Increased Expression of FN1 in Head and Neck Squamous Cell Carcinoma Predicts Poor Prognosis


 Background: A few recent studies have addressed the function of FN1 (Fibronectin 1) in head and neck cancer. The clinical information from 500 HNSCC (Head and neck squamous cell carcinoma) patients with FN1 gene expression data set was published by The Cancer Genome Atlas (TCGA). The correlation between clinicopathologic characteristics and FN1 expression was analyzed by Logistic regression and Wilcoxon signed rank test. Survival function was performed employing Kaplan-Meier estimator, and the relationship between clinicopathological characteristics, prognostic outcome, and FN1 expression were examined by using Cox regression analysis. As Gene set enrichment analysis (GSEA) was performed, we investigated the correlation between FN1 expression and immune cell infiltrates with single-sample gene set enrichment analysis (ssGSEA). Results: Patients with high FN1 expression revealed a significantly decreased overall survival (OS), and disease-specific survival (DSS) than those with low FN1 expression in Kaplan-Meier survival analyses. According to the above results, univariate and multivariate analysis revealed that patients with high FN1 expression had lower OS than those with low FN1 expression.Conclusions: The findings of this research provide insights for FN1 may be potential prognostic biomarkers for diagnosis as well as therapeutic targets in HNSCC patients.

metastasis. [10] FN1 could down-regulate P53 and inhibit cell growth and increase apoptosis in colon cancer. [11] Although FN1 overexpression is regarded as a poor prognosis factor in radiotherapy resistance in oral squamous cell carcinoma, the potential functions of the FN1 mechanism have not been de ned in HNSCC yet.
The goal of the current study was to use a comprehensive expression analysis of the FN1 and its relationship with the prognosis of patients with head and neck squamous cell carcinoma (HNSCC) by The Cancer Genome Atlas (TCGA), Kaplan-Meier plotter, and gene-set enrichment analysis (GSEA). Further, we estimated the correlation between FN1 and in ltration abundances of immune cells by single-sample Gene Set Enrichment (ssGSEA) analysis. In addition, the limma package administers the removeBatchEffect function, which performs batch effect correction on the TCGA dataset [12]. Our results purport the prognostic value of FN1 in HNSCC and explore the underlying mechanism of FN1 in regulating HNSCC patients.

Characteristics of patients
Gene expression data and clinical data of 500 cases of head and neck cancers were collected from the TCGA database in August 2021 (Table 1).
In this study, under the current eighth edition of the TNM staging system, the total size of most patients was stage IV (272 cases; 55.7%), followed by stage III (102 cases; 20.9%), and the least patients were stage II ( 95 cases; 19.5%) and stage I disease (19 cases; 3.9%). The primary therapy outcomes included complete response (CR) 87.3%, partial response (PR) 1.4%, stable disease (SD) 1.4%, and disease progression (PD) 9.9%. The median follow-up time of overall survival (OS) in the FN1 high expression group and FN1 low expression group at last contact were 57.3 months and 40.1 months, respectively.
Among 441 patients who were assessed for lymphovascular invasion (LI), 122 (35.8%) patients had LIpositive, while 219 (64.2%) patients had LI-negative. In this study, there were 244 patients under 60 years old, accounting for 48.9% of the total number of patients, and 255 patients 60 years or older, accounting for 51.1%.

Association with FN1 gene expression and clinical pathological Variable
To better determine the relevance and clinical signi cance of FN1 expression in HNSCC, the association of FN1 expression with clinical characteristics was investigated of 500 HNSCC samples. Our result are plotted in Figure 1 and showed increased FN1 expression was enriched in advanced stage III and IV (p = 0.979), objective clinical response (CR and PR, p = 0.415), high histological grade G3 and G4 (p = 0.027), and lymphovascular invasion-positive (p = 0.114).
To identify which poor prognostic clinicopathologic and characteristics were independently associated with FN1 expression, we performed a univariate analysis. Speci cally, poor clinicopathological did not associate with co-pathology in any group via logistic regression analysis (Table 2).

Survival Outcomes and Multivariate Analysis
Kaplan-Meier survival plots showed a signi cant correlation between high expression FN1 levels and revealing a poor prognostic value for FN1. In univariable analysis (Figure 2), low FN1 expression was associated with a better OS, and DSS (p = 0.046, 0.041, respectively).
When other clinicopathologic features that were signi cant with poor OS in primary therapy outcome, lymphovascular invasion and metastasis stage in univariate analysis were adopted as covariates (Table   3). Multivariate analysis revealed that primary therapy outcome was an prognostic factor for OS (HR = 0.161; p < 0.001) but not for lymphovascular invasion and metastasis stage (HR = 1.446, 2.107 respectively; p = 0.065, 0.337, respectively). FN1 was still independently associated with the OS.
When other clinicopathologic features that were signi cant with poor DSS in primary therapy outcome, degree of spread to regional lymph nodes, metastasis stage, and lymphovascular invasion in univariate analysis were adopted as covariates (Table 4). Multivariate analysis revealed that primary therapy outcome was an prognostic factor for DSS (HR = 0.091; p < 0.001) but not for degree of spread to regional lymph nodes, metastasis stage, and lymphovascular invasion (HR =1.386, 2.374, 1.292 ; p = 0.136, 0.243, 0.310, respectively ). Associations with DSS were independent of risk factors and persisted even after additional adjustment.

GSEA Related Signaling Pathway
To explore FN1-related catalogue signaling pathways in HNSCC, we utilized Gene Set Enrichment Analysis (GSEA) between low-and high-expression FN1. We used the median score as a cut-off point to divide HNSCC samples into either high-expression or low-expression FN1. GSEA demonstrated signi cant enrichment differences (false discovery rate (FDR)-adjusted pooled p-values < 0.05) of the MSigDB molecular signatures databases Collection (c2.cp.v7.2.symbols.gmt). To investigate the function of the target gene, we then selected the most enriching signaling pathways via their normalized enrichment score (NES) value (Table 5 and Figure 3). The high expression of FN1 revealed the differential enrichment of amino acid and its derivatives metabolism and olfactory signaling pathway categories.

Statistical Correlations between the Expression of FN1 and Immune Cells
To identify immune cell-types and genes related to the anti-tumour function that might be a key factor to immunotherapy, we focused on the correlations between FN1 and immune cells of HNSCC in the databases of Bindea et al to clarify their relation to each other [16]. Finally, this analysis demonstrated FN1 expression was correlated with 24 immune cell subsets. For example, T helper (Th)1, Th2, Th17, effector memory (Tem), γδ, central memory, regulatory, and cytotoxic T cells, three dendritic cells types (immature, activated, and plasmacytoid), two subtypes of natural killer cells, as well as B cells, neutrophils, and mast cells ( Figure 4). The nding showed a signi cant positive association between FN1 and macrophages.
Signi cant correlations and moderately positive were found between FN1 and NK cells, Eosinophils, immature dendritic cells, and Tem cells. An insigni cant relationship between FN1 and the other helper T cell-types was shown in Among them, the extra domain A (EDA) and extra domain B (EDB) are expressed well during chronic in ammation and cancer. Thus, it is a potential diagnosis and treatment target. [20][21][22] One interesting nding is that EDB overexpression appears in various human cancers, including Hodgkin's lymphoma, non-small cell lung cancer, and prostate cancer. [23][24][25] Very little was found in the literature on the association between Fibronectin and head and neck cancer (HNC). Herein, we validated the highthroughput RNA sequencing data downloaded from TCGA showing the increase of FN1 expression in head and neck squamous cell carcinoma (HNSCC) is associated with poor prognosis, short survival time, and poor clinicopathological characteristics. Besides, our analysis revealed that the expression level of FN1 in HNSCC is related to in ammatory in ltrates in diverse immune cell types. Thus, this study con rms, and supports evidence from our understanding of the potential target of FN1 in HNSCC immunity and as a diagnostic marker of this malignancy or potential prognostic. This study explored the expression level of FN1 and visualizes the prognostic outcome in head and neck cancer (HNSCC) using TCGA datasets. Fibronectin is one of the de ning features of EMT, it alters the composition of the ECM and is believed to contribute to the invasive properties of cancers. Indeed, previous studies provided evidence that bronectin has malignant effects and is carcinogenic in several malignant neoplasms. [23,26,27] As mentioned in the literature review, it can be hypothesized that FN1 is an oncogenic factor for developing squamous cell carcinoma, but there is no su cient research on the role of FN1 in head and neck squamous cell carcinoma after reviewing the literature.
Based on the TCGA cohort, FN1-related gene expression was negatively correlated with OS and DSS, and there was signi cant statistical interaction between the parameters. Univariate and multivariate Cox regression analysis indicated that high FN1 expression was associated with poor prognosis of HNSCC.
Moreover, when FN1 was highly expressed in HNSCC, the current study have found that higher expression levels of FN1 were strongly correlated with worse primary outcome and lymphovascular invasion for poorer OS and DSS.
Increasing evidence demonstrated that HNSCC was considered as an immunogenic tumor. replicates which contributed to fewer data integration. In order to solve the problem, ComBat-and limma-corrected methods have been performed. Though we used the removeBatchEffect function (limma package in R) to minimize batch effects, the sequencing may still cause some batch effects.

Conclusion
This study found that increased FN1 expression indicated poor long-term prognosis of HNSCC with increased abundance of macrophages and immune cells subsets proportion. In addition, the expression of FN1 in HNSCC regulated memory T cells, NK cells, and neutrophils, by activating different signal cascades to activate the level of TGF-β1 receptor associated with the growth of HSNCC. Therefore, FN1 may be a prognosis predictive biomarker, and plays a signi cant role in immune cell in ltration for HNSCC patients.

Materials And Methods
Data from TCGA Analysis of mRNA SeqV2 expression in head and neck squamous cancer were pro ling from The Cancer Genome Atlas (TCGA) database. ( https://tcga-data.nci.nih.gov/tcga/ ) [13]. Gene expression pro ling of FN1 was obtained from TCGA RNA-seq data of 500 Head and neck squamous cell carcinoma (HNSCC) patients. Clinical performance of overall survival (OS) and disease-speci c survival (DSS) endpoints were derived from the TCGA pan-cancer Clinical Data Resource (CDR) [14]. HTSeq-FPKM was used to calculate fragments per kilobase of transcript per million mapped reads, normalized gene expression. We analyzed the clinical data of level 3 of expression level in HNSCC patients simultaneously. According to the CDR, the clinical endpoints used for HNSCC were selected as OS and DSS. Patients who were off study or died were counted as nonresponses in this measurement.

Gene enrichment analysis
Gene set enrichment analysis (GSEA) was used to determine gene sets signi cance and concordant differences between two ranked gene states. (http://software.broadinstitute.org/gsea/index.jsp) [15] In this analysis, GSEA was the rst to the rank-ordered list based on the Spearmen's correlation with FN1 expression. GSEA was used to determine signi cance of survival differences between high and low FN1 The collected data had analyzed the relationship between clinicopathological features and FN1 by logistic regression test and Wilcoxon sign-rank test. Uni-and multivariate binary regression analysis were performed to evaluate FN1 expression scores. Survival curves were constructed using Kaplan-Meier method, and differences between survival curves were analyzed using the log-rank test. Hazard ratios (HRs) of the operating system were estimated by univariate Cox proportional hazards regression models.
The p values were two-sided and p values less than 0.05 indicating signi cance.  Tables Table 1 The Cancer Genome Atlas (TCGA) head and neck squamous cell carcinoma (HNSCC) patient characteristics.