Integrative Analysis of MUC4 to Prognosis and Immune Infiltration in Pan-Cancer: Friend or Foe?

MUC4, a transmembrane mucin, plays important roles in epithelial renewal and differentiation. Recent studies suggest that MUC4 has been implicated in pancreatic cancer pathogenesis and is expressed in various normal and cancer tissues. The underlying features of MUC4 across various cancer types may allow us to ensure appropriate treatment and patient monitoring. However, the contributions of MUC4 to pan-cancer have not been well characterized. In this study, we investigated the expression pattern and prognostic value of MUC4 across multiple databases. We further explored genomic and epigenetic alterations of MUC4, its association with proliferation and metastasis, and the correlation with immune infiltration in different cancers. Our results characterized the distinct expression profile and prognostic values of MUC4 in pan-cancer. Through examining its association with genomic alteration, tumor proliferation, and metastasis, as well as tumor infiltration, we revealed multiple function effects of MUC4. MUC4 may influence prognosis, proliferation, metastasis, and immune response in opposite directions. In conclusion, our findings suggested the necessity to more carefully evaluate MUC4 as a biomarker and therapeutic target and develop the new antibodies for cancer detection and intervention.


INTRODUCTION
In general, mucins are a family of glycosylated proteins, which are expressed by epithelial cells and provide protection and lubrication to epithelial surfaces (Kufe, 2009). However, aberrant expression of mucins occurs in various cancers and has been implicated in cancer progression and prognosis (Bhatia et al., 2019).
MUC4, a transmembrane mucin, is localized on chromosome band 3q29. Human MUC4 was first identified in 1991 from a tracheal library (Carraway et al., 2009). MUC4 is synthesized as two subunits: MUC4α and MUC4β. MUC4α contains a tandem-repeat domain altering glycosylation and epitope multiplicity, a nidogen-like domain, and an adhesion-associated domain. MUC4β consists of a von Willebrand factor-type D domain and three epidermal growth factor (EGF)-like domains. On the basis of the specific structure, MUC4 was suggested to modulate HER2/ERBB2 signaling and play a critical role in cancer.
Normally, MUC4 is expressed in the salivary glands, trachea and bronchioles, reproductive tract, colon, and mammary epithelium. In the past few years, many studies have reported that MUC4 is aberrantly produced in a variety of cancers, including lung, breast, pancreatic, prostate, ovarian, and bladder, and functionally links to tumor initiation, metastasis, and interaction of tumor cells with the components of the tumor microenvironment. The available evidence indicates that MUC4 is overexpressed in pancreatic cancer and contributes to the aggressiveness and metastasis of pancreatic cancer (Gautam et al., 2020;Sagar et al., 2021). MUC4 is also overexpressed in ovarian cancer and promotes the pathobiology and aggressiveness of ovarian cancer cells (Ponnusamy et al., 2011;Bae et al., 2017). MUC4 also plays a pivotal role in intestinal cell proliferation during tumorigenesis (Das et al., 2016). Rowson-Hodel et al. have provided considerable evidence that aberrantly expressed MUC4 can lead to the metastatic efficiency of breast cancer (Rowson-Hodel et al., 2018). In contrast, MUC4 expression was associated with improved survival and decreased recurrence in squamous cell carcinoma of the upper aerodigestive tract (Weed et al., 2004). These results suggested that the role of MUC4 appears to be complicated depending on the particular cancer and cell context. Thus, the expression and function of MUC4 in human tumors remain unclear and need to be analyzed in detail.

Differential Expression Analysis
We compared mRNA expression levels of MUC4 between normal and tumor tissues of each cancer type using Oncomine and GEPIA2. In Oncomine, a t-test was used to calculate the p-value, and the threshold was set as a p-value of 0.0001 and a fold change of 2. In GEPIA2, we used analysis of variance (ANOVA) for differential expression analysis and considered genes as differentially expressed genes with a fold change >2 and an FDR < 0.01. We also used GEPIA2 to assess the differential expression between different stages across independent cancer types by t-test and defined p-value <0.05 as significant.

Survival Analysis
We performed overall survival analysis based on MUC4 expression using PrognoScan and Kaplan-Meier Plotter. PrognoScan employed a univariate Cox regression model to find the optimal cut point in continuous gene expression measurement without prior biological knowledge or assumption and calculate the minimum p-value and hazard ratios with 95% confidence intervals for grouping patients. The Kaplan-Meier plotter split all patients into high-or low-expression groups according to the median value of MUC4 and used the log-rank test for hypothesis testing. We also used cBioPortal to evaluate the prognosis of MUC4 alteration. p-value <0.05 was considered significant.

Characterize the Alteration and Methylation Profile in Pan-Cancer
We explored the genomic alteration frequency including nonsynonymous mutation and copy number alteration of MUC4 in the TCGA PanCancer Atlas dataset via cBioPortal. We obtained the promoter DNA methylation level of normal and tumor samples from UALCAN. Significance of difference was estimated by t-test. A p-value under 0.05 was considered statistically significant.

Construct a PPI Network and Estimate the Associations Between MUC4 and Tumor Proliferation and Metastasis
We carried out PPI analysis with STRING, and the max number of interactions to show was 10. The proliferation marker ki67 was used to reflect tumor proliferation across tumor samples. We used CVCDAP to evaluate the correlation between MUC4 and MKI67 by Pearson's correlation and considered |R| > 0.3 and p-value <0.05 to indicate significance. We also used CVCDAP to divide tumor samples into a high and low group by the median expression value of MUC4 for each cancer type and calculate the epithelial-mesenchymal transition (EMT) enrichment score by GSEA. The gene set of EMT was collected from MSigDB. FDR < 0.25 was set as the threshold for screening.

Correlation Between MUC4 and Immune Infiltration
We employed TIMER2.0 for correlation between MUC4 and tumor immune infiltration. Via TIMER2.0, we analyzed the correlation between MUC4 and six immune cells, including CD8 + and CD4 + T cells, B cells, macrophages, neutrophils, and dendritic cells, in KIRC and PAAD. We also explored the relationships between MUC4 and immune gene markers in KIRC and PAAD. The association was generated with tumor purification adjusted. We used TISIDB to calculate the correlation between expression of MUC4 and abundance of immunomodulators and chemokines. The correlation was statistically assessed by Spearman's correlation. p-value <0.05 was considered significant.

Differential Expression of MUC4
To understand the differences in MUC4 expression between human cancer and normal tissues, MUC4 expression was explored via Oncomine. Our results revealed that MUC4 expression is upregulated or downregulated in different types of cancer ( Figure 1A). Compared to normal tissues, the expression of MUC4 was significantly higher in bladder cancer, cervical cancer, lung cancer, and pancreatic cancer. In contrast, the expression of MUC4 was lower in colorectal cancer, head and neck cancer, prostate cancer. Of interest, MUC4 was upregulated in two datasets while it was downregulated in four datasets in kidney cancer. Similarly, MUC4 was upregulated in one dataset while it was downregulated in one dataset in sarcoma.
We further confirmed the differential gene expression between tumor samples and normal tissues using TCGA datasets in GEPIA2 ( Figure 1B). Compared to normal tissues, the expression of MUC4 was significantly higher in CESC, LUAD, PAAD, and STAD, while the expression of MUC4 was lower in HNSC and PRAD.

Prognostic Analysis of MUC4
In cancer research, the relevance of MUC4 to clinical outcome may suggest the potential pathogenesis of disease and stimulate further researches. The impact of MUC4 on overall survival was evaluated through PrognoScan (Figure 2A). The results showed that high expression of MUC4 was associated with a poor prognosis in ovarian cancer (HR = 1.14, p = 0.0335), brain cancer (HR = 1.42, p = 0.0312), and lung cancer (HR = 1.25, p = 0.0335). However, in breast cancer (HR = 0.83, p = 0.0014) and colorectal cancer (HR = 0.02, p = 0.0456), increased expression of MUC4 was significantly correlated with good survival.
In addition, we examined the potential effects of MUC4 on prognosis across different cancer types via the Kaplan-Meier plotter ( Figure 2B). Our study revealed that the poor prognosis of ovarian cancer (HR = 1.42, 95% CI: 1.02-1.72, p = 0.0312), pancreatic ductal adenocarcinoma (HR = 1.55, 95% CI: 1.02-2.36, p = 0.039), thymoma (HR = 4.58, 95% CI: 0.93-22.53, p = 0.041), and bladder cancer (HR = 1.36, 95% CI: 1.01-1.83, p = 0.04) was related to the high expression of MUC4. Furthermore, the increased expression of MUC4 was related with prolonged overall survival in kidney renal clear cell carcinoma (HR = 0.67, 95% CI: 0.5-0.91, p = 0.0099). These findings revealed that the expression of MUC4 has important significance in the prognosis in pan-cancer and can be used as a prognostic factor. To further understand the relevance of MUC4 expression in cancer, we used the TCGA database to study the relationship between MUC4 expression and pathological stage via GEPIA2 ( Figure 2C). The MUC4 expression profile observed in KIRC and TGCT may suggest a link between the level of MUC4 and the tumor stage.
Cell proliferation is one of the significant hallmarks of cancer. To characterize the functional roles of MUC4 in cell proliferation, we calculated the Rs between MUC4 and the well-known proliferation marker ki67 across cancer types. We identified a total of three significant associations ( Figure 4B). MUC4 negatively correlated with cell proliferation in DLBC (R = −0.47, p = 0.00086) and positively correlated with cell proliferation in HMSC (R = 0.31, p = 3e-13) and PAAD (R = 0.49, p = 2.2e-12).
Metastasis is the major cause of death among cancer patients. Recent studies have heralded that EMT plays a critical role in metastasis (Heerboth et al., 2015;Mittal, 2018). To investigate the functional roles of MUC4 in metastasis, we assessed their enrichment associated with EMT through GSEA. We identified seven significant MUC4 enrichments (FDR < 0.25) ( Figure 4C).

Correlation Between MUC4 and Immune Infiltration Level
Given the distinctive roles of MUC4 in immunomodulation during cancer progression and metastasis Peng et al., 2021), we used TIMER2.0 to investigate the impact of the expression of MUC4 on tumor immune infiltration levels. The detailed results are shown in Figure 5. According to the TIMER2.0 results, we identified MUC4 expression has weak positive relevance with tumor purity and the immune-infiltrating levels of CD4 + T cells (R = 0.147, p = 1.55e-03) and B cells (R = 0.119, p = 1.07e-02) but a negative correlation with dendritic cells (R = −0.127, p = 6.19e-03) in KIRC. In PAAD, MUC4 expression has significant positive correlations with the immuneinfiltrating levels of B cells (R = 0.27, p = 3.6e-04) but no significant correlation with the infiltrating levels of CD8 + T cells, CD4 + T cells, macrophages, neutrophils, and dendritic cells.

DISCUSSION
Past research has shown the critical roles of mucins in tumorigenesis (van Putten and Strijbis, 2017;Maeda et al., 2018;Ganguly et al., 2020;Liu et al., 2020), which indicated their prospective usefulness in cancer diagnosis, prognosis, and therapy (Nabavinia et al., 2017;Aithal et al., 2018;Guo et al., 2018;Lucchetta et al., 2019;Pothuraju et al., 2020). Previous studies have raised hope that MUC4 can be a good candidate marker for pancreatic malignancy (Trabbic et al., 2019;Gautam et al., 2020). MUC4 could also be used in combination with MUC16 for detection of advanced ovarian cancer (Chauhan et al., 2006). However, it remains unclear whether MUC4 can be characterized as a friend or foe across the cancer types (Jonckheere and Van Seuningen, 2018). To address this issue, we performed an integrative analysis about MUC4 to understand its effect on survival and immunomodulation, which is necessary to develop the MUC4-based cancer therapy. Through a comprehensive analysis in large-scale datasets, we identified consistent expression levels of MUC4 in pancancer using independent datasets via Oncomine and GEPIA2, which showed that MUC4 expression compared with normal samples was upregulated in cervical cancer, lung cancer, and pancreatic cancer, while it was downregulated in head and neck cancer and prostate cancer. We analyzed its prognostic significance statistically via PrognoScan and Kaplan-Meier plotter. Increased MUC4 expression was significantly correlated with prolonged survival time in breast cancer, colorectal cancer, and kidney renal cell carcinoma, while it was associated with poor survival in ovarian cancer, brain cancer, lung cancer, pancreatic cancer, thymoma, bladder carcinoma. Overall, these findings strongly suggest that MUC4 can serve as a prognostic biomarker in pan-cancer. Furthermore, we explored the mechanism by which MUC4 influences prognosis. We depicted global alterations and epigenetic regulation of MUC4 across multiple cancer types, which showed that genomic alteration was an unfavorable factor for overall survival in LUSC and UCEC. As we expected, alteration in mRNA level and genomic level may influence tumor-malignant traits through different mechanisms. A previous study showed a high correlation between hypomethylation status and mRNA expression MUC4, and patients with MUC4 hypomethylation correlated with poor prognosis in pancreatic cancer (Yokoyama et al., 2016). Yamada et al. reported that the mRNA expression of MUC4 negatively correlated with its DNA methylation status at promoter regions in human lung cancer cell lines (Yamada et al., 2009), which was consistent with our results.
The effects on protein network, tumor growth, and metastasis were also concerned. Highly expressed MUC4 was correlated with MKI67 expression and negatively enriched with EMT in HNSC, which was consistent with its expression profile. In HNSC, upregulated MUC4 expression was enriched with EMT, which was identical with its poor prognosis. In addition, we focused on the functional roles on the tumor-associated microenvironment, especially in KIRC and PAAD. Our results demonstrated that the correlation with immune gene markers was not always the same as the overall trend (the relationships with immune cells). For example, in PAAD, B cells significantly correlated with MUC4 expression; however, gene markers of B cell had no relation to MUC4 expression. These discrepant implications on immune response and survival need further experiments for validation.
When considered together, our data demonstrated that MUC4 expression and promoter methylation status are potential prognosis biomarkers for lung cancer. MUC4 can be explored in pancreatic cancer as an early diagnostic tool. Thus, these findings in our study may provide new avenues for understanding the biological characterization of MUC4 and make sense in the design of therapeutic strategies.

DATA AVAILABILITY STATEMENT
The original contributions presented in the study are included in the article/supplementary material, further inquiries can be directed to the corresponding author.

AUTHOR CONTRIBUTIONS
X-PG and XG led the bioinformatic and biostatistical data analysis. X-PG and J-JD collected the literature, wrote the manuscript, and made the figures. XG and TX edited and made significant revisions to the manuscript. XG contributed to the study design and project supervision. All authors contributed to the article and approved the submitted version.