Analysis of genome-wide mutation profile and establishment of risk signature for prognosis of bladder cancer

doi:10.21203/rs.3.rs-29487/v1

Background: Emerging studies have shown that a variety of gene mutations occur in development and progression of cancer and highly mutation genes could play oncogenic or tumor suppressive roles in cancer. Therefore, our aim is to explore mutation genes which affect the prognosis of bladder.

Methods: Mutation profile was obtained and analyzed from TCGA data set. A mutation-based signature was established by multivariable Cox regression analysis. Kaplan-Meier was performed to assess the prognostic power of signature. Time-dependent ROC was conducted to evaluate predictive accuracy of signature for bladder cancer patients.

Results: There are 20177 genes have alteration in 403 bladder patients and 662 of them were frequently variation (mutation frequency > 5%). In this study, we assessed the prognostic predictive ability of 662 highly mutated genes and identified a mutation signature as an independent indicator for predicting the prognosis of bladder. The time-dependent ROC showed that AUC were 0.893, 0.896, 0.916 and 0.965 at 1, 3, 5 and 10 year, respectively. Stratified analysis and Multivariate Cox analysis showed that this mutation signature was reliable and independent biomarker. Furthermore, the nomogram predictive model can be used to effectively predict clinical prognosis of bladder patients. The decision analysis curve showed patients with risk threshold of 0.03-0.92 potentially yielded clinical net benefit. Finally, we identified several signaling pathways that associated with risk score by GSEA and KEGG analysis including PI3K-Akt signaling pathway and so on.

Conclusions: In general, this study provide an optimal mutation signature as potential prognosis biomarker for bladder patients.

Cancer Biology

Bladder cancer

Biomarkers

Mutation gene

Prognosis

Nomogram

Decision Curve Analysis

Bladder cancer (BC), the most common malignancy of the urinary tract, was the 9th most common cause of tumor-related death worldwide and was ranked fourth of the most common cancers in male, with an estimated 17,670 deaths (12,870 cases for male and 4800 cases for female) and 80,470 new cases (61,700 cases for male and 18,770 cases for female) in 2019 cancer statistics [1]. Indicated by the degree of detrusor involvement, bladder cancer can be clinically classified into the following categories: non-muscle invasive (NMIBC) or muscle invasive bladder cancer (MIBC) through the stage [2]. Despite early prevention and intervention, many patients with MIBC are still at high risk of recurrence and death [2]. Although the prognosis of NMIBC is better than MIBC, we should not ignore the high recurrence risk of NMIBC patients [3]. Therefore, it is imperative to screen a novel potentially reliable prognostic indicators to forecast the survival of bladder patients.

So far, gene mutation theory, widely accepted by the scientific community, believes that mutation in gene level are main cause of cancer. When it comes to the development of cancers, we have to talk about somatic mutations. It is reported that partial and complete tumor genome sequences exist hundreds to thousands of variants in most cancers including bladder cancer [4]. Consistently, in recent years, with the deepening research for cancer, increasing reports demonstrated that gene mutation exerted a positive or negative role in cancer [5–8]. Previous study have identified multiple frequently mutation genes in bladder cancer, including TP53, FGFR3, RB1, PIK3CA, KDM6A, CREBBP, EP300 and so on [9–12]. Several gene variants have been demonstrated to confer susceptibility to bladder cancer [13–15] and some mutation genes were identified novel biomarker for bladder cancer such as TERT and CCNE1 [16–18]. However, due to the high heterogeneity of cancer, including massive random genome alteration, it is therefore necessary to continuously research and evaluate the utility of gene mutation.

In this study, we investigated the whole mutation profile in bladder patients from TCGA database. Then, we identified a mutation signature consisting of 23 mutated genes. This signature is an independent and high accuracy indicator to appraise the survival of bladder patients. Our results confirmed a novel panel of combined mutation genes that connection with the survival of bladder patients and combined risk score with conventional clinical characteristics have high clinical application value.

2.1 Patient Samples

Mutation data and expression profiles of genes and corresponding clinical information for bladder patients were available at The Cancer Genome Atlas of the National Cancer Institute (TCGA, http://cancergenome.nih.gov) and UCSC Xena (https://tcga.xenahubs.net). In this study, patients were removed if follow-up information was lacking (1 cases). The remaining 403 samples were further analyzed and clinical characteristics include age, gender, TNM stage and so on.

2.2 Mutation gene collection and signature construction

Using univariable and multivariable Cox regression analysis, a prognostic mutation signature and the risk score model formula were constructed as follows: risk score = β₁ gene₁ * mutation status of gene₁ + β₂ gene₂ * mutation status of gene₂+…β_n gene_n * mutation status of gene_n. In this formula, β indicates the coefficients of every mutated gene in multivariable Cox regression analysis. "0" means that the gene is not mutated, and "1" means that the gene is mutated including Missense Mutation, In Frame Del, Silent, RNA, Splice Site, Frame Shift Del, Frame Shift Ins and Transition Start Site.

2.3 Predictive nomogram of the mutation signature

The development and progression can be diagnosed and predicted using nomogram with a variety of indicators [19]. To elucidate the clinical application of our mutation signature, we assembled a predictive nomogram by utilizing the "rms", "nomogramEx" and "regplot" R package to estimate the 1-, 3- and 5-year OS of bladder patients. The factors included risk score, TNM stage, T stage, N stage, tumor status and lymphovascular invasion. To evaluate the clinical application of our nomogram, we performed decision curve analysis (DCA) and clinical impact curve. In DCA, the x-axis represented the percentage of threshold probability and y-axis represented net benefit. The clinical impact curve can appraise the number of high-risk individuals with different possibility and true positive number among 1000 participants [20].

2.4 Functional analysis

The GSEA (http://www.broadinstitue.org/gsea/index.jsp) was conducted by the JAVA.8 program using Molecular Signatures Database (MSigDB) [21]. Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway was implemented to illustrate related signaling pathways and potential biological functions related to high risk or low risk groups using the Database for Annotation, Visualization, and Integrated Discovery website (DAVID, https://david.ncifcrf.gov/summary.jsp).

2.5 Statistical Analysis

The Kaplan-Meier analysis as well as univariate and multivariate Cox regression analysis were performed to identify and validate mutation genes significantly related to the survival of bladder patients. Univariate and multivariate Cox proportional hazard regression analysis of mutated genes in bladder were performed with R package 3.5.1 (https://www.r-project.org/, v3.5.1). The cBioPortal database was used to identify the co-occurrence and mutual exclusivity of mutated genes within the TCGA Pancancer Atlas cohort [22]. Univariate and multivariate Cox proportional hazard regression analysis of clinical characteristics were constructed using SPSS 22.0 (SPSS Inc., Chicago, IL). The time-dependent ROC curve was used to evaluate predictive accuracy of prognostic model using R package. The GEPIA (http://gepia.cancer-pku.cn) [23] and UALCAN (http://ualcan.path.uab.edu/index.html) [24] website were used to analyze differentially expression, expression based on clinical stage, survival of genes and so on.

3.1 Somatic mutation profile of bladder patients

The variant classification and variant type of bladder patients shown in Fig. 1a-b. The transition and transversion plot summarizes SNVs into six categories (Fig. 1c-e). There are 20177 genes have alteration in 403 bladder patients and 662 of them were frequently variation (mutation frequency > 5%) [25]. The top 30 of the individual mutation frequencies are illustrated in Fig. 1 f. The five most-altered genes were TTN (0.56), TP53 (0.41), MUC16 (0.36), KMT2B (0.24) and PIK3CA (0.22).

3.2 Identification of mutation genes related to the survival of bladder patients

To evaluate the influence of somatic mutation genes on overall survival, we assembled data encompassing 403 bladder patients form TCGA database, listed in Table 1. By subjecting 662 highly mutated genes to univariate Cox analysis, we identified 57 mutated genes related to the overall survival of bladder patients (P < 0.05, Table 2). To construct a prognostic classifier, these 57 candidate mutated genes were considered candidates for further multivariate Cox analysis. The optimal tuning parameter identified twenty-three mutated genes: AHNAK2, RYR2, USH2A, DCHS2, BRCA2, RP1L1, FBN2, ZFYVE26, ANK1, NCOR1, F8, WNK1, IGSF10, COL11A1, RANBP2, MED1, FANCM, MYO18B, ZNFX1, MROH2B, AUTS2, CNOT1, COL5A1 (Table 3). The mutation profile of these mutated genes generated from cBioPortal for Cancer Genomics (Additional file 1a). The interactions between these mutated genes were also generated from cBioPortal. As shown in Additional file 1b, most of these mutated genes are co-occurrence, suggesting that these genes might play an essential role together. Furthermore, we built a risk score formula so as to calculate the overall survival of bladder patients based on the mutation status of individual mutated genes and their respective coefficients as follows: 23 mutated genes risk score = (-0.6344 * mutation status of AHNAK2) + (-0.5828 * mutation status of RYR2) + (-1.0395 * mutation status of USH2A) + (-0.9495 * mutation status of DCHS2) + (1.0108 * mutation status of BRCA2) + (-1.2290 * mutation status of RP1L1) + (-1.2238 * mutation status of FBN2) + (1.7481 * mutation status of ZFYVE26) + (-1.5427 * mutation status of ANK1) + (-1.4791 * mutation status of NCOR1) + (-1.1631 * mutation status of F8) + (-1.1986 * mutation status of WNK1) + (-0.8949 * mutation status of IGSF10) + (-0.7952 * mutation status of COL11A1) + (-1.3281 * mutation status of RANBP2) + (-1.8627 * mutation status of MED1) + (-2.0523 * mutation status of FANCM) + (-1.6193 * mutation status of MYO18B) + (-1.3542 * mutation status of ZNFX1) + (-1.0805 * mutation status of MROH2B) + (-1.5807 * mutation status of AUTS2) + (-1.5486 * mutation status of CNOT1) + (0.6871 * mutation status of COL5A1). Next, we computed the 23-mutation signature risk score for per patient. The distribution of the mutation risk score, the survival situation of bladder patients and the alteration of mutated genes were obtained. As shown in Fig. 2a, patients in high risk group have shorter survival and higher mortality than those in low risk group. Then, we used the median risk score as the cut-off point to divide bladder patients into high-risk group (n = 201) or low-risk group (n = 202). Kaplan-Meier curve were performed and the result appeared that there are shorter OS and poor prognosis for patients with high risk score (P < 0.0001, Fig. 2b). The time-dependent ROC curve displayed that the predictive sensitivity and specificity of the mutation signature were 0.893, 0.896, 0.916 and 0.965 at 1-, 3-, 5- and 10 year, respectively (Fig. 2c). Taken together, this mutation signature have high accuracy and is a potentially novel indicator to evaluate the patients' survival.

3.3 The mutation signature is an independent prognostic indicator

We conducted the univariate and multivariate Cox analysis so as to further determine if the risk score was an independent risk factor for clinical outcome of bladder patients. The univariate Cox analysis showed that risk score, TNM, T stage, N stage, tumor status as well as lymphovascular invasion were significantly associated with the prognosis of patients (Fig. 3a). Furthermore, Risk score and tumor status were considered as independent risk elements for the prognosis of bladder patients by using the multivariable Cox analysis (Fig. 3a). These results demonstrated that the mutation signature remain independently related to overall survival after adjustment by conventional clinical characteristics.

Next, we performed Kaplan-Meier curve to examine prognostic effect of clinical characteristics. The results showed that TNM stage (P < 0.0001), T stage (P = 0.0003), N stage (P = 0.0001), Tumor status (P < 0.0001) and Lymphovascular invasion (P = 0.0006) were importantly related to the survival of bladder patients (Fig. 3b-d, 3 h-i). However, other clinical characteristics were not related to the patient' survival, such as age (P = 0.0941, Additional file 2a-e). Furthermore, we explored the correlation between risk score and clinical features. The results revealed that risk score was elevated in higher T stage and N stage (Fig. 3f-g). Contrast to patients with tumor, patients with tumor free have lower risk score (Fig. 3j). Further, higher risk score was chiefly enriched in patients with lymphovascular invasion (Fig. 3k). However, there are no significant correlation between the mutation signature and other clinical characteristic, including TNM stage and so on (Fig. 3e and Additional file 2f-j).

To appraise whether our mutation signature manifest strong prognostic ability within same clinical features, stratified analysis were further implemented. Patients were stratified into different groups according to each clinical features. The stratified survival analysis results showed that patients were divided into high-risk or low-risk group and there are shorter OS and poor prognosis for patients in high-risk group (Fig. 4). In general, the results above suggested that our mutation signature was independent risk factor to evaluate the survival of patients and could divide patients into subtypes with different prognosis.

3.4 Clinical Application of the mutation signature

Based on univariate Cox analysis and relationship analysis, we further constructed a prognostic nomogram including risk score, stage, T stage, N stage, tumor status and lymphovascular invasion to estimate the survival of 1-, 3- and 5-year of bladder patients (Fig. 5a). The points of each prognostic factor add up to get the total points and the higher the patients' points, the worse the prognosis. The internally validated Harrell's C-index was 0.817, demonstrating the mutation signature based nomogram exhibited superior performance in clinical condition. Taken together, these results suggested that the predictive nomogram has potentially high clinical application value. Furthermore, to evaluate clinical utility of predictive nomogram, we performed decision curve analysis as well as clinical impact curves. The decision curve analysis result showed that the risk prediction nomogram with the risk threshold probabilities ranging from 0.03–0.92. When using 0.55 as the risk threshold, patients yielded the best clinical net benefit (Fig. 5b). Clinical impact curve was performed to estimate the numbers of high risk people for each risk threshold and true positives. For instance, if 1000 patients were examined at a risk threshold of 0.55, about 220 patients would be deemed high risk, whereas about 198 patients were truly positives (Fig. 5c).

3.5 The mutation signature are better than other biomarkers in survival prediction

To determine if our mutation signature are better than other biomarkers, we also constructed a mRNA signature and a miRNA signature by univariate and multivariate analysis from differentially mRNA and miRNA between cancer and adjacent cancer, respectively. Then formula as follows: mRNA signature = (0.1446 * EMP1) + (-0.1375 * CCDC80) + (0.0596 * SERPINB12) + (0.1690 * LRP1) + (0.1087 * ANXA1) + (-0.1606 * CHMP4C) + (-0.1290 * UNC5CL) + (-0.2277 * DCAF4L1) + (0.0838 * FER1L4) + (-0.1501 * OAS1) + (-0.1222 * APOBEC3H) + (0.1578 * PEX5L) + (0.1132 * CLSTN2) + (-0.1317 * CATSPER4) and miRNA signature = (-0.3472 * miR-590) + (0.2458 * miR-365a) + (0.6026 * miR-337) + (0.2907 * miR-192) + (-0.5377 * miR-127) + (-0.1839 * miR-378c) + (-0.2279 * miR-576) + (-0.3928 * miR-194-2) + (-0.1462 * miR-4636) + (0.0967 * miR-217) + (-0.2276 * miR-3913-2). Simultaneously, we collected other miRNA signature for predicting bladder patients' survival from previous studies [26]. Kaplan-Meier curve as well as time-dependent ROC analysis were conducted to compare their clinical outcome (Fig. 6a-c). The Kaplan-Meier results showed that compared to patients with other signature's low risk score, patients in our mutation signature's low-risk group have distinctly longer OS and better prognosis, similarly, compared to patients with other signature's high risk score, patients in our mutation signature's high-risk group have distinctly shorter OS and poor prognosis. The results of time-dependent ROC analysis displayed that the mutation signature have clearly higher AUC than other signature (Fig. 6c-d). In general, our mutation signature exhibit a better predictive ability for prognosis of bladder patients.

3.6 The higher risk score was related to higher expression of APOBEC3B

Recent studies has suggested that APOBEC3B is a source of mutations in a variety of cancer embracing bladder cancer, ovarian cancer and so on [27–30]. Hence, we first investigated the expression of APOBEC3B in various cancers using GEPIA website (Fig. 7a-b). The results confirmed that APOBEC3B is highly expressed in many cancer types, such as bladder cancer (Fig. 7a-c). In addition, the result also showed that APOBEC3B is highly expressed in bladder cancer using ulacan website (Fig. 7d). Then, we investigated the correlation of APOBEC3B expression with TNM stage as well as N stage. The results showed higher expression of APOBEC3B mainly enriched in advanced stage and higher N stage (Fig. 7e-g). Next, we analyzed the relationship between our mutation signature and APOBEC3B to explore if our signature is associated with APOBEC3B and result showed that expression of APOBEC3B is higher in patients with high risk score (Fig. 7h). In general, these results demonstrated that APOBEC3B was not only high expression in bladder cancer, but also had a close connection with our signature.

3.7 Identification of signaling pathways correlated with the mutation signature

To identify the potential signaling pathways correlated with this mutation signature, we performed GSEA analysis and KEGG analysis. In the light of the results of GSEA analysis, we observed that hallmarks involving "Inflammatory response", "Interferon alpha response" and "Interferon gamma response", were mainly enriched in low-risk group (Fig. 8a). "Epithelial mesenchymal transition"(EMT) and "Hedgehog signaling" were enriched in high-risk group although they are not statistically significant (Fig. 8b). Simultaneously, we analyzed the differential gene expression data between high- and low-risk groups (|LogFC| > 2, P < 0.05). Then, based on the results of KEGG, upregulated genes in low risk group were primarily involved in "Pancreatic secretion", "Neuroactive ligand-receptor interaction" and so on, while upregulated genes in high risk group were primarily involved in "PI3K-Akt signaling pathway", "Ras signaling pathway" and so on (Fig. 8c). These results suggested this mutation signature might be participated in deterioration of BC through these pathways.

BC has become one of the most common malignancy due to its increasing incidence and high recurrence rates [31]. Therefore, early diagnosis and prognosis prediction need to be taken seriously. It is common knowledge that the accumulation of genetic alteration in tumor-related genes is necessary for the development and progress of carcinogenesis [32]. Therefore, our study concentrated on addressing the clinical significance of mutated genes with highly frequency for bladder cancer.

Single markers have been used clinically for many years, but there are still shortcomings such as low sensitivity and specificity. So far, numerous studies have demonstrated that combination markers could improve sensitivity and specificity and better predict the prognosis of patients [33, 34]. Furthermore, given the heterogeneity of bladder cancer, it is unlikely to identify its progression by a single marker. This underlines the importance of biomarker combinations [35]. In this study, we analyzed the whole mutation profile and identified 23 mutated genes which had a clearly correlation with the prognosis of bladder patients. Furthermore, we found this mutation signature has high accuracy by time-dependent ROC analysis and was an independent indicator by multivariable Cox regression analysis.

Besides, more and more researches have demonstrated that combined molecular with clinical features have higher predictive ability than single biomarker [36, 37]. Thus, nomogram and DCA analysis were established by combining risk score with clinical features involving TNM stage, T stage, N stage, tumor status and lymphovascular invasion. The results suggested that combined parameters performs better in survival prediction for bladder patients and has good clinical application. Our findings provide a superior and helpful prognostic biomarker for bladder patients.

Recent evidence has implicated bladder exist four mutational signatures, including signature 1B, signature 2, signature 5 and signature 13. Both Signature 2 and 13 are mainly featured by C > T as well as C > G and have a close connection with the APOBEC families of cytidine deaminases. In our study, we also observed alteration are mainly featured by C > T together with C > G. APOBEC enzymes usually play essential role in innate immune responses, embracing target retroviruses, demonstrating that there are interrelated among immunity mutagenesis, and viral infection in the progression of cancer development [38]. Based on the similarity of the sequence background of cytosine mutations caused by APOBEC enzymes in experimental systems, APOBEC1, APOBEC3A and APOBEC3B appear to play a greater role in human cancer than other members of the family. Hence, we investigated the expression of APOBEC1, APOBEC3A and APOBEC3B and their expression based on clinical characteristics in bladder. APOBEC1 and APOBEC3A are not significantly differential expressed in bladder tumor (Additional file 3). And we found APOBEC3B is not only highly expressed in tumor, but also its higher expression is related to advanced TNM stage and higher N stage (Fig. 7). Thus, we further explored relationship between our mutation signature and APOBEC3B and found patients with high risk score have higher APOBEC3B expression, suggesting our signature is a reliable predictor to estimate survival of bladder patients.

It is well known that the epithelial-mesenchymal transition (EMT) process plays a significant role in embryonic development as well as tissue repair and tightly associated with cancer metastasis and invasiveness including bladder cancer [39, 40]. In human, hedgehog homologs was divided into three types including sonic hedgehog (Shh), desert hedgehog (Dhh) and Indian hedgehog (Ihh) [41]. Shh is currently the best studied of these three homologs, [42]. Previous study suggested that sonic hedgehog (Shh) is a crucial element for the development and progression of bladder cancer [43]. The PI3K-AKT signaling pathway, a well-known cancer-related signaling pathway, has close relationships with cell growth, cell cycle, survival as well as metastasis in various cancer including bladder cancer [44, 45]. Therefore, it is highly possible that these selected genes may influence the bladder progression through these pathways. Besides, interferon is a cytokine produced by monocytes and lymphocytes. It is divided into three types: α, β and γ. Many studies have shown that interferon (especially α-interferon and γ-interferon) has an obvious anti-cell proliferation effect [46]. Therefore, low-risk patients have a better prognosis and reduce mortality through these immune response and inflammatory response.

It has been reported that some of the screened genes play an essential role be in the tumorigenesis, including bladder cancer. For example, NCOR1 gene was found to be high frequently mutation in bladder cancer. In our study, NCOR1 is not only a highly mutated gene in bladder cancer but also associated with better prognosis of patients. Another candidate, bladder patients with BRCA2 gene mutation have a better prognosis [47]. In our study, we also observed mutated BRCA2 exerts a protective role. However, when combined with other mutated genes, it played a risky role for bladder patients. As for the rest of mutated genes, such as AHNAK2, RYR2, DCHS2, USH2A, RP1L1 and so on were found to be mutated in various cancer and diseases [48–55]. Although the effect of these genes' mutation on bladder cancer were previously unclear, our findings indicated that these genes were associated with patients' survival and deserve further investigations in bladder cancer.

In conclusion, we established a mutation signature related to clinical outcome of bladder patients. Furthermore, this signature is a novel independent indicator and high accuracy for survival prediction. More importantly, we perform a predictive model to effectively predict clinical outcome of bladder patients through integrating risk score with clinical characteristics.

Ethics approval and consent to participate

The data obtained in this study were rooted mainly in the UCSC and TCGA databases, GEPIA and UALCAN website, which are open access.

Consent for publication

Not applicable.

Competing interests

The authors declare that they have no competing interests.

Acknowledgments

Not applicable.

Finding

This work was supported by National Natural Science Foundation of China and Liaoning joint fund key program (No.U1608281), Major Special S&T Projects in Liaoning Province (2019JH1/10300005), Key R&D Guidance Plan Projects in Liaoning Province (2019JH8/10300011), Shenyang S&T Projects (19-109-4-09), and Liaoning Revitalization Talents Program (XLYC1807201).

Availability of data and material

The gene mutation information can be download from UCSC Xena (https://tcga.xenahubs.net). The gene expression data and corresponding clinical information can be download from TCGA database (http://cancergenome.nih.gov). The partial data analysis in this study from GEPIA (http://gepia.cancer-pku.cn) and UALCAN (http://ualcan.path.uab.edu/index.html).

Code availability

Note applicable.

Author contribution statement

YYZ, RM, MH and MW designed the study. YYZ, RM and FL performed research. LZ, XL, PS, YF, YFZ and ML analyzed the data. YYZ and RM wrote the first draft of the paper, and YYZ and MH edited the manuscript.

Siegel RL, Miller KD, Jemal A. Cancer statistics, 2019. Cancer J Clin. 2019;69(1):7–34.
Zhang S, Yu YH, Zhang Y, Qu W, Li J. Radiotherapy in muscle-invasive bladder cancer: the latest research progress and clinical application. American journal of cancer research. 2015;5(2):854–68.
Babjuk M, Burger M, Comperat EM, Gontero P, Mostafid AH, Palou J, van Rhijn BWG, Roupret M, Shariat SF, Sylvester R, et al: European Association of Urology Guidelines on Non-muscle-invasive Bladder Cancer (TaT1 and Carcinoma In Situ) – 2019 Update. European urology 2019.
Burns MB, Temiz NA, Harris RS. Evidence for APOBEC3B mutagenesis in multiple human cancers. Nat Genet. 2013;45(9):977–83.
Vousden KH, Lu X. Live or let die: the cell's response to p53. Nature reviews Cancer. 2002;2(8):594–604.
Therkildsen C, Bergmann TK, Henrichsen-Schnack T, Ladelund S, Nilbert M. The predictive value of KRAS, NRAS, BRAF, PIK3CA and PTEN for anti-EGFR treatment in metastatic colorectal cancer: A systematic review and meta-analysis. Acta oncologica (Stockholm Sweden). 2014;53(7):852–64.
Akagi Y, Tachibana T. KRAS mutations in tongue squamous cell carcinoma. 2019:1–5.
Miyake N, Chikumi H, Yamaguchi K, Takata M, Takata M, Okada K, Kitaura T, Nakamoto M, Yamasaki A. Effect of Cetuximab and EGFR Small Interfering RNA Combination Treatment in NSCLC Cell Lines with Wild Type EGFR and Use of KRAS as a Possible Biomarker for Treatment Responsiveness. Yonago acta medica. 2019;62(1):85–93.
Goebell PJ, Knowles MA. Bladder cancer or bladder cancers? Genetically distinct malignant conditions of the urothelium. Urol Oncol. 2010;28(4):409–28.
Cancer Genome Atlas Research N. Comprehensive molecular characterization of urothelial bladder carcinoma. Nature. 2014;507(7492):315–22.
Gui Y, Guo G, Huang Y, Hu X, Tang A, Gao S, Wu R, Chen C, Li X, Zhou L, et al. Frequent mutations of chromatin remodeling genes in transitional cell carcinoma of the bladder. Nat Genet. 2011;43(9):875–8.
Rodriguez Pena MDC, Tregnago AC, Eich ML, Springer S, Wang Y, Taheri D, Ertoy D, Fujita K, Bezerra SM, Cunha IW, et al. Spectrum of genetic mutations in de novo PUNLMP of the urinary bladder. Virchows Archiv: an international journal of pathology. 2017;471(6):761–7.
Wu X, Ye Y, Kiemeney LA, Sulem P, Rafnar T, Matullo G, Seminara D, Yoshida T, Saeki N, Andrew AS, et al. Genetic variation in the prostate stem cell antigen gene PSCA confers susceptibility to urinary bladder cancer. Nat Genet. 2009;41(9):991–5.
Kiemeney LA, Sulem P, Besenbacher S, Vermeulen SH, Sigurdsson A, Thorleifsson G, Gudbjartsson DF, Stacey SN, Gudmundsson J, Zanon C, et al. A sequence variant at 4p16.3 confers susceptibility to urinary bladder cancer. Nat Genet. 2010;42(5):415–9.
Figueroa JD, Ye Y, Siddiq A, Garcia-Closas M, Chatterjee N, Prokunina-Olsson L, Cortessis VK, Kooperberg C, Cussenot O, Benhamou S, et al. Genome-wide association study identifies multiple loci associated with bladder cancer risk. Human molecular genetics. 2014;23(5):1387–98.
Song BN, Kim SK, Chu IS. Bioinformatic identification of prognostic signature defined by copy number alteration and expression of CCNE1 in non-muscle invasive bladder cancer. Exp Mol Med. 2017;49(1):e282.
Leao R, Lee D, Figueiredo A, Hermanns T, Wild P, Komosa M, Lau I, Mistry M, Nunes NM, Price AJ, et al. Combined genetic and epigenetic alterations of the TERT promoter affect clinical and biological behavior of bladder cancer. International journal of cancer. 2019;144(7):1676–84.
Qiu JG, Shi DY, Liu X, Zheng XX, Wang L, Li Q. Chromatin-regulatory genes served as potential therapeutic targets for patients with urothelial bladder carcinoma. Journal of cellular physiology. 2019;234(5):6976–82.
Williams SB, Huo J, Chu Y, Baillargeon JG, Daskivich T, Kuo YF, Kosarek CD, Kim SP, Orihuela E, Tyler DS, et al: Cancer and All-cause Mortality in Bladder Cancer Patients Undergoing Radical Cystectomy: Development and Validation of a Nomogram for Treatment Decision-making. Urology 2017, 110:76–83.
Kerr KF, Brown MD, Zhu K, Janes H. Assessing the Clinical Impact of Risk Prediction Models With Decision Curves: Guidance for Correct Interpretation and Appropriate Use. Journal of clinical oncology: official journal of the American Society of Clinical Oncology. 2016;34(21):2534–40.
Subramanian A, Tamayo P, Mootha VK, Mukherjee S, Ebert BL, Gillette MA, Paulovich A, Pomeroy SL, Golub TR, Lander ES, et al. Gene set enrichment analysis: a knowledge-based approach for interpreting genome-wide expression profiles. Proc Natl Acad Sci USA. 2005;102(43):15545–50.
Gao J, Aksoy BA, Dogrusoz U, Dresdner G, Gross B, Sumer SO, Sun Y, Jacobsen A, Sinha R, Larsson E, et al. Integrative analysis of complex cancer genomics and clinical profiles using the cBioPortal. Sci Signal. 2013;6(269):pl1.
Tang Z, Li C, Kang B, Gao G, Li C, Zhang Z. GEPIA: a web server for cancer and normal gene expression profiling and interactive analyses. Nucleic acids research. 2017;45(W1):W98-w102.
Chandrashekar DS, Bashel B, Balasubramanya SAH, Creighton CJ, Ponce-Rodriguez I, Chakravarthi B, Varambally S: UALCAN: A Portal for Facilitating Tumor Subgroup Gene Expression and Survival Analyses. Neoplasia (New York, NY) 2017, 19(8):649–658.
Liu K, Newbury PA. Evaluating cell lines as models for metastatic breast cancer through integrative analysis of genomic data. 2019, 10(1):2138.
Zhou H, Tang K, Xiao H, Zeng J, Guan W, Guo X, Xu H, Ye Z. A panel of eight-miRNA signature as a potential biomarker for predicting survival in bladder cancer. Journal of experimental clinical cancer research: CR. 2015;34:53.
Zou J, Wang C, Ma X, Wang E, Peng G. APOBEC3B, a molecular driver of mutagenesis in human cancers. Cell bioscience. 2017;7:29.
Burns MB, Lackey L, Carpenter MA, Rathore A, Land AM, Leonard B, Refsland EW, Kotandeniya D, Tretyakova N, Nikas JB, et al. APOBEC3B is an enzymatic source of mutation in breast cancer. Nature. 2013;494(7437):366–70.
Leonard B, Hart SN, Burns MB, Carpenter MA, Temiz NA, Rathore A, Vogel RI, Nikas JB, Law EK, Brown WL, et al. APOBEC3B Upregulation and Genomic Mutation Patterns in Serous Ovarian Carcinoma. Can Res. 2013;73(24):7222–31.
Wang S, Jia M, He Z, Liu XS. APOBEC3B and APOBEC mutational signature as potential predictive markers for immunotherapy response in non-small cell lung cancer. Oncogene. 2018;37(29):3924–36.
Siegel RL, Miller KD, Jemal A. Cancer statistics, 2019. CA Cancer J Clin. 2019;69(1):7–34.
Hahn WC, Weinberg RA. Rules for making human tumor cells. N Engl J Med. 2002;347(20):1593–603.
Wild N, Andres H, Rollinger W, Krause F, Dilba P, Tacke M, Karl J. A combination of serum markers for the early detection of colorectal cancer. Clinical cancer research: an official journal of the American Association for Cancer Research. 2010;16(24):6111–21.
Tamhane M, Cabrera-Ghayouri S, Abelian G, Viswanath V. Review of Biomarkers in Ocular Matrices: Challenges and Opportunities. Pharmaceutical research. 2019;36(3):40.
van der Heijden AG, Mengual L, Lozano JJ, Ingelmo-Torres M, Ribal MJ, Fernandez PL, Oosterwijk E, Schalken JA, Alcaraz A, Witjes JA. A five-gene expression signature to predict progression in T1G3 bladder cancer. European journal of cancer. 2016;64:127–36.
Zhu X, Tian X, Yu C, Shen C, Yan T, Hong J, Wang Z, Fang JY, Chen H. A long non-coding RNA signature to improve prognosis prediction of gastric cancer. Mol Cancer. 2016;15(1):60.
Qu L, Wang ZL, Chen Q, Li YM, He HW, Hsieh JJ, Xue S, Wu ZJ, Liu B, Tang H, et al. Prognostic Value of a Long Non-coding RNA Signature in Localized Clear Cell Renal Cell Carcinoma. European urology. 2018;74(6):756–63.
Kuong KJ, Loeb LA. APOBEC3B mutagenesis in cancer. Nat Genet. 2013;45(9):964–5.
Gonzalez DM, Medici D. Signaling mechanisms of the epithelial-mesenchymal transition. Sci Signal. 2014;7(344):re8.
Lei Y, Yang L, Hongwei J, Hongyuan Y, Tao L. TIS111D can affect bladder cancer cells by regulating epithelial-mesenchymal transition. Life sciences 2019:116832.
Ingham PW, McMahon AP. Hedgehog signaling in animal development: paradigms and principles. Genes Dev. 2001;15(23):3059–87.
Gonnissen A, Isebaert S, Haustermans K. Hedgehog signaling in prostate cancer and its therapeutic implication. Int J Mol Sci. 2013;14(7):13979–4007.
Syed IS, Pedram A, Farhat WA. Role of Sonic Hedgehog (Shh) Signaling in Bladder Cancer Stemness and Tumorigenesis. Curr Urol Rep. 2016;17(2):11.
Engelman JA, Luo J, Cantley LC. The evolution of phosphatidylinositol 3-kinases as regulators of growth and metabolism. Nat Rev Genet. 2006;7(8):606–19.
Sathe A, Nawroth R. Targeting the PI3K/AKT/mTOR Pathway in Bladder Cancer. Methods in molecular biology. 2018;1655:335–50.
Bandurska K, Krol I, Myga-Nowak M. [Interferons: between structure and function]. Postepy higieny i medycyny doswiadczalnej (Online). 2014;68:428–40.
Kuang S, Li H, Feng J, Xu S, Le Y. Correlation of BRCA2 gene mutation and prognosis as well as variant genes in invasive urothelial carcinoma of the bladder. Cancer Biomark A. 2019;25(2):203–12.
Wolff RK, Hoffman MD, Wolff EC, Herrick JS, Sakoda LC, Samowitz WS, Slattery ML. Mutation analysis of adenomas and carcinomas of the colon: Early and late drivers. Genes Chromosom Cancer. 2018;57(7):366–76.
Yu XJ, Yang MJ, Zhou B, Wang GZ, Huang YC, Wu LC, Cheng X, Wen ZS, Huang JY, Zhang YD, et al. Characterization of Somatic Mutations in Air Pollution-Related Lung Cancer. EBioMedicine. 2015;2(6):583–90.
Monies D, Abouelhoda M, AlSayed M, Alhassnan Z, Alotaibi M, Kayyali H, Al-Owain M, Shah A, Rahbeeni Z, Al-Muhaizea MA, et al. The landscape of genetic diseases in Saudi Arabia based on the first 1000 diagnostic panels and exomes. Human genetics. 2017;136(8):921–39.
An CH, Je EM, Yoo NJ, Lee SH. Frameshift mutations of cadherin genes DCHS2, CDH10 and CDH24 genes in gastric and colorectal cancers with high microsatellite instability. Pathology oncology research: POR. 2015;21(1):181–5.
Pearsall N, Bhattacharya G, Wisecarver J, Adams J, Cosgrove D, Kimberling W. Usherin expression is highly conserved in mouse and human tissues. Hearing research. 2002;174(1–2):55–63.
Piermarocchi S, Segato T, Leon A, Colavito D, Miotto S. Occult macular dystrophy in an Italian family carrying a mutation in the RP1L1 gene. Mol Med Rep. 2016;13(3):2308–12.
Zhou S, Wang F, Dou Y, Zhou J, Hao G, Xu C. A novel FBN2 mutation cosegregates with congenital contractural arachnodactyly in a five-generation Chinese family. 2018, 6(8):1612–1617.
Vinci M, Fchera M, Antonino Musumeci S, Cali F, Aurelio Vitello G. Novel c.C2254T (p.Q752*) mutation in ZFYVE26 (SPG15) gene in a patient with hereditary spastic paraparesis. Journal of genetics. 2018;97(5):1469–72.

Table 1. Patients' clinical characteristics in TCGA database

Characteristics	Total (n=403)	Characteristics	Total (n=403)
Gender		M stage
Female	104	M0	193
male	299/403	M1	10
Age		Race
<69	198	Non-White	67
≥69	205	White	318
TNM stage		Tumor status
I+II	132	Tumor free	233
III+IV	269	With tumor	133
T stage		Lymphovascular invasion
T1+T2	120	No	129
T3+T4	249	Yes	150
N stage		Histological subtype
N0	238	Non-papillary	267
N1+2+3	127	Papillary	131

Table 2. Univariable Cox regression analysis to access the prognostic value of mutated genes

Gene symbol	HR^a	P value^b	Gene symbol	HR^a	P value^b
AHNAK2	0.4586	0.0043	HEATR5B	0.2909	0.0354
BIRC6	0.3333	0.0050	WDR96	0.2230	0.0359
NAV3	0.2449	0.0059	NAV2	0.2238	0.0361
USH2A	0.3159	0.0063	AUTS2	0.2240	0.0362
LRP1B	0.4053	0.0067	KIF1B	0.2248	0.0367
RYR2	0.3793	0.0085	ERCC2	0.3462	0.0376
RP1L1	0.1646	0.0120	SCN11A	0.2271	0.0379
SRRM2	0.2811	0.0129	LRRK2	0.4427	0.0379
GPR98	0.3199	0.0130	IGSF10	0.2965	0.0381
DYNC1H1	0.3246	0.0142	PRKDC	0.2975	0.0389
DCHS2	0.2432	0.0158	ABCA12	0.2981	0.0390
FBN2	0.1818	0.0170	CASC5	0.2293	0.0392
FANCM	0.2477	0.0176	ZNFX1	0.2303	0.0398
WNK1	0.1834	0.0176	FREM2	0.4472	0.0401
CACNA1D	0.1851	0.0182	TECTA	0.3001	0.0402
MROH2B	0.1874	0.0192	FLNC	0.2313	0.0403
VPS13B	0.4245	0.0202	GTF3C1	0.2314	0.0405
MED1	0.0974	0.0206	RANBP2	0.1281	0.0409
EP300	0.3482	0.0214	XIRP2	0.5457	0.0412
BAZ2B	0.1993	0.0239	F8	0.3026	0.0413
NCOR1	0.2031	0.0256	SCAF4	0.2348	0.0434
DST	0.4916	0.0261	POLE	0.3062	0.0434
HSPG2	0.4438	0.0274	LAMA3	0.3980	0.0446
COL11A1	0.2746	0.0275	BRCA2	0.4327	0.0466
ZFYVE26	1.9460	0.0305	ASTN1	0.2423	0.0476
ANK1	0.2183	0.0331	MYO18B	0.1368	0.0477
CNOT1	0.2190	0.0335	ARFGEF1	0.2446	0.0486
MGA	0.2193	0.0342	COL5A1	2.1655	0.0487
AKAP13	0.4367	0.0348

Abbreviation: HR, hazard ratio.

^aValues >1.0 indicate that expression is positively associated with poor survival.

^bLikelihood ratio test P value.

Table 3. The HR and P value of mutated gene

Gene symbol	Coefficient^a	HR^b	P value^c
AHNAK2	-0.6344	0.5302	0.0273
RYR2	-0.5828	0.5583	0.1346
USH2A	-1.0395	0.3536	0.0171
DCHS2	-0.9495	0.3869	0.1139
BRCA2	1.0108	2.7479	0.0421
RP1L1	-1.2290	0.2926	0.0976
FBN2	-1.2238	0.2941	0.0934
ZFYVE26	1.7481	5.7436	0.0000
ANK1	-1.5427	0.2138	0.0351
NCOR1	-1.4791	0.2278	0.0401
F8	-1.1631	0.3125	0.0543
WNK1	-1.1986	0.3016	0.1009
IGSF10	-0.8949	0.4086	0.1365
COL11A1	-0.7952	0.4515	0.1954
RANBP2	-1.3281	0.2650	0.1891
MED1	-1.8627	0.1553	0.0667
FANCM	-2.0523	0.1284	0.0020
MYO18B	-1.6193	0.1980	0.1122
ZNFX1	-1.3542	0.2582	0.0829
MROH2B	-1.0805	0.3394	0.1419
AUTS2	-1.5807	0.2058	0.0403
CNOT1	-1.5486	0.2125	0.0342
COL5A1	0.6871	1.9880	0.0866

Abbreviation: HR, hazard ratio.

^aValues >0.0 and ^bValues >1.0 indicate that expression is positively associated with poor survival.

^cLikelihood ratio test P value.

Fig S1 The interactions between these mutated genes. (a) The alteration of these mutated genes from cBioPortal database. (b) The interactions between these mutated genes generated from cBioPortal database.

Fig S2 The Kaplan-Meier analysis of clinical characteristics and correlation analysis between risk score and clinical characteristics. Kaplan-Meier analysis of age (a), gender (b), race (c), M stage (d) and histological subtype (e). The distribution of risk score in patients stratified by age (f), gender (g), race (h), M stage (i) and histological subtype (j).

Fig S3 The differentially expression of APOBEC1 and APOBEC3A in normal and tumor tissues and in different TNM stage.

Analysis of genome-wide mutation profile and establishment of risk signature for prognosis of bladder cancer

Status:

Version 1

Abstract

Figures

1 Background

2 Materials And Methods

2.1 Patient Samples

2.2 Mutation gene collection and signature construction

2.3 Predictive nomogram of the mutation signature

2.4 Functional analysis

2.5 Statistical Analysis

3 Results

3.1 Somatic mutation profile of bladder patients

3.2 Identification of mutation genes related to the survival of bladder patients

3.3 The mutation signature is an independent prognostic indicator

3.4 Clinical Application of the mutation signature

3.5 The mutation signature are better than other biomarkers in survival prediction

3.6 The higher risk score was related to higher expression of APOBEC3B

3.7 Identification of signaling pathways correlated with the mutation signature

4 Discussion

5 Conclusions

Declarations

References

Tables

Supplementary file

Supplementary Files

Status:

Version 1