PNN expression and promoter region methylation level, especially in STAD ( Figure A). It indicated that in STAD, abnormal methylation of promoter region is one of the important causes of PNN gene overexpression, but in other digestive tract cancers, there may exist other more regulatory mechanisms in uencing the expression of PNN.
The prognostic value of PNN in digestive tract cancers Kaplan-Meier analyses indicated that PNN played different prognostic role in different types of digestive tract cancers (Figure 3). In ESCA, patients with higher PNN expression level associated with poorer OS (p=0.076) and PFS (p=0.044). In STAD, patients with high PNN expression had a better prognosis, but unfortunately, there was no statistically difference observed. In COAD, higher expression of PNN was signi cantly associated with poorer OS (p=0.003) and poorer PFS (p=0.009). And in patients with READ, the higher expression level of PNN was correlated with longer OS (p=0.02), but there was no statistical difference in PFS (p =0.082).
We further explored the genetic alterations in PNN and their effects on prognosis in patients with digestive tract cancers by utilizing cBioPortal database. The proportion of various genetic alterations of PNN in different digestive tract cancers were similar: 6% in ESCA, 8% in STAD and 6% in CRC ( Figure 4). It was worth noting that the types of genetic alteration were diverse. Ampli cation and deep deletion were more common in ESCA, missense mutation and truncating mutation were more frequent in STAD and CRC. It was with regret that there was no signi cant correlation between PNN genetic alterations and OS in these cancers.
Moreover, we investigated the DNA methylation of PNN CpG sites and corresponding prognostic effects in four digestive tract cancers by the MethSurv database. The results were illustrated in Table 1. We observed that cg18648343, cg12087797, cg15592059, cg20337385 were remarkably associated with the prognosis in patients with STAD. In COAD, cg15592059, cg24034629 and cg10250651 were indicated as signi cant factors for prognosis. As for READ, the meaningful cg sites in prognosis included cg02969452, cg18648343 and cg12087797. However, there were no statistically signi cant DNA methylation CpG sites observed for predicting OS in ESCA.
Finally, considering the signi cantly prognostic effect of PNN expression in COAD, we performed univariate and multivariate cox regression analyses to assess the independent prognostic effect in COAD. After controlling the clinical parameters, the univariate cox regression analysis suggested that the expression of PNN had signi cant effect both in survival (p=0.004) and recurrence (p=0.016) ( Table 2, 3). Subsequently, the expression of PNN was identi ed as an independent prognostic biomarker for predicting the OS in patients with COAD with multivariate cox regression analysis (p = 0.041, HR=1.7, 95% CI 1.02-2.8, Figure 5). Table 1 The prognostic effect of CpGs in PNN

Functional enrichment analysis of PNN expression in COAD
Owing to the independent prognostic value of PNN expression in recurrence and survival, we explored the biological functions of PNN in COAD by GO analysis based on Metascape. In this research, GO pathway and process enrichment analysis included: molecular functions (MFs, functional set), biological processes (BPs, pathway) and cellular components (CCs, structural complex). Top 15 clusters were displayed in Figure 6A (regulation of small GTPase mediated signal transduction) and GO 0061136 (regulation of proteasomal protein catabolic process).
The GSEA was performed for evaluating the underlying signaling pathways, which were involved in carcinogenesis of PNN in COAD. The study indicated that PNN high expression was positively associated with "spliceosome", "basal transcription factors", "WNT signaling pathway", "ERBB signaling pathway", "mTOR signaling pathway" and "adherens junction" ( Figure 6B).

Associations between TIICs and PNN in COAD
In the last few years, increasing researches have revealed the crucial relationships between the immune microenvironment and cancer progression. In the current study, we investigated the effects of different PNN expression levels on tumor-in ltrating immune cells in COAD by using CIBERSORT algorithm. The results illustrated that the distribution of 22 immune cell types in each COAD sample varied markedly ( Figure 7A). Furthermore, we observed that CD8+ T cells, regulatory T cells (Tregs) and resting dendritic cells were signi cantly enriched in low PNN expression group, nevertheless, resting NK cells and naive CD4+ T cells were markedly enriched in high PNN expression group ( Figure 7B). Moreover, TIMER database analysis demonstrated that high expression of PNN was positively associated with several types of TIICs, including B cells, CD8+ T cells, CD4+ T cells, macrophages, neutrophils and dendritic cells ( Figure 7C).

A prognostic nomogram for patients with COAD
A prognostic nomogram incorporating the PNN expression and common clinicopathological characteristics was successfully established for predicting the 3, 5, 10-year overall survival probability, which might be promisingly applied in clinical evaluation of patients with COAD ( Figure 8).

Discussion
PNN was initially reported as a novel factor involved in the mature desmosomes of the epithelia cell(1). Studies have revealed that PNN participated in apoptosis, proliferation, and migration regulation by affecting mRNA splicing and gene transcription (7  (3). PNN could promote cell proliferation and metastasis via activating egfr/erk signaling pathway in colorectal cancer. The upregulated PNN was also con rmed as an independent adverse prognostic factor in hepatocellular carcinoma patients (2). In ovarian cancer, PNN high expression was observed to increase cell adhesion and clonogenicity of epithelial cell (4). Besides, Tang et al. found increased expression of PNN promoted migration and invasion of nasopharyngeal cancer cells in vitro, as well as metastasis in vivo (9). In addition, recently, a study demonstrated that patients with stage III colorectal cancer would not bene t from the adjuvant uoropyrimidine-based chemotherapy who expressing high mRNA levels of PNN (5). In the current study, we found that PNN high expression had signi cantly poor OS and DFS in colon cancer. Further analysis con rmed that PNN expression was an independent prognostic factor for predicting the OS in colon cancer. Also, we found high expression of PNN was signi cantly related with poorer PFS in esophageal cancer. However, on the contrary, the upregulated PNN expression was markedly related with longer OS, and tend to have longer PFS in rectal cancer. But beyond that, our study rejected the independent prognostic effect of PNN in esophageal, gastric, and rectal cancer. Since previous studies has not subdivided colorectal cancer, the controversial results may contribute by heterogeneity of tumor site, or insu cient sample numbers of rectal cancer in TCGA.
It was commonly known that the prognosis and drug response of colorectal cancer patients were closely associated with speci c gene mutation status, such as KRAS, BRAF and PIK3CA (14,15 (2). And recently, studies showed that PNN participated in histone modi cation and pre-mRNA splicing events (18,19). In our study, the results of functional enrichment analysis demonstrated that PNN was involved in "spliceosome", "adherens junction and mRNA processing". The KEGG analysis showed PNN affect cell-cell adhesion and tumor invasion and metastasis via variety signaling pathways (e.g., WNT signaling pathway, ErbB signaling pathway, mTOR signaling pathway). WNT signaling pathway is known as one of the most important signaling pathways, its activation is very common during development of many tumors by facilitating cell differentiation, polarization and migration (20). The ErbB belongs to the receptor tyrosine kinase receptor family, includes four distinct members: EGFR (also known as ErbB-1/HER1), ErbB-2 (HER2), ErbB-3 (HER3) and ErbB-4 (HER4). The ErbB pathway is one of the most extensively studied areas of signal transduction, which best exempli es the pathogenic power of aberrations in biological information transfer (21,22). The mTOR signaling pathway is frequently activated in cancer, regulates cell growth and various cellular metabolic processes (23).
Until now, few studies have investigated the effect of PNN expression on tumor-in ltrating immune cells.
A study reported that PNN was markedly related to T cell receptor signaling pathway in renal cell carcinoma, and has a positive correlation with TIICs (24). At present study, we found that overexpressed PNN was signi cantly associated with resting NK cells and naive CD4 + T cells in COAD, conversely, PNN low expression were more enriched with CD8 + T cells, regulatory T cells (Tregs) and resting dendritic cells. Therefore, we considered that regulating the tumor immune microenvironment may be a crucial underlying mechanism by which PNN promotes tumor progression in COAD, which deserves further study.
Since our study validated a powerful e ciency of prognostic value of PNN in predicting the OS in patients with COAD, a promising prognostic nomogram incorporating the PNN expression and common clinicopathological characteristics was successfully established for predicting 3, 5, 10-year overall survival probability, which might be well applied in clinical evaluation.
However, there were still several limitations in this study. First, we only got the results through bioinformatics and database analysis, further experimental veri cation is required. Second, the limited sample size of sub-group may affect the result. Finally, the prognostic nomogram for patients with COAD needs more clinical veri cation. However, our study has its convincing power for its larger sample-based study by TCGA database.

Conclusions
In conclusion, our bioinformatics analyses demonstrated that PNN was highly expressed in digestive tract cancers, which could act an independent prognostic factor for COAD. Furthermore, the potential molecular mechanisms by which PNN participated in carcinogenicity and its correlations with TIICs need to be investigated further.

Data resource
Data Thereafter, we analyzed the correlations of PNN mRNA expression and clinicopathologic characteristics with OS and PFS among ESCA, STAD, COAD and READ patients respectively using univariate Cox regression analyses. Furthermore, we conducted multivariate Cox regression analyses to assess whether PNN expression was an independent factor in four digestive tract cancers. All analyses were conducted by survival and survminer packages of R software, and the forest plot was drawed by the ggplot package.
CpG sites methylation of PNN and its prognostic effect in digestive tract cancers We evaluated the DNA methylation of PNN CpG sites and its prognostic value for OS in ESCA, STAD, COAD and READ by the excellent online MethSurv database. The MethSurv is an online bioinformatics platform for multivariable prognosis assessment according to the massive DNA methylation data PNN miRNA expression and Tumor Immune In ltrating Cells (TIICs) in COAD According to the optimal cutoff value of PNN mRNA expression in the survival analysis based on the Kaplan-Meier method, the COAD samples were divided into two groups, high and low expression.
CIBERSORT was applied to estimate the relationships between PNN and the subsets of each immune cell types in COAD. The result was visualized by the vioplot package of R software. CIBERSORT is an online platform to estimate the proportions of 22 tumor-in ltrating lymphocyte subsets in the tumor microenvironment (https://cibersort.stanford.edu/) (29). Furthermore, we utilized the Tumor Immune Estimation Resource (TIMER) to evaluate the effect of PNN expression on six major TIICs (CD4+ T-cells,
A prognostic nomogram for patients with COAD For better clinical application of PNN expression, we further developed a nomogram incorporating the PNN expression and clinicopathological characteristics to estimate the survival probability of COAD patients by rms package of R software. The 3-, 5-, and 10-year OS rates served as endpoints in the nomogram. The clinicopathological characteristics involved in the nomogram including age, gender, stage, T stage and N stage.

Statistical analysis
Perl software 5.32 was used to extract and structure the HTSeq FPKM data, DNA methylation data and GESA preparation documents. R 4.0.3 software with speci c packages was used to perform analyses for gene differential expression, Pearson correlation, prognostic value evaluation and nomogram development. The comparisons of intergroup variables were conducted by using the chi-square test with SPSS software 20.0 (IBM, Chicago, USA). P<0.05 was considered to be statistically signi cant.

Ethical approval statement
Since there was no personal identifying information was used in the current study, it was granted an exemption from ethics approval from the Institutional Review Board of the A liated Lihuili Hospital, Ningbo University.

Consent for publication
Not applicable.

Data availability statement
The data used for bioinformatics analyses in this study are freely available on The Cancer Genome Atlas