Abstract Background Bladder cancer (BLCA) is one of the most frequently-diagnosed tumors globally. Disulfidptosis represents a critical framework for cell death mechanism in cancer therapy. Our study constructed a predictive model utilizing disulfidptosis-related lncRNAs (DRLs) to provide value in evaluating diagnosis, drug sensibility, and prognosis of BLCA patients. Methods The study data of BLCA patients retrieved from TCGA-BLCA database. Cox and LASSO regression analysis were used to identify DRLs. Kaplan–Meier survival analysis, ROC curve, and nomograms were constructed to assess and forecast survival events. GSEA were performed to illustrate relevant enrichments results. Tumor mutation burden (TMB), immune status, and drug sensitivity were assessed. Muscle invasive bladder cancer (MIBC) tumor and tumor-adjacent normal tissues samples were collected in our department to validate the DRLs expression levels by RT-PCR. Results Overall, nine DRLs ([28]AL590428.1, LSAMP-AS1, LINC01184, LINC-PINT, [29]AC023825.2, [30]AC010331.1, [31]AC009716.1, [32]AC104785.1, [33]AC008764.6) were identified. These DRLs were used to calculate risk scores and create a prognostic model. ROC revealed higher diagnostic efficiency of the model than other clinical characteristics. Nomogram was constructed using the risk scores, age, and tumor stage, which showed excellent predictive power and was verified by calibration graph. BLCA patients were further classified into high-risk group and low-risk group using median risk score as the cut-off value. The high-risk group showed lesser TMB levels and developed worse prognosis. GSEA of the high-risk group identified pathways associated with BLCA progression such as WNT signaling pathway. Immune cells including CD4^+ and CD8^+ T cells, and immune-related function like T cell co-stimulation also showed remarkable differences between two risk groups. Furthermore, IC50 values of twelve drugs such as Sorafenib, Nilotinib, and Navitoclax were significantly higher in the high-risk group. RT-PCR results revealed that 9 DRLs expression levels were statistically significant between tumor tissues samples and tumor-adjacent normal tissues samples. The expression trends of these DRLs in clinical tissues samples were the same as the findings in TCGA dataset. Conclusion Based on this study, it would be advisable to identify the key DRLs with potential prognostic value in BLCA to enhance the evaluation of clinical outcomes in this context. Keywords: Bladder cancer, Disulfidptosis, lncRNA, Signature, Immunotherapy Introduction Bladder cancer (BLCA) is a frequent urogenital carcinoma and ranks as the tenth most frequently-diagnosed cancer globally, with an increasing rate of new cases yearly [[34]1]. About 25% of patients exhibit muscle-invasive disease during diagnosis [[35]2]. The current standard treatments include surgery and adjuvant treatments. Owing to increased rates of relapse and metastasis, prognostic outcomes of BLCA cases are usually unfavorable [[36]3]. A recent remarkable study [[37]4] indicated that actin cytoskeleton proteins were specifically vulnerable to disulfide stress resulted from extreme accumulation of disulfide molecules in the cell. Abnormal and unrepaired disulfide bonding among actin cytoskeleton proteins could cause actin network collapse and cell death. The authors of the study termed this type of cell death as disulfidptosis, because it did not fall under any frequent forms of the regulated cell death. Conceptualization of disulfidptosis represents a crucial configuration for farther comprehension and target of the individual cell death progress in tumor therapy [[38]5]. Hence, identification of disulfidptosis-associated molecule characteristics could enhance our comprehension of disease progression, facilitate the personalized therapies, and predict treatment response in BLCA. To enhance inchoate treatment and improve rates of BLCA survivors, it should be imperative to identify novel molecule signatures for predicting cases prognosis and medicine sensibility. Intriguingly, many investigations have reported potential associations between long non-coding RNAs (lncRNAs) and programmed cell death in cancer cells [[39]6, [40]7]. Whereas, potential impact of disulfidptosis-related lncRNAs (DRLs) on predicting BLCA prognosis has not been studied yet, and their utility as an independent prognostic factor remains unclear. Hence, we aim to analyze the data for recognizing DRLs signature to predict BLCA patients’ prognosis. We created a predictive model utilizing DRLs, which provides value in evaluating clinical outcomes of BLCA cases. We further carried out some significant analysis such as gene set enrichment analysis (GSEA), tumor mutation burden (TMB), immune status, and medicine sensitivity analysis for illustrating underlying mechanisms between DRLs and BLCA patients. Furthermore, our research set up a risk score model based on DRLs in BLCA cases, that could explore disulfidptosis mechanism in cancer, assess patient prognosis, and provide individual management schemes. Material and methods BLCA cases and datasets The research data consisted of 412 BLCA cases and retrieved from The Cancer Genome Atlas (TCGA) database ([41]https://portal.gdc.cancer.gov/). Datasets pertaining to case information and clinicopathologic records were also procured from TCGA database. The entire transcriptome sequencing data was standardized utilizing the FPKM approach. Seven BLCA patients had to be eliminated because of fragmentary information. Data on gene mutations for BLCA retrieved through TCGA database. After downloading initial data, count and location of mutations in every gene of every sample was analyzed and utilized to generate the matrix. Ten disulfidptosis-related genes (DRGs) were obtained from a previously published article [[42]4], including GYS1, NDUFS1, OXSM, LRPPRC, NDUFA11, NUBPL, NCKAP1, RPN1, SLC3A2, and SLC7A11. Identification of DRLs Correlations between DRGs and lncRNAs were computed through R software "limma" package, with the value of p < 0.001 and |r|> 0.4 utilized as thresholds. Based on these conditions, we identified 306 DRLs in BLCA. To screen DRLs linked to overall survival (OS), a total of 405 BLCA specimens were randomly divided into training and test cohorts using the “caret” package. Then, univariate Cox regression analysis was performed using the training cohort. For preventing model overfitting, least absolute shrinkage and selection operator (LASSO) regression analysis was employed by "glmnet" package for filtering lncRNAs associated with DRGs in BLCA. Subsequently, the screened lncRNAs was utilized for multivariate Cox regression. Eventually, 9 DRLs were recognized and applied for calculating risk scores and creating a prognostic model. This involved calculating individual risk scores by multiplying each lncRNA’s expression level by its corresponding regression coefficient. BLCA patients were then classified into high-risk group and low-risk group on the basis of risk scores calculated with expression levels of 9 DRLs. The cutoff value of high and low risk group was the median risk scores. Subsequently, we utilized "survival" and "survminer" packages to conduct Kaplan–Meier survival curve for the purpose of contrasting OS in high- and low-risk groups. To assess ability of novel DRLs model in forecasting OS, we employed "timeROC" package. For gauging practicality of DRLs model, BLCA cases were categorized as distinct clusters based on the patients' age and tumor stage. We subsequently employed Kaplan–Meier survival analysis to scrutinize its precision in forecasting prognosis of distinct clusters. Nomograms for prognosis To improve the assessment of clinical outcomes, risk scores calculated by DRLs model had been integrated with clinical characteristics, such as age and tumor staging. The "rms," "regplot," and "survival" packages were utilized to create nomogram that could predict OS at 1 year, 3 years, and 5 years. Accuracy and dependability of constructed nomogram were scrutinized by Calibration plots. Principle component analysis The "limma" and "Scatterplot3d" packages were utilized to perform principal component analysis (PCA), which could display the distinct distribution of different risk groups. By utilizing PCA, we compared expression distribution of whole gene, DRGs, DRLs, and the modeling lncRNAs, respectively. Functional enrichment and immune-related analysis The "limma" package was employed to filter differentially expressed genes (DEGs) of two risk groups, using criteria of |log2 FC|≥ 1 and FDR < 0.05. The "clusterProfiler" package has been utilized to perform GO (Gene Ontology) and KEGG (Kyoto Encyclopedia of Genes and Genomes) enrichment analysis. Gene set enrichment analysis (GSEA) using the curated gene set (c2.cp.kegg.Hs.symbols.gmt) was also carried out. Furthermore, status of immunizing cell and immunizing function were evaluated on the basis of DEGs. The "maftools" package had been employed to assess gene mutation counts and sites between low- and high-risk groups, with visualization of results. Comparison of TMB was performed through "ggpubr" package. The "survminer" package was used to assess survival outcomes for BLCA cases having diverse TMB levels. Drugs sensitivity analysis For assessing capability of DRLs model on drug treatment response in BLCA, we employed "oncoPredict" package to compute half-maximal inhibitory concentration (IC50) values of frequent drugs for managing BLCA in clinic. Subsequently, IC50 values were compared between two risk groups. RT-PCR validation Ten pairs of tumor and tumor-adjacent normal tissues samples were extracted from muscle invasive bladder cancer (MIBC) cases who performed radical cystectomy in our hospital. Samples were cut and instantly put in liquid nitrogen. The study was approved by the Ethics Committee of The First Affiliated Hospital of Chongqing Medical University (Approval Number: 2020–155). Written informed consent was obtained from all cases and this research was performed based on the Declaration of Helsinki. Total RNA from MIBC and tumor-adjacent normal tissues samples were extracted by TRIzol reagent (Invitrogen) as per the manufacturer's instructions. The cDNA was amplified by the first-strand cDNA Synthesis Super Mix for qPCR (Yeasen). Quantitative real-time polymerase chain reaction (RT-PCR) was completed using a real-time detector (MA-6000; Molarray). GAPDH was regarded as internal control gene. The primers for 9 DRLs were present in Table [43]1. Table 1. Primers for disulfidptosis-related lncRNA LncRNA Forward Reverse [44]AL590428.1 ATCTGCTCAGCATTCCTGTGTTT GGGATCTGGGAAAAGTAAGAAATTG [45]AC023825.2 TCCGGCAGGAAAGAGCATT ATCGGCATTGGCGAAGAA [46]AC010331.1 TGCCCTGAGAGAGCAAACGT TGGAACAGGCTCTCCAGTATTGT [47]AC009716.1 GGCCATTCAGAGGTGAAATCTG TTTACACCGCCATAAGCATGTC [48]AC104785.1 AAAACCCAGCTGAAGAGCAATT CCAGCAATCCTCAGCCATTC [49]AC008764.6 GAGAGGCAGCACATTTCATTCC CCACACCCTGCTCTAAGAGTGA LSAMP-AS1 AGAAGACGGTGATTTCTGCATT TCGTTTGTGCTTAGAGACCAGTAG LINC01184 CTGCATTTAGTGAGATAAGCCACA AGAATGTATTCCAGGCTCAAGTTC LINC-PINT GAGGTCATATCTCCGTACCTCACT TTACGTTCCTCCAGTAACTGTTGA GAPDH GGCAAGTTCAACGGCACAG CGCCAGTAGACTCCACGACA [50]Open in a new tab Statistical analysis The R software (version 4.2.2) was applied for organizing and analyzing all data. For variables exhibiting normal distribution, differences between two groups were assessed using an unpaired two tailed t-test, while non-normally distributed variables were compared with the Wilcoxon signed-rank test. The value of two-sided p < 0.05 was statistically significant. Results Identification of DRLs The flowchart of present investigation was depicted in Fig. [51]1. Overall, 306 DRLs was identified in BC datasets (Fig. [52]2A). DRLs marker was explored through following steps. Fig. 1. [53]Fig. 1 [54]Open in a new tab The flowchart of the present study Fig. 2. [55]Fig. 2 [56]Open in a new tab Identification of the DRLs signature in bladder cancer. A The correlation between disulfidptosis-related genes and differentially expressed lncRNAs. B Forest plot of univariate Cox regression analysis for the 26 DRLs. C, D Lasso regression was performed to calculate the (C) coefficients and (D) minimum criteria. E Correlation heatmap of disulfidptosis-related genes with 9 DRLs markers in prognostic signature. F Sankey diagrams of visualizing protective and risk factors in 9 DRLs signature DRLs model Initially, 405 BLCA patients were divided into a train cluster and a test cluster at random. Clinical and pathological features of both clusters were comparable (Table [57]2). Secondly, 26 DRLs associated with OS were identified using univariate Cox analysis (Fig. [58]2B). These 26 DRLs underwent further analysis via LASSO regression, resulting in the selection of 18 DRLs (Fig. [59]2C, D). Then, a multivariate Cox regression analysis was conducted to find markers with the lowest Akaike information criterion (AIC) value, which ultimately led to the generation of a prognostic model comprising 9 DRLs (Table [60]3). A heatmap was generated to depict the correlations between these DRLs and DRGs (Fig. [61]2E). Among the 9 DRLs, 3 acted as protective factors ([62]AL590428.1, LSAMP-AS1, LINC01184), and the other 6 acted as dangerous factors (LINC-PINT, [63]AC023825.2, [64]AC010331.1, [65]AC009716.1, [66]AC104785.1, [67]AC008764.6) (Fig. [68]2F). Table 2. Clinical characteristics of bladder cancer patients in different groups Covariates Total Test group Train group p-value Age, years 0.8076 ≤ 65 159(39.26%) 81(40.1%) 78(38.42%) > 65 246(60.74%) 121(59.9%) 125(61.58%) Gender 0.7123 Female 106(26.17%) 55(27.23%) 51(25.12%) Male 299(73.83%) 147(72.77%) 152(74.88%) Grade 0.9813 High Grade 381(94.07%) 189(93.56%) 192(94.58%) Low Grade 21(5.19%) 11(5.45%) 10(4.93%) Unknow 3(0.74%) 2(0.99%) 1(0.49%) Stage 0.4307 Stage I 2(0.49%) 1(0.5%) 1(0.49%) Stage II 129(31.85%) 72(35.64%) 57(28.08%) Stage III 140(34.57%) 67(33.17%) 73(35.96%) Stage IV 132(32.59%) 61(30.2%) 71(34.98%) Unknow 2(0.49%) 1(0.5%) 1(0.49%) T stage 0.3705 T0 1(0.25%) 1(0.5%) 0(0%) T1 3(0.74%) 1(0.5%) 2(0.99%) T2 118(29.14%) 65(32.18%) 53(26.11%) T3 193(47.65%) 87(43.07%) 106(52.22%) T4 57(14.07%) 28(13.86%) 29(14.29%) Unknow 33(8.15%) 20(9.9%) 13(6.4%) M stage 0.2678 M0 195(48.15%) 99(49.01%) 96(47.29%) M1 11(2.72%) 8(3.96%) 3(1.48%) Unknow 199(49.14%) 95(47.03%) 104(51.23%) N stage 0.2083 N0 236(58.27%) 125(61.88%) 111(54.68%) N1 46(11.36%) 26(12.87%) 20(9.85%) N2 75(18.52%) 30(14.85%) 45(22.17%) N3 6(1.48%) 3(1.49%) 3(1.48%) Unknow 42(10.37%) 18(8.91%) 24(11.82%) [69]Open in a new tab Table 3. Nine disulfidptosis-related lncRNA identified by multivariate Cox regression analysis Disulfidptosis-related lncRNA Coefficient LINC-PINT − 0.87371 [70]AC023825.2 − 0.75373 [71]AC010331.1 − 0.36977 [72]AL590428.1 0.670142 LSAMP-AS1 0.287124 [73]AC009716.1 − 0.61749 [74]AC104785.1 − 0.39311 [75]AC008764.6 − 0.34857 LINC01184 1.152618 [76]Open in a new tab These 9 DRLs were applied for calculating risk scores by the formula as follows: (1.1526 × LINC01184)−(0.3486 × [77]AC008764.6)−(0.3931 × [78]AC104785.1 )−(0.6175 × [79]AC009716.1) + (0.2871 × LSAMP-AS1) + (0.6701 × [80]AL59 0428.1)−(0.3698 × [81]AC010331.1)−(0.7537 × [82]AC023825.2)−(0.8737 × L INC-PINT). All cases were classified into high- (n = 207) and low-risk (n = 198) groups using median risk score as the cut-off value. Furthermore, Kaplan–Meier survival analysis indicated cases in high-risk group developed obviously lower progression free survival (PFS) rate and OS (Fig. [83]3A, B). When risk scores rose, dead cases rose and surviving time declined (Fig. [84]3C, D). A heatmap displayed expression distribution and correction of 9 marker DRLs in two risk groups (Fig. [85]3E). Fig. 3. [86]Fig. 3 [87]Open in a new tab Validation of the DRLs prognostic signature in BLCA. Kaplan–Meier survival curves of progression-free survival (A) and overall survival (B) in the high- and low-risk groups of BLCA patients. C Distribution of the risk score in the total data set. D Survival status of the patients in the total data set. E The 9 DRLs expression heatmap in the high- and low-risk groups. F C-index curve of the prognostic signature Regression analysis and survival analysis Univariate Cox regression analysis showed risk score, tumor stage, and age were statistically relevant to BLCA cases’ OS (Fig. [88]4A). Multivariate Cox regression analysis confirmed risk score, tumor stage, and age as independent prognostic factors for OS (Fig. [89]4B). Notably, ROC (Fig. [90]4C) indicated risk score (AUC = 0.696) exhibited superior predictive ability for OS compared to age (AUC = 0.617), gender (AUC = 0.491), grade (AUC = 0.465), and tumor stage (AUC = 0.642), respectively. Risk score was also used to assess OS at 1 year, 3 years, and 5 years, the AUCs of which were 0.644, 0.696, and 0.699, respectively (Fig. [91]4D). Furthermore, the Concordance index also demonstrated that the risk score outperformed clinicopathological variables (Fig. [92]3F). Patients’ age (≤ 70y vs. > 70y) and tumor stage (stage I-II vs. III-IV) were employed for subgroup assessment. Our model can differentiate two risk groups (Fig. [93]5A–D). Thus, cases in high-risk group displayed significantly lower survival probability than those in low-risk group. Collectively, these findings reveal our model possesses significant capability to prognosticate outcomes. Fig. 4. [94]Fig. 4 [95]Open in a new tab Independent prognostic factors for overall survival (OS) and construction of the predictive nomogram based on the DRLs in bladder cancer (BLCA) patients. A, B Forest plots for univariate (A) and multivariate (B) Cox regression analysis. C Receiver operating characteristic (ROC) curves and areas under the curve (AUCs) for the risk score and clinical variables. D ROC curves and AUCs of 1-, 3-, and 5-year OS for the predictive signature. E Nomogram for predicting the OS of BLCA patients at 1, 3, and 5 years based on the risk score and significant clinical variables. F Calibration curves for evaluating the nomogram of OS prediction at 1, 3, and 5 years Fig. 5. [96]Fig. 5 [97]Open in a new tab Survival curve based on subgroup analysis and principal component analysis (PCA) considering the different gene profiles of bladder cancer (BLCA) patients. A, B Survival curve of the high- and low-risk groups in BLCA patients with age ≤ 70 years (A) and age > 70 years (B). C, D Survival curve of the high- and low-risk groups in BLCA patients with tumor stage I–II (C) and stage III–IV (D). PCA of BLCA patients between high- and low-risk groups based on the whole gene module (E), disulfidptosis-related gene module (F), disulfidptosis-related lncRNA module (G), and disulfidptosis-related lncRNA prognostic module (H) A nomogram (Fig. [98]4E) was constructed for measuring survival rates at one year, three years, and five years. Nice conformance between predictable and actual OS outcomes were demonstrated by calibration curves (Fig. [99]4F). Furthermore, the expression distribution of whole genes (Fig. [100]5E), DRGs (Fig. [101]5F), DRLs (Fig. [102]5G), and model lncRNAs (Fig. [103]5H) were subjected to PCA assessment. The PCA outcomes indicated our model effectively distinguished two risk groups. These findings demonstrated the reliable utility of the model constructed by 9 marker DRLs. Function and pathways enrichment analysis Biological process of GO enrichment primarily consisted of epidermis development and epidermal cell differentiation (Fig. [104]6A, B). Cellular component category primarily consisted of cornified envelope, intermediate filament cytoskeleton, and intermediate filament (Fig. [105]6A, B). Molecular function primarily consisted of structural constituent of skin epidermis, peptidase regulator activity, serine hydrolase activity, and endopeptidase regulator activity (Fig. [106]6A, B). KEGG pathway enrichment primarily consisted of staphylococcus aureus infection, estrogen signaling pathway, and retinol metabolism (Fig. [107]6C, D). Fig. 6. [108]Fig. 6 [109]Open in a new tab Enrichment analysis of differentially expressed genes in the low- and high-risk groups. The enrichment items in GO (A, B) and KEGG (C, D) analysis. The gene set enrichment analysis (GSEA) results in the low- (E) and high-risk (F) groups To further assess feasible effect of the model, GSEA was conducted separately. GSEA of low-risk group indicated significant enrichment in the allograft rejection and graft versus host disease pathways (Fig. [110]6E). In contrast, GSEA of high-risk group revealed significant enrichment in the WNT signaling pathway, regulation of actin cytoskeleton, and pathways in cancer (Fig. [111]6F). Immune cell and immune-related function The low-risk group showed a significant increase in B cells naïve, plasma cells, CD8^+ T cells, activated memory CD4^+ T cells, and Tregs, while resting-memory CD4^+ T cells, activated dendritic cells, eosinophils, and neutrophils were significantly upregulated in the high-risk group (Fig. [112]7A). Immune function analysis indicated that Th2 cells and T cell co-stimulation had significantly greater activity in low-risk group (Fig. [113]7B). Fig. 7. [114]Fig. 7 [115]Open in a new tab Immune-related activity and tumor mutation burden (TMB) analysis in the two groups. The differences in tumor-infiltrating immune cells (A) and immune-related function (B). Waterfall plots of TMB in the low-risk (C) and high-risk (D) groups. E Comparison of TMB levels between the low- and high groups. F Kaplan–Meier survival curves for different TMB groups. G Kaplan–Meier survival curves for different TMB and risk groups. H Comparison of IC50 values on anti-tumor drug SCH772984 between the low- and high groups Tumor mutation burden assessment The waterfall plots showed TMB outcomes in low- (Fig. [116]7C) and high-risk groups (Fig. [117]7D). The majority of genes showed less mutation frequencies in cases with high-risk scores, and TMB levels of these cases were obviously lower (Fig. [118]7E). All cases had been classified into high- and low-TMB groups on the basis of median TMB level for evaluating TMB impact on survival outcomes. The OS in cases with high TMB was considerably higher than that in cases with low TMB (Fig. [119]7F). Further assessment was performed by combining risk scores and TMB levels. Patients with low TMB and high-risk scores had the worst OS, whereas those with high TMB and low-risk scores had the best OS (Fig. [120]7G). Drug sensibility analysis The half-maximal inhibitory concentration (IC50), which referred to the concentration of a drug to inhibit 50% of tumor cell proliferation, was applied for evaluating the impact of our model on forecasting the response to anti-tumor drugs. On one hand, SCH772984 had significantly lower IC50 values in the high-risk group (Fig. [121]7H), indicating that this drug could be more beneficial for high-risk cases. On the other hand, the low-risk group had significantly lower IC50 values for 12 drugs, including ABT737, AZD1208, EPZ004777, EPZ5676, GSK343, MIRA-1, MK-2206, PRIMA-1MET, Venetoclax, Sorafenib, Nilotinib, and Navitoclax, suggesting that these drugs could be more beneficial to BLCA patients in the low-risk group (Fig. [122]8A–L). Fig. 8. [123]Fig. 8 [124]Open in a new tab Comparison of anti-tumor drugs sensitivity between the high- and low-risk groups. The half-maximal inhibitory concentration (IC50) values of ABT737 (A), AZD1208 (B), EPZ004777 (C), EPZ5676 (D), GSK343 (E), MIRA-1 (F), MK-2206 (G), Navitoclax (H), Nilotinib (I), PRIMA-1MET (J), Sorafenib (K), and Venetoclax (L) RT-PCR validation in clinical samples Nine DRLs expression levels were analyzed and compared between the tumor samples and normal samples from TCGA-BLCA dataset. Significant differences were found in all 9 DRLs expression levels between two groups (Fig. [125]9A–I). Moreover, we collected MIBC tumor and tumor-adjacent normal tissues samples in our department to validate 9 DRLs expression levels by RT-PCR. RT-PCR results revealed that 9 DRLs expression levels were statistically significant between tumor tissues samples and tumor-adjacent normal tissues samples (Fig. [126]10). The expression trends of these DRLs in clinical tissues samples were the same as the findings in TCGA dataset. Fig. 9. [127]Fig. 9 [128]Open in a new tab Expression levels of nine DRLs markers between the tumor samples and normal samples from TCGA-BLCA dataset. Expression levels of [129]AL590428.1 (A), LSAMP-AS1 (B), LINC01184 (C), LINC-PINT (D), [130]AC023825.2 (E), [131]AC010331.1 (F), [132]AC009716.1 (G), [133]AC104785.1 (H), and [134]AC008764.6 (I) Fig. 10. [135]Fig. 10 [136]Open in a new tab Nine DRLs markers expression levels validated by RT-PCR using tumor tissues samples and tumor-adjacent normal tissues samples collected from our department Discussion BLCA has significant impacts on human health and incurs high expenses for treatment [[137]8]. A novel finding of non-apoptotic cell death, naming disulfidptosis, has been reported recently [[138]4]. LncRNA is associated with onset, advancement, and metastasis of BLCA [[139]9, [140]10]. However, the characteristics and functions of DRLs in BLCA are unknown. Hence, we explored the influence of DRLs on cases prognosis and tumor microenvironment (TME), and constructed a DRLs model for forecasting cases prognosis. Our study established a prognostic signature using 9 DRLs ([141]AL590428.1, LSAMP-AS1, LINC01184, LINC-PINT, [142]AC023825.2, [143]AC010331.1, [144]AC009716.1, [145]AC104785.1, and [146]AC008764.6). Prior research demonstrated that part of these lncRNAs were associated with cancer immunizing response and progression. For instance, [147]AL590428.1 had been found to obviously reduce in human pterygium fibroblasts and strongly associated with cell circle and death [[148]11, [149]12]. LINC-PINT had a potential function in controlling cancer and was closely relevant to cells invasion resistance for kidney, lung, and colon cancers [[150]13]. It could also curb DNA restore and boost the reaction caused by radiation treatment for nasopharyngeal cancer [[151]14]. LINC-PINT could play a crucial role in controlling thyroid cancer through miR-767-5p/TET2 axis, which might be thyroid carcinoma treatment target [[152]15]. Additionally, up-regulated LINC-PINT could restrain BLCA cell diffusion, expansion, and metastasis through miR-155-5p, which was regarded as a potential prognostic marker of BLCA [[153]16]. These 9 lncRNAs were found to form a DRLs prognostic signature model for BLCA. Subsequent analysis disclosed that prognostic accuracy of the model was significantly superior to clinicopathologic parameters. The model exhibited satisfactory reliability to forecast OS at 1 year, 3 years, and 5 years. Enrichment assessment indicated DEGs between two risk groups were primarily enriched in epidermis development, T-cell receptor complex, peptidase regulator activity, and estrogen signaling pathway. KEGG pathway enrichment demonstrated the association of Staphylococcus infection, which was closely associated with the development and progression of bladder cancer. These findings indicate that the abnormal interactions between bioactive molecules and cellular signaling pathways play a critical role in illness advancement and contribute to poor prognosis in clinic [[154]17–[155]19]. Additionally, the high-risk group showed enrichment in the WNT signaling pathway and pathways in cancer, as identified through GSEA assessment. The WNT signaling pathway was relevant to cancer cell expansion and migration in BLCA [[156]20], and the TME of BLCA was also linked to this pathway [[157]21]. Hence, DRLs could influence biological processes in BLCA through these pathways. It is noteworthy that the DRLs prognostic signature model exhibited a strong association with the invasion of immunizing cells in tumors. Assessment in immunizing cells revealed BLCA cases in high-risk group exhibited higher expression of resting-memory CD4^+ T cells, activated dendritic cells, eosinophils, and neutrophils. We further assessed the impact of DRLs on immune functions and analyzed BLCA cases prognosis relevant to immunizing function. The analysis of immune function showed that BLCA cases with high-risk scores had fewer activity in Th2 cells and T cell co-stimulation, which may indicate immunizing resistance [[158]22–[159]24]. While lncRNAs themselves did not encode proteins, they had been shown to be involved in regulating cancer immunotherapy [[160]25] and immune responses [[161]26]. All kinds of immunizing cells in TME involved mutual effect between lncRNAs and tumor immune treatment [[162]27]. This conformed to immunizing functions in our findings. Immunotherapy is now a vital part of treating advanced BLCA [[163]28]. However, its effectiveness is varied and limited in advanced tumor patients [[164]29]. Thus, it’s worthwhile for finding new signature that can forecast immunotherapy response. TMB has been widely recognized as the signature of immunizing effect, and some studies have shown that it has good predictive value in certain subpopulations [[165]30]. In the present study, high-risk group showed lower TMB value, so risking scores and TMB levels synergistically predicted BLCA cases outcomes. The accumulation of somatic mutations promotes the generation of neoantigens, which activate T-cell immunogenicity and enhance T cells ability of eliminating tumor cells [[166]31, [167]32]. High TMB tumors could produce novel antigens for absorbing all kinds of immunizing cells [[168]33, [169]34]. Our results are consistent with previous studies, indicating higher TMB might be associated with better outcomes. Moreover, higher TMB cases could restrain tumor advancement by enhancing T-cell immunogenicity and increasing sensitivity to immune treatment [[170]35]. In summary, BLCA cases with lower risking scores and higher TMB levels might be more efficient by immunizing treatment. Our study highlights the importance of TMB as helpful signature to forecast immunizing treatment effect in BLCA cases. Drug sensibility was a significant obstacle in treating cancers [[171]36]. Several cases could encounter insufficient therapeutic medicine options when trying multiple anti-cancer therapies [[172]37]. The detection for novel medicine was usually challenging, costly, and inconclusive. Computational drug repositioning was an efficient way to discover novel functions for classical medicine [[173]38]. Moreover, IC50 is a vital index to evaluate medicine effect and patient’s treating reaction [[174]39]. Cases having high IC50 values are more likely to have the medicine resistance. In this study, BLCA cases in high-risk group exhibited a higher IC50 value for several medicines than patients in low-risk group. These medicines included ABT737, AZD1208, EPZ004777, EPZ5676, GSK343, MIRA-1, MK-2206, PRIMA-1MET, Venetoclax, Sorafenib, Nilotinib, and Navitoclax. Sorafenib is a multi-kinase inhibiting agent which facilitates apoptosis, alleviates angiogenesis, and inhibits proliferation [[175]40]. It can suppress cancer cell proliferation by restraining Raf-1, B-Raf, and kinase activity in the Ras/Raf/MEK/ERK signaling pathways [[176]41]. Additionally, sorafenib can target PDGFR-β, VEGFR 2, c-KIT, and additional factors for restraining cancer angiogenesis [[177]42]. Venetoclax was a kind of BCL-2 inhibiting agent and it could result in beneficial results in acute myeloid leukemia [[178]43]. Nilotinib is a second-generation tyrosine kinase inhibitor and primarily used in first-line treatment in newly diagnosed chronic myeloid leukemia [[179]44]. Navitoclax was also a kind of BCL-2 inhibiting agent and it is usually used for managing myelofibrosis [[180]45]. These results indicated potential medicines could be used to manage different subgroups of BLCA cases. Therefore, BLCA cases in high-risk group could be prone to resistance of chemical and immunizing treatment, and our model was useful in guiding management of BLCA cases. A limitation was that exact mechanisms relevant to DRLs need to be explored through further experiments. Therefore, it is required to validate the accuracy and illustrate the exact mechanism of these DRLs signature in BLCA in the future. Besides, our study did not analysis non-muscle invasive bladder cancer samples, and a comparison between the NMIBC and MIBC groups could provide clearer insights into tumor progression. Notably, our study has some advantages. First, this is the detailed report on illustrating connection between DRLs and clinical outcomes in BLCA cases. Second, 9 DRLs markers were identified, which may be potential targets for reversing drug resistance in BLCA. Third, we constructed a prognostic model consisting of 9 DRLs, revealing a high predictive value. Last but not the least, we found differential tumor mutation burden and immune status in high- and low-risk groups. Therefore, our study has greater clinical implications for the prognostic assessment and treatment options for patients with BLCA. Conclusion Based on this study, it would be advisable to identify the key DRLs with potential prognostic value in BLCA to enhance the evaluation of clinical outcomes in this context. In the future, there is the necessity of confirmatory clinical studies involving larger cohorts and multicenter studies and additional basic clinical data validations to substantiate these findings. Author contributions The manuscript was conceived by GC. XPL and HC were responsible for software operation and analysis. XPL completed the manuscript. HC and GC were responsible for manuscript review and revision. The article was submitted with the authorization of all authors who contributed to the article. Funding No funding. Data availability The datasets presented in this study can be found in online repositories. RNA-seq data and relevant clinical information were obtained from The Cancer Genome Atlas database (TCGA-BLCA). Mutation data of patients with BLCA were also downloaded from TCGA database (TCGA-BLCA) in MAF format. These data can be found here: ([181]https://portal.gdc.cancer.gov/). Declarations Competing interests All authors declare that they have no potential conflict of interest. Footnotes Publisher's Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Xiaoping Leng and Han Chen contributed equally to this work. References