Prognostic Signicance of SQSTM1 in Breast Cancer: A Comprehensive Analysis

Background: SQSTM1 (Sequestosome 1, p62) is degraded by activated autophagy and involved in the progression of in various types of cancers. However, the prognostic role and underlying regulation mechanism of SQSTM1 in the progression and development of breast cancer remain unclear. Methods: In this study, 1336 samples with available mRNA data from Molecular Taxonomy of Breast Cancer International Consortium (METABRIC) database and 27 formalin xation and paran embedding (FFPE) tissue samples from the First Aliated Hospital of Xi’an Jiaotong University were collected to evaluate SQSTM1 expression in mRNA and protein levels. Kaplan–Meier and Cox regression were used for revealing prognostic value in three independent breast cancer independent datasets. Tumor Immune Estimation Resource (TIMER) database and Gene Set Variation Analysis (GSVA) was used to explore the relationship of SQSTM1 mRNA expression and immune inltration level in breast cancer. Dysregulation mechanisms of SQSTM1 were also explored including copy number variation (CNV), somatic mutation, epigenetic alterations and other transcription and post-transcription level using multiple datasets. Finally, Gene Set Enrichment Analysis (GSEA) was constructed to elucidate functional regulating performance of SQSTM1 in breast cancer. Results: The results showed that mRNA and protein level of SQSTM1 were signicantly elevated in breast cancer and receiver operating characteristic (ROC) curve showed that p62 may act as diagnostic biomarker. Lower expression of SQSTM1 predicted better outcome through multiple datasets. It was also found that SQSTM1 correlated with immune inltrates in breast cancer. Moreover, CNV and methylation of SQSTM1 DNA was correlated with SQSTM1 dysregulation and act as prognostic factors for breast cancer patients. Yet, somatic mutation status of SQSTM1 didn’t show any prognostic relevance. We also identied diverse transcription factors that directly bound to SQSTM1 DNA and the miRNAs which may regulate SQSTM1 mRNA. Finally, functional enrichment analysis revealed that SQSTM1 is related to cell signal transduction, oxidative stress and autophagy in breast cancer. Conclusion: Our ndings revealed that overexpression of SQSTM1 signicantly to poor survival and immune inltrations in breast cancer. In addition, SQSTM1 plays a key role in the progression of breast cancer and might be a promising biomarker for the diagnosis and personalized treatment of breast cancer patients. oxidative stress and autophagy in breast cancer. Our study provides new insights into the biological and clinical characteristics of SQSTM1 in breast cancer. Further large-scale multicentre clinical trials and studies are urgently needed to reveal complete gene expression proles and provide therapeutic regimens for breast cancer patients.


Introduction
Breast cancer is the one of the most frequently diagnosed malignant cancers and the leading cause of cancer-related deaths among women worldwide (1). The Global Cancer Statistics 2018 reported that 2,088,849 new cases and 626,679 deaths of breast cancer occurred globally in 2018 (2).The combination of surgery, chemotherapy and other strategies have made remarkable progress during the past few years. However, the clinical outcome of breast cancer patients still remains poor due to lack of reliable tumor biomarkers and personalized therapies. It is widely known that there is high heterogeneity in breast Page 4/30 cancer with different clinical, histological, and prognostic characteristics (3). Therefore, exploring effective diagnostic and therapeutic biomarkers to help stratify patients and optimize appropriate therapy strategies is signi cantly urgent. SQSTM1, also called p62, has been reported as an adaptor protein involved in autophagy and played as a central hub in various signal pathway and regulate multiple effectors, such as NF-kappaB and mTOR (4,5). Dysregulation of SQSTM1 is considered to be involved in various types of cancers, including hepatocellular carcinoma, lung adenocarcinoma, breast cancer, colon cancer (6)(7)(8)(9)(10). In human liver tissues, high expression of SQSTM1 correlates with rapid recurrence of resectable hepatocellular carcinoma which demonstrated its oncogenic role (11). According to a meta-analysis, high expression of SQSTM1 was associated with poor overall survival in lung cancer and might be useful to predict prognosis of lung cancer (12). On the other hand, previous study indicated that SQSTM1 enhanced breast cancer stem-like properties and promoted breast cancer metastasis promoter by binding vimentin (13,14). However, other researchers also revealed that SQSTM1 expression didn't show signi cant difference between breast cancer tissues and healthy adjacent tissues (15). These ndings suggest the SQSTM1 hold the promise as a novel biological and therapeutic marker and also needed further studies to elucidate its detailed role in human cancers.
The aim of our study is to shed light on the impact of SQSTM1 in the development and prognostic outcome of breast cancer. We evaluated SQSTM1 expression in mRNA and protein levels. It was found that SQSTM1 was upregulated in breast cancer tissues and indicated a poor prognosis in patients with breast cancer. Based on this, we investigated the mechanism of the dysregulation of SQSTM1 in breast cancer by identifying diverse regulation levels. Furthermore, GSVA and GSEA analysis were constructed to elucidate functional regulating performance of SQSTM1 in breast cancer. Our nding indicated that SQSTM1 is a novel therapeutic biomarker and may be useful for improvement of breast cancer treatment ( Fig. 1).

Patients and samples selection
This study was approved by the Human Ethics Committee of the First A liated Hospital of Xi'an Jiaotong University. Between 2011 and 2017, para n-embedded breast cancer tissues (n = 27) and adjacent non-tumorous tissues (n = 27) from 27 patients underwent breast cancer surgery at Department of Breast Surgery was collected in our study. SQSTM1 mRNA expression analysis SQSTM1 mRNA expression in multiple The Cancer Genome Atlas (TCGA) tumors and adjacent normal tissues was explored by "DiffExp" module of TIMER (https://cistrome.shinyapps.io/timer/). The mRNA expression (microarray) differences between breast cancer tissues and normal tissues were selected from the Gene Expression Omnibus (GEO) database (GSE54002, GSE42568) using GEOquery. SQSTM1 mRNA expression, SQSTM1 CNV data, SQSTM1 mutation pro le as well as complete clinical data of breast cancer patients were downloaded from METABRIC using cBioPortal (https://www.cbioportal.org/) and TCGA using UCSC Xena (http://xena.ucsc.edu/) via R (version 3.6.3).
Immunohistochemical staining SQSTM1 expression in protein level (p62) was estimated by immunohistochemical (IHC). All tissues were formalin-xed para n-embedded and cut in 3-µm sections. All the experiment procedures were performed based on manufacturer's protocol. The rabbit anti-SQSTM1 antibody was provided by Abcam (ab121146). The extent of positively stained cells was graded as: 0 (positive cells 0-5% of the cells), 1 (6-25% of the cells), 2 (26-50% of the cells), 3 (51-75% of the cells), 4 (76-100% of the cells). The staining intensity score was classi ed by four grades: negative, 0; weak, 1; medium, 2; and strong, 3. Expression levels of p62 were determined by nal staining scores, which were calculated by multiplying the positive cells scores and intensity and ranging from 0 to 9.

Immune In ltration of SQSTM1 in breast cancer
The correlation of SQSTM1 expression with the tumor-in ltrating levels (B cells, CD4 + T cells, CD8 + T cells, neutrophils, macrophages and dendritic cells) in BRCA (Breast Invasive Carcinoma) was evaluated by the module of TIMER. Based on the previous study, we conducted the gene set variance analysis (GSVA) to estimate the correlation between SQSTM1 expression and immune cell abundance in breast cancer tissue samples using GSVA package in R software (16).

The prognostic value of SQSTM1 mRNA in public databases
To investigate the prognostic role of SQSTM1 in breast cancer, Kaplan-Meier (Log-rank tests) analysis was conducted to determine the prognostic signi cance using METABRIC and GEO breast cancer cohorts (GSE1456, GSE9195).
Transcription factors and miRNA identi cation GCBI (https://www.gcbi.com.cn) is a web which integrates diverse genetic, clinical and bioinformatic data (17). In this study, we used GCBI to identify transcription factors which interacts with SQSTM1. The Cistrome (http://cistrome.org/db/) is a platform which includes ChIP-seq, DNase-seq and ATAC-seq data from multiple public databases of human and mouse (18). In our study, we used Cistrome's Chip-seq data to con rm the transcription factors directly bound to SQSTM1 DNA. The miRDB (http://www.mirdb.org/), DIANA tool (www.microrna.gr), TargetScan (http://www.targetscan.org/) are online resources for miRNAs and targets predictions. In this study, they were used to predict potential microRNAs targeting on SQSTM1 mRNA.

Functional enrichment analysis
The limma package in R was performed for TCGA dataset differential expression analysis. The cutoff for log (Fold change) FC in our study was 0.477 and a P value < 0.05 was statistically signi cant. The coexpressed genes related to SQSTM1 in breast cancer was retrieved from the Coexpedia (http://www.coexpedia.org/). The Gene-set enrichment analysis for Gene ontology (GO), the Broad Molecular Signatures Database (MSigDB) and Kyoto Encyclopedia of Genes and Genomes pathway (KEGG) was performed by clusterPro ler package and the GSEA program (version 3.0).

Statistical analysis
The statistical analysis in this study was performed by SPSS (version 19.0) software. The association between expression of SQSTM1 and clinical parameters were analyzed by Chi-square test. Multivariate logistic regression was used to nd independent in uence factors for SQSTM1 mRNA expression. Screening for prognostic factors for breast cancer patients was performed by univariate and multivariate Cox regression and visualized by Review Manager (version 5.3). Overall survival of breast cancer was examined with Kaplan-Meier curve and compared by log-rank test using GraphPad prism (version 7.0) software. A value of P < 0.05 was considered to be statistically signi cant.

SQSTM1 is overexpressed in patients with breast cancer at RNA and protein levels
We rst explored SQSTM1 mRNA expression distribution in various tumors using TIMER dataset. It was shown that SQSTM1 mRNA was signi cantly higher in most common tumor tissues compared with normal tissues. It is noteworthy that SQSTM1 acted as an oncogene in different subtypes (Luminal, Her2, basal) of breast cancer ( Fig. 2A). Furthermore, same results were validated in 2 independent GEO cohorts (GSE54002, GSE42568) when compared SQSTM1 mRNA expression between breast cancer tissues and normal tissues (Fig. 2B). SQSTM1 expression in protein level was examined by IHC staining. As shown in Fig. 3, p62 was mainly expressed in the cytoplasm of breast cancer cells and signi cantly overexpressed in breast cancer tissues compared with adjacent non-tumorous tissues (P < 0.001). The diagnostic performance of p62 for distinguishing breast cancer from non-breast cancer was assessed by ROC analysis with AUC of 0.846 (95% CI = 0.760-0.933, P < 0.001). Our results indicated that p62 can be used as a diagnostic biomarker ( Fig. 3). All the above data demonstrated the oncogene role of SQSTM1 in breast cancer and can be utilized as a predictive tool for achieving precision medicine.

Correlation of SQSTM1 mRNA expression with clinicopathologic characteristics
In order to explore the clinical signi cance of SQSTM1 in breast cancer, we explored RNA-seq data from METABRIC database with SQSTM1 mRNA expression (n = 1336) and detailed clinical information. Using Chi-square test, we assessed the correlation between SQSTM1 mRNA expression and clinical-pathologic characteristics. As were shown in Table 1, the expression level of SQSTM1 mRNA was signi cantly associated with ER status (P = 0.018), hormone therapy (P = 0.006).Multivariate logistic regression indicated that ER status (OR = 1.762, P < 0.05) and PR status (OR = 0.743, P < 0.05) were independent in uence factors of SQSTM1 mRNA expression in breast cancer patients ( Table 2).

Prognostic value of SQSTM1 in breast cancer
In our METABRIC cohort, by plotting Kaplan-Meier curve, we found that breast cancer patients with higher SQSTM1 mRNA expression (median survival time = 130.7 months) tended to have a worse overall survival (OS) than patients with lower SQSTM1 mRNA expression (median survival time = 172.9 months, P < 0.001) (Fig. 5A). By using GEO database, we validated the prognostic role of SQSTM1 in breast cancer patients. Lower SQSTM1 expression indicated favorable prognosis (OS, DFS and RFS) in breast cancer from 2 independent cohorts (GSE1456, GSE9195) ( Fig. 5B and C). All these results indicated that high SQSTM1 mRNA expression may be a poor prognostic biomarker of breast cancer.

Association between SQSTM1 expression and immune in ltration level in breast cancer
Tumor microenvironment has been demonstrated to serve as a "complex network" of different tumor cells, extracellular matrix components, chemotactic factor and other types of cells which forms the basis for tumor cancer cell proliferation and metastasis (19). Here, we analyzed the correlation between SQSTM1 expression and immune in ltration levels in breast cancer. As were shown in Fig. 6A, SQSTM1 expression was inversely associated with in ltrating levels in breast cancer. It is noteworthy that SQSTM1 expression has positive correlation with tumor purity in breast cancer.
Next, we compared SQSTM1 expression in breast cancer patients with different GSVA score (lowest 25% versus highest 25%) of multiple immune cells through TCGA dataset. The results indicated that higher SQSTM1 expression was signi cantly correlated with higher in ltration of memory B cell, activated CD4 + T cell and neutrophil. However, it was shown different result in activated CD8 + T cells (Fig. 6B). These results further demonstrated that SQSTM1 may serve as an immune modulatory role in breast cancer and large-scale projects are still urgently needed in the near future.
The mechanism of SQSTM1 expression dysregulation in patients with breast cancer Numerous studies have indicated that CNV unbalanced gene expression by disrupting the structure of gene coding regions. Next, we evaluated the copy number alterations of SQSTM1 using a cohort of 1904 breast cancer patients from METABRIC database (Shallow Deletion, n = 192; Diploid, n = 1460; Ampli cation, n = 31; Gain, n = 221). We found that 56.8% (252/444) patients in the altered group harboring SQSTM1 ampli cation/gain. This result indicated that the ampli cation/gain of gene copy numbers was likely to be one of the main mechanisms of over-expression of SQSTM1 in breast cancer patients. Consistently, breast cancer patients with SQSTM1 ampli cation/gain exhibited higher SQSTM1 mRNA expression compared with shallow deletion and diploid (no alteration) group. By drawing the Kaplan-Meier survival curve, the results revealed that patients with SQSTM1 ampli cation/gain signi cantly associated with worse overall survival compared with other groups (Fig. 7).
In order to explore the potential clinical signi cance of SQSTM1 mutation, we rst evaluated its mutation pro le in the METABRIC database. The results showed that there was no SQSTM1 mutation in the selected patients. Next, patients obtained from TCGA database with mutation pro les were validated. Compared with the high-frequency altered genes such as PIK3CA, AKT1 and PTEN, SQSTM1 mutation frequency is rare and has no predictive value on the prognosis of breast cancer patients (P = 0.338) (Fig. 8).
In addition to point mutations and CNV, epigenetic changes (especially DNA methylation) also play an important role in regulating speci c genes expression and the development of breast cancer. We then investigated characteristics of the SQSTM1 promoter methylation in breast cancer. First, the heat map of the SQSTM1 methylation value used different probes were drawn from TCGA dataset. The Kaplan-Meier survival analysis showed that patients with lower methylation of SQSTM1 experienced longer overall survival and disease speci c survival signi cantly, which further suggested that the high expression of SQSTM1 plays a critical prognostic role in breast cancer (Fig. 9). All of the above data showed that upregulation of SQSTM1 expression involved in the development and progression of breast cancer.  Supplementary Fig. 1). Based on this analysis, we then used Chip-seq data of Cistrome and con rmed that CTCF, ERG, EP300, E2F1, FOXA1 can directly bind to SQSTM1 DNA in breast cancer (Supplementary Table 1).

SQSTM1 is related to cell signal transduction, oxidative stress and autophagy
To clarify the biological molecular mechanism of SQSTM1 in breast cancer, we rst performed differential gene expression analysis based on LIMMA package in samples with high expression of SQSTM1 (N = 552) and low expression of SQSTM1 (N = 552) from TCGA database. Our analysis found that a total of 387 genes were signi cantly up-regulated and 561 genes were signi cantly down-regulated (Fig. 10A). In addition, SQSTM1 was observed to be associated with various signal transduction pathways according to KEGG analysis, such as JAK/STAT and PI3K/Akt, which was consistent with previous reports (Fig. 10B). Next, we used the MSigDB Hallmark gene set (Fig. 10C) for GSEA. The results showed that compared with high expression levels of SQSTM1, low levels of SQSTM1 were signi cantly related to oxidative phosphorylation, peroxisome, DNA repair and reactive oxygen species pathway.
Coexpedia is a distinct co-expression database which offers biomedical hypotheses through medical subject headings. In our study, the co-expression genes in breast cancer associated with SQSTM1 were explored from Coexpedia database in order to clarify the underlying regulation network and mechanism of breast cancer. Through exploring GSE12237, GSE7848 and GSE14018, a total of 19 genes, such as LAMTOR2, PIR, and GULP1 were identi ed ( Supplementary Fig. 3) ( Table 4).
The GO analysis based on SQSTM1 and its related genes were then constructed. The top 20 Go terms enrichment of the gene lists was showed in Supplementary Fig. 4. The most signi cantly enriched GO terms of BP, CC and MF for SQSTM1 and co-expressed genes were negative regulation of endoplasmic reticulum unfolded protein response (GO: 1900102; P = 7.12E-05), ionotropic glutamate receptor binding (GO:0035255; P = 0.0005), DSIF complex (GO:0032044; P = 0.0017) and amphisome (GO:0044753, P = 0.001693), respectively. Altogether, these data indicated that SQSTM1 is related to cell signal transduction, oxidative stress and autophagy thus plays a key role in the progression of breast cancer.

Discussion
Breast cancer remains a global health concern as a type of aggressive tumor. Over the past few years, a great number of studies have demonstrated the molecular characteristics of breast cancer with genetic and clinical heterogeneity which restrict the accuracy of typical morphological and pathological classi cation. However, newly molecular targeted drugs by identifying and discovering diagnostic and prognostic biomarkers have offered new elds of breast cancer treatment. Therefore, discovering new therapeutic target involved in the progression of tumor to improve the prognosis of breast cancer is urgent nowadays.
In the current study, we explored the clinical signi cance of SQSTM1 based on RNA expression data from METABRIC/GEO databases and protein expression data from our hospital cohort. We found that SQSTM1 mRNA and protein level were signi cantly higher in breast cancer tissues than adjacent non-tumorous tissues. It was also showed that SQSTM1 was a high-risk factor and could be an independent prognostic factor in patients with breast cancer using univariate and multivariate Cox analyses. Besides, we also found high SQSTM1 mRNA expression predicted poor outcome in breast cancer patients through multiple databases. By plotting ROC curve, we observed the AUC value for p62 was 0.846 which was a potential predictor of breast cancer. All these data suggest that SQSTM1 might be a therapeutic target for breast cancer.
Numerous studies have demonstrated the role of autophagy activity on tumor cells involving in modulating functions of T cells such as CD8 + cytotoxic T cells and regulatory T cells (20,21). Although it is widely believed that breast cancer was a relatively non-immunogenic cancer and showed poor response to immunotherapy, immuno-oncology focused on tumor-in ltrating lymphocytes have showed remarkable progress in treatment of breast cancer recently, especially for those with hormone receptor negative subtypes (22). Since the role of SQSTM1 in immunity of breast cancer remains unclear, in this study, we found that SQSTM1 mRNA expression was correlated with diverse immune in ltration levels signi cantly. In addition, we also found that higher SQSTM1 expression was signi cantly correlated with higher in ltration of memory B cell, activated CD4 + T cell and neutrophil. These data indicated the underlying role of SQSTM1 in the breast cancer microenvironment Then, we tried to investigate the mechanisms of SQSTM1 dysregulation. Through examining its CNV, DNA methylation and somatic mutation status in patients with breast cancer, it was found that copy number ampli cation/ gain of SQSTM1 could be the key driver mechanism for its overexpression. In addition, we also observed SQSTM1 CNV and methylation status were signi cantly associated with survival of breast cancer patients. In the future, detection of SQSTM1 copy number ampli cation, methylation, and overexpression status may provide new guidelines of evaluation and adjustment of breast cancer treatment strategies.
Predicting the transcription factors regulating SQSTM1 made it possible to get a better understanding of the gene expression patterns and regulation mechanisms in breast cancer. In our study, we identi ed CTCF, ERG, EP300, E2F1, FOXA1 can directly bind to SQSTM1 DNA in breast cancer and may help complement the regulatory network thus develop novel effective targeted therapeutic strategies for patients.
miRNAs are a class of endogenous non-coding RNAs that regulate the expression of different transcription factors at the post-transcriptional level. Since 2002, George and other scientists rst reported that miRNAs are dysregulated in tumors, more and more scholars devoted to studying the role of miRNAs in tumorigenesis and development (23). Current ndings showed that miRNAs were dysregulated in a variety of malignant tumors, and the regulation of tumorigenesis and affected various activities, including tumor cell proliferation, invasion and metastasis, drug resistance, angiogenesis and immune escape. Previous study demonstrated that miR-17/20/93/106 targeted SQSTM1 and promoted hematopoietic cell expansion (24) thus implicated SQSTM1 expression was regulated by miRNAs in different malignancies. In our study, we identi ed 6 common miRNAs (miR-106b-5p. miR-20a-5p, miR-106a-5p, miR-93-5p, miR-17-5p, and miR-20b-5p) by different databases and more convincing evidences in the future which may lead to novel therapeutic strategies for breast cancer.
A large number of in vivo and in vitro studies have reported that SQSTM1 can promote tumor development and malignant phenotypes such as tumor growth, invasion, migration and apoptosis inhibition via multiple signal transduction pathways (13,25,26). Previous studies have con rmed the double-edged sword effect of autophagy in regulating tumors which highlighted the importance of SQSTM1 expression pattern. It has recently been reported that the oxidation of SQSTM1 promoted its oligomerization via disul de-linked conjugates then activated autophagy which facilized cell homeostasis and survival under oxidative stress from aging or cancer (27). In addition, an increase of SQSTM1 in autophagy-de cient cells directly bonded to and inhibited nuclear RNF168, an E3 ligase essential for histone H2A ubiquitination and DNA damage responses (28). Combining previous results and our ndings, it can be concluded that autophagy defects with SQSTM1 accumulation can impair the DNA repair activity of cells thus leading to tumorigenesis.
Then, a total of 19 co-expressed genes of SQSTM1 in breast cancer were explored by Coexpedia. The highest score gene was LAMTOR2 which is a convergence point for RAF/MEK/ERK and PI3K/AKT/mTOR pathways (29). Numerous studies have veri ed the signi cance of two signaling pathway in the progression of breast cancer. In addition, Lin and his colleagues have reported that LAMTOR2 interacted with SQSTM1 and was required for recruiting TAX1BP1 to autophagosomes (30). However, to the best of our knowledge, there is no study about the associated between LAMTOR2 and SQSTM1 in breast cancer and the underlying regulation network requires further investigations.
Furthermore, the potential biological processes mainly involved in regulation of endoplasmic reticulum unfolded protein response has also been discussed. Unfolded protein response (UPR) is a protective cellular response activated by endoplasmic reticulum stress (31). SQSTM1 was known to protect cells against tunicamycin (TM)-mediated oxidative damage through Nrf2 activation (32, 33). This nding further demonstrated the importance of SQSTM1 in regulating oxidative stress of tumor progression.
Accumulating evidence demonstrated that SQSTM1 dysregulation involved in multiple tumor progression.
In the current study, we found that SQSTM1 acted as an oncogene in breast cancer. Importantly, the overexpression of SQSTM1 is related to poor prognosis. We also explored the dysregulation mechanism of SQSTM1 and found CNV and methylation might be the potential targets for patients. In the upstream of SQSTM1, several transcription factors and miRNA have also been identi ed. The miRNA mimic/inhibitor might be a promising target for new cancer therapy in the future. The precise mechanism of SQSTM1 needed further in vivo and in vitro experiments to elucidate its biological function. These results have provided such an exciting future, in which focusing on SQSTM1 pro ling might on one day help tailor therapy strategies and achieve better management of breast cancer.

Conclusion
This study indicated that SQSTM1 is a promising diagnostic and prognostic target in breast cancer patients by exploring multiple cohorts. Overexpression of SQSTM1 was correlated with tumor progression, poor survival, immune in ltrations in breast cancer. Multiple mechanisms involved in transcription and post-transcription levels are responsible for the dysregulation of SQSTM1. In addition, elevated SQSTM1 was also associated with cell signal transduction, oxidative stress and autophagy in breast cancer. Our study provides new insights into the biological and clinical characteristics of SQSTM1 in breast cancer. Further large-scale multicentre clinical trials and studies are urgently needed to reveal complete gene expression pro les and provide therapeutic regimens for breast cancer patients.  Figure 1 Work ow of this study.