Identification and validation of disulfidptosis-related signature to evaluate clinical outcomes, immune infiltration and drug sensitivity in osteosarcoma

Identification and validation of disulfidptosis-related signature to evaluate clinical outcomes, immune infiltration and drug sensitivity in osteosarcoma | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Identification and validation of disulfidptosis-related signature to evaluate clinical outcomes, immune infiltration and drug sensitivity in osteosarcoma Yonghui Zhao, Xiaochen Su, Menghao Teng, Hao Ru, Ziliang lu, Yulong Zhang, and 2 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-4426108/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Disulfidptosis is a novel form of programmed cell death discovered by Liu et al. It's initiated in cells highly expressing SLC7A11, especially in cancers. Our principal aim is to establish and validate a prognostic prediction model for osteosarcoma patients, potentially providing a fresh perspective on the characteristics of disulfidptosis in osteosarcoma and its treatment. Osteosarcoma cohorts obtained from the TARGET and GEO databases were classified into disulfidptosis-high/low-related groups to analyze the Differentially Expressed Genes (DEGs) using the ssGSEA method. DEGs were subsequently analyzed by the Weighted Gene Co-expression Network Analysis (WGCNA) method. Various machine learning algorithms, including the log-rank test, univariate Cox analysis, and LASSO algorithm, were employed, yielding 5 Disulfidptosis-Related Genes (DRGs). GSVA and ssGSEA, were also conducted to investigate the underlying mechanisms of disulfidptosis in osteosarcoma. We established a reliable disulfidptosis-related classification, aand our subsequent analysis has suggested intriguing disparities in the expression of the pentose phosphate pathway (PPP) and cytoskeleton regulation among the groups, indicating that the high-related group was more susceptible to disulfidptosis. 5 disulfidptosis-related genes were selected from the differentially expressed genes (DEGs) , and samples in the cohorts were divided into high-/low-risk groups based on the risk score. Functional analysis demonstrated significantly higher expression of the regulation of the cytoskeleton pathway in the high-risk group. Additionally, immune cell-associated pathways such as the T cell receptor signaling pathway and NOD/TOLL-like receptor signaling pathway showed significant decreases in the high-risk group. We then analyzed the infiltration of immune cells in the tumor microenvironment, revealing lower infiltration of almost every immune cell in the high-risk group. To gain insights into the clinical treatment of osteosarcoma patients, we also analyzed the differences in drug sensitivity between the risk groups, identifying 8 drugs that were more sensitive in the high-risk group. Biological sciences/Biotechnology Biological sciences/Cancer Biological sciences/Computational biology and bioinformatics Biological sciences/Genetics Biological sciences/Immunology osteosarcoma disulfidptosis immunotherapy prognostic model Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Introduction Osteosarcoma is a malignant tumor that significantly impacts the health of children, adolescents, and individuals over 60 years old[ 1 ]. It is most prevalent in adolescents, followed by the elderly in terms of incidence. It typically occurs at the metaphysis of long bones, especially around the knees, and rarely affects the skull, mandible, and pelvis[ 2 ]. For localized OS patients, the standard treatment involves preoperative (neoadjuvant) and postoperative (adjuvant) polychemotherapy in conjunction with surgical resection[ 3 ]. Luckily, 68% of the OS patients are localized and the 5-year survival rate for localized OS patients can exceed 60%. However, for the remaining 20%-30% of OS patients experiencing metastasis or recurrence, the 5-year survival rate is only around 20%[ 2 , 4 ]. Due to the heterogeneity within OS and inter-individual variability[ 5 ], the progress in both developing and applying new molecular targeted drugs has been slow[ 6 ]. Evidence proving the efficacy of immune therapy for OS remains few[ 7 ]. Therefore, further research into the mechanisms of progression and metastasis in OS remains crucial. Disulfidoptosis is a recently identified form of programmed cell death initiated in Solute carrier family 7 member 11 (SLC7A11)-overexpressing cells[ 8 ]. Due to the high oxidative stress induced by the mutation of chromosomes and metabolic reprograming of cancer cells, GSH is severely needed to protect against oxidative-dependent cell death[ 9 ]. For GSH synthesis, cysteine is the rate-limiting precursor, and intracellular cysteine is mainly supplied by SLC7A11-mediated cystine uptake[ 10 ]. Therefore, to counterpart the excessive ROS, most cancer cells typically overexpress SLC7A11[ 11 ]. SLC7A11, also known as xCT, is one of the catalytic subunits of system xc–[ 9 ]. The system xc– is located on the cell membrane and facilitates a 1:1 exchange of extracellular cystine with intracellular glutamate, providing cysteine for GSH synthesis[ 12 ]. And the other subunit of system xc– is the regulatory subunit solute carrier family 3 member 2 (SLC3A2; also known as 4F2hc or CD98), which acts as a chaperone for SLC7A11[ 9 ]. Due to the toxicity of cystine, cancer cells are compelled to reduce it to cysteine upon uptake. This process depletes the cellular NADPH pool. To counteract this NADPH consumption, cancer cells upregulate the glucose pentose phosphate pathway (PPP), resulting in a dependence on glucose[ 13 ]. Consequently, when glucose is limited, high cystine uptake leads to NADPH depletion, causing aberrant disulfide bonding in the actin cytoskeleton. This results in collapse of the actin network and ultimately programmed cell death[ 10 ]. However, the precise mechanism of how disulfide bond-mediated cell death, termed disulfidptosis, functions in osteosarcoma (OS) cells remains unclear. In our study, we identified 24 marked disulfidptosis genes from previous research[ 10 ]. We gathered publicly accessible information on OS patients from the TARGET and GEO datasets (GSE21257). Subsequently, we developed an innoative risk model based on these disulfidptosis-related genes. This model identified 5 Disulfidptosis-Related Genes (DRGs) strongly correlated with both disulfidptosis and OS prognosis. Furthermore, we analyzed the differences in immune characteristics and actin metabolism between the risk groups to elucidate the underlying mechanisms of our risk model. Methods Data collection We obtained the expression profile and the clinical data of OS patients from the TARGETs dataset ( https://www.cancer.gov/ccg/research/genome-sequencing/tcga ) and the GEO dataset ( https://www.ncbi.nlm.nih.gov/geo/ ) [ 14 ]. After excluding samples without expression profile or clinical data, we subsequently got 85 samples from TARGETs dataset for training cohort and 53 samples from GEO dataset for validation cohort. All expression profiles had been transformed with the following formula: log2(exp i +1) , where the exp i is the expression value of each gene. 24 marked disulfidptosis genes were collected from Liu et al.’s study[ 10 ]. Calculation of the disulfidptosis score on the training cohort and DEGs analysis Utilizing the “GSVA” R package, we performed the single-sample gene set enrichment analysis (ssGSEA) technique to calculate a score of each sample that assessed the relevance of the disulfidptosis in training cohort[ 15 ]. The gene set for ssGSEA analysis was the 24 marked disulfidptosis genes we collected before (supplementary table 1 ). Then the cohort was classified into high-relevance class and low relevance class. With the classification of the training cohort, Differentially Expressed Genes (DEGs) between the two classes were analyzed by “limma” R package[ 16 ]. Volcano plot was made by “ggplot2” R package[ 17 ]. Furtherly, utilizing the ssGSEA technique[ 15 ], the expression of pentose phosphate pathway and regulation of cytoskeleton pathway were assessed with ssGSEA score and individually compared in disulfidposis-high/low-related groups to validate the differenet disulfidptosis procession between the two groups. Weighted Gene Co-expression Network Analysis (WGCNA) of the DEGs WGCNA (Weighted Gene Co-expression Network Analysis) is a technique utilized to identify co-expressed genes and evaluate the significance of individual genes and co-expressed gene modules concerning the clinical phenotypes being studied. To delve into the relationship between the Differentially Expressed Genes (DEGs) and disulfidptosis scores within the training cohort, we employed the WGCNA method to partition the DEGs into distinct modules and evaluated the correlation between each module and disulfidptosis scores using the 'WGCNA' R package[ 18 ]. Specifically, we set the minimum number of genes in each module to 50 and the deepSplit parameter to 2. Modules with similarity coefficients below 0.25 were merged. Subsequently, we determined the correlation between gene modules and disulfidptosis scores. The two modules displaying correlation coefficient values greater than 0.5 were chosen as the gene sets for further investigation. Establishment and validation of prognostic model for DRGs The key process of our study is to establish a prognostic model. Univariate Cox proportional hazards regression analysis, log-rank test and least absolute shrinkage and selection operator (LASSO) was used to select the most significant disulfidptosis related genes for prognosis (DRGs) by the genes modules mentioned above[ 19 ]. “glmnet” R package was utilized[ 20 ]. We overlapped the outcomes of univariate Cox proportional hazards regression analysis and the log-rank test, followed by further analysis using the LASSO algorithm. Particularly, we set penalty parameter tuning through 10-fold crossvalidation for the LASSO algorithm. After DRGs being selected, the following formula were performed to calculate the risk score: Risk score = \({\sum }_{i=1}^{n}\left({coef}_{{i}^{ }}*{exp}_{{i}^{ }}\right)\) , where the \({exp}_{{i}^{ }}\) is the gene expression level of the DRGs, and \({coef}_{{i}^{ }}\) is the corresponding regression coefficient from the calculation of the LASSO algorithm. The “survminer” R package was used to calculate the best cutoff value and divide the samples into high risk group and low risk group based on the risk score[ 21 ]. Additionally, Receiver Operating Characteristic (ROC) curve analysis was performed to assess the precision of the risk model using the “timeROC” R package[ 22 ]. The same method was used on another independent cohort (GSE21257) to validate the strength of the prognostic model. Development and validation of a nomogram The nomogram was constructed based on the risk score and clinical factors, with statistical significance at P < 0.05. To assess and refine the nomogram, calibration curves and decision curve analysis were performed, utilizing the 'rms' R package[ 23 ]. Survival analysis To compare the prognosis differences between high risk group and low risk group, Kaplan-Meier survival analysis was utilized using “survival” R package and “survminer” R package[ 24 , 25 ]. Besides, to furtherly analyze the underling molecular mechanism and identify the independent predictors of OS, we conducted Kaplan-Meier survival analysis for each of the disulfidptosis related gens (DRGs) based on their expression levels. ssGSEA and Gene Set Variation Analysis To gain a comprehensive understanding of the functional disparities between the high-risk and low-risk groups, we employed Gene Set Variation Analysis (GSVA). This analysis offered an overarching perspective on the differences in biological processes between the two groups, aiding in the identification of key pathways associated with disulfidptosis in osteosarcoma. Furthermore, we acquired six gene sets from GENE CARD, TIPs ( http://biocc.hrbmu.edu.cn/TIP/ ), and FerrDb ( http://www.zhounan.org/ferrdb/current/ ), covering various biological processes such as glucose metabolism, metabolic metabolism, amino acid metabolism, cancer-immunity cycle, and cytoskeleton metabolism[ 26 ]. Subsequently, we performed single-sample Gene Set Enrichment Analysis (ssGSEA) on the cohort using the 'GSVA' R package[ 15 ]. This analysis enabled us to calculate scores for each biological process, providing insights into their activity levels within individual samples. We then conducted an assessment and comparison of these scores between the high-risk and low-risk groups using six bar plots. These plots visually depicted the differences in activity levels of the respective biological processes, facilitating a clear comparison between the two groups. Immune infiltration, regulation of cytoskeleton characteristics along with drug sensitivity analysis Drawing from the outcomes of our analysis, immune infiltration analysis based on the ssGSEA method was conducted to further explore the immune characteristics in the OS cohorts. We then performed immunophenotyping analysis using the 'ImmuneSubtypeClassifier' R package in R[ 27 ]. This analysis allowed us to evaluate the immune cell composition within the tumor microenvironment and assess any differences between the high-risk and low-risk groups. Overall, these comprehensive analyses provided valuable insights into the molecular and immune landscape associated with disulfidptosis in osteosarcoma. Additionally, we generated gene heatmaps to scrutinize the variances in the Cancer-Immunity Cycle and the regulation of the cytoskeleton pathway between the high-risk and low-risk groups. These heatmaps provided a visual representation of the expression patterns of genes involved in these processes, and significant different genes between the groups was selected. Furthermore, correlation analysis was conducted to explore the relationship between the risk scores and genes in the cytoskeleton regulation pathway. This analysis aimed to identify any potential associations between the risk of disulfidptosis and the expression levels of genes involved in cytoskeleton regulation. Lastly, we conducted drug sensitivity analysis between the groups using the "oncoPredict" R package. Drug sensitivity data was obtained from the Genomics of Drug Sensitivity in Cancer 2 (GDSC2) database ( https://www.cancerrxgene.org/ ), allowing us to assess potential differences in drug responses between the high-risk and low-risk groups of osteosarcoma patients[ 28 ]. Results Disulfidptosis associated classification of the cohorts & different genes analysis Using ssGSEA technique, we scored each of the sample in training cohort based on the 24 marked disulfidptosis genes (supplementary table 1 ), and the training cohort was divided into high relation group and low relation group, where the high relation group indicates a high relevance to the disulfidptosis (Fig. 1 A, supplementary table2). Then, different expression genes analysis was conducted for the two groups, resulting in the identification of DEGs (Fig. 1 B, supplementary table 3). As previous study shows, the occurrence of disulfidptosis relies on the impairment of the PPP pathway and depletion of NADPH[ 13 ]. The regulation of the cytoskeleton pathway promotes the synthesis of the cellular framework, which also plays a role in accelerating the collapse and contraction of the cytoskeleton during disulfidptosis-induced programmed cell death, thereby protecting the cell[ 29 , 10 ]. Accordingly, we conducted further analysis of the expression differences between the two groups and found a significant decrease in the expression of PPP and increase in the expression of the regulation of cytoskeleton pathway in disulfidptosis-high-related groups. This suggested OS cells in high-related groups were more likely to suffer from the disulfidptosis (Fig. 1 C). With the examination of WGCNA, when the soft threshold was 6, the distribution of the expression profile best fit a power-law distribution, and the connections between networks were more stable (Fig. 1 D, Fig. 1 E), which ensured the accuracy of the subsequent analysis. After modules with similarity coefficients bellowing 0.25 being merged, 2 major non-grey genes modules were left (Fig. 1 F), and the correlation coefficient of both the modules were greater than 0.5, presenting that both the modules were significantly connecting with disulfidptosis (Fig. 1 G). With the further analysis of STRING and Cytoscope, we got several genes as the disulfidptosis associated genes set (supplementary table 4,5). Prognostic prediction model establishment and risk score related classification Log-rank test, Univariate Cox regression analysis, along with LASSO algorithm was performed to assess the significance of the genes in the modules to prognosis and subsequently selected 5 DRGs (Fig. 1 H, I). As the random forest plot show, the C index of the 5 DRGs mentioned above was 0.84, indicating a high level of predictive accuracy and a strong consistency between the model’s predictions and actual observations. What’s more, each of the DRGs presented an independent protective factor related to disulfidptosis (Fig. 1 J). Based on the expression levels of the 5 DRGs and the corresponding coefficient values, we calculated the risk scores for each sample and determined the optimal cutoff values for the samples based on these risk scores. The training cohort and validation cohort were then divided into high-risk and low-risk groups accordingly (Fig. 1 K, L). By considering the relevance to disulfidptosis and patients’ prognosis, we established a prognostic model for disulfidptosis in OS. Validation of the prognostic prediction model and establishment of nomogram The classification based on survival and the DRGs expression profile of both the training cohort and the validation cohort were shown in Fig. 2 A, B. By calculating the area under the curve (AUC) for the prediction model at 1 year, 3 years, and 5 years in the training cohort, we obtained values of 0.87, 0.85, and 0.88, respectively, indicating excellent performance of the prognostic model in the training set (Fig. 2 C). The same steps were performed in the validation set to obtain the AUC value, which were 0.84, 0.68, and 0.71, demonstrating good performance of the model in the validation set and confirming its reliability (Fig. 2 D). Following the validation of the model's reliability, we analyzed the hazard ratio of clinical factors related to prognosis in OS patients. The P values for metastasis status and risk score were all less than 0.05, with hazard ratio of 11.13(4.60–26.9), 2.98(1.34–6.6), respectively, indicating a negative correlation with prognosis. And the P values for the gender has a marginally significant effect with hazard ratio of 0.43(0.18-1.0) (Fig. 2 E) (One of the samples in the training cohorts lacked the clinical data and was omitted from the cohorts). Subsequently, combining these clinical factors, we constructed a random forest plot of the model to accurately estimate patients' prognoses, helping with the targeted formulation of treatment plans in clinical practice (Fig. 2 F). Furthermore, we conducted calibration and decision curve analyses to calibrate and validate the clinical prognostic model, comprehensively assessing its accuracy and reliability. The results demonstrated higher accuracy in patient prognosis evaluation compared to other individual clinical markers, confirming the reliability of the predictive model in forecasting patient outcomes (Fig. 2 G, Fig. 2 H). Survival analysis After comprehensively validating the model's reliability, we conducted Kaplan Meier analysis on the 5 genes among DRGs to obtain survival curves depicting the relationship between the high or low expression of each gene and patient survival. The results revealed a significant correlation between the expression levels of all the 5 DRGs and patient survival rates. Among them, CORT and RASGRP2 showed a negative correlation with patient survival time. Others showed a positive correlation. (Fig. 3 A-E). In conjunction with the earlier study, RASGRP2, CORT, CAMK4 and GAB3 had been demonstrated to have an independent relationship with patient prognosis as well as a large extent of difference in both LASSO regression analysis and univariate Cox regression analysis, suggesting a potentially significant role for the 4 genes in the occurrence and progression of OS. Subsequently, we plotted survival curves between the high and low-risk groups in both the training and validation cohorts, based on the previous classification. The curves presented that the high-risk group had significantly inferior survival outcomes compared to the low-risk group in both the training and validation cohorts, further confirming the accuracy of the predictive model (Fig. 3 F, G). GSVA and drug sensitivity analysis GSVA was employed on the risk groups to uncover the underlying mechanisms associated with the risk score. The GSVA-KEGG results unveiled notable disparities in biological pathways, including nicotinate and nicotinamide metabolism, regulation of actin cytoskeleton, and several immune cell-associated pathways between the two groups (Fig. 4 A). Notably, the initiation steps of cell death induced by disulfidptosis rely on cytoskeleton contraction, which significantly correlates with the heightened expression of the regulation of actin cytoskeleton pathway. The GSVA-GO analysis results also showcased significant distinctions in biological processes, such as cysteine catabolic process, alpha ketoglutarate transport, and alanine catabolic process between the two groups (Fig. 4 A, B). However, both the KEGG and GO analyses for GSVA revealed no significant differences in PPP between the two groups. Additionally, further analysis of lipid, glucose, amino acid metabolism, and TIP, ferroptosis, and cytoskeleton biological processes using ssGSEA confirmed that there were no significant differences in nutrient metabolism between the two groups (Fig. 4 C-H). Immune infiltration, regulation of cytoskeleton characteristics along with drug sensitivity analysis among the risk groups As indicated by the results, cytoskeleton metabolism was significantly underexpressed in the high-risk group, suggesting a resistance to disulfidptosis. However, no significant difference was observed in the pentose phosphate pathway (PPP) between the two groups, and nicotinamide metabolism, responsible for generating NADPH precursor substances. To delve into the underlying mechanism of disulfidptosis in osteosarcoma and its influence on outcomes, we also analysed other different biological processes between the risk groups. Findings from the KEGG-GSVA analysis and TIPs highlighted notable variances in immune biological processes between the two groups, especially in pathways associated with immunity like T, NOD, TOLL-like receptor signaling pathways, and the immune cell cycle. Immune cells such as neutrophil granulocytes, macrophages, B lymphocytes, and T lymphocytes can generate ROS through NADPH-dependent ROS-generating oxidases and release ROS into the microenvironment of tumor cells[ 30 , 31 , 32 ]. We proceeded to analyze the immune phenotype. Moreover, upon comparing immune infiltration between risk groups (Fig. 5 A), we observed a significantly higher infiltration of most immune cells in the high-risk group. We also grouped the training and validation cohorts by immune subtype, with the C4 and C6 subtypes having the worst prognosis, the C1 and C2 subtypes having a relatively better prognosis, and the C3 subtype having the best prognosis. We found that regardless of the high or low-risk group, the C4 subtype accounted for a considerable proportion, with a small proportion of the C2 subtype present. The training group also included some C1 subtypes and a small amount of C6 and C3 subtypes. The proportion of the C4 subtype in the high-risk group was significantly higher than that in the low-risk group, while the proportion of the C2 subtype was significantly lower in the high-risk group. In the training set, the proportion of the C1 subtype in the high-risk group was significantly lower than that in the low-risk group. The low-risk group in the training set included small amounts of the C3 and C6 subtypes, which align with the predicted poorer prognosis for the high-risk group, further demonstrating the reliability of the prognostic model (Fig. 5 B). Subsequently, to explore the molecular mechanism, we further plotted heatmaps of gene sets regulating cell cytoskeleton structure and TIPs gene sets between the high and low-risk groups in the training set, presenting a significant difference between the genes regulating the biological process (Fig. 5 C, D). Correlation analysis was also conducted. The result showed that some of the genes showed a positive correlation with the risk-score, indicating a negative correlation with the disulfidptosis. Among them, CYFIP2, which is the subunits of WRC complex, showed the most positive correlation with risk score. According to Liu et al.'s study, CYFIP1, another member of CYFIPs family, demonstrated significant potential to promote disulfidptosis[ 10 ], indicating a potential antagonistic relationship between CYFIP2 and CYFIP1 in disulfidptosis (Fig. 5 E). The drug sensitivity analysis revealed that compounds such as AZ960_1250, XAV939_1268, SB212763_1025, Ruxolitinib_1507, NU7441_1038, MK-8776_2046, JAK_8517_1739, and KU-55933_1030 exhibited greater sensitivity in patients classified in the high-risk group (Fig. 5 F-M). This discovery provides valuable insight into potential clinical treatments for patients with poor prognoses. Discussion OS is the most common primary tumor of the bone, typically occurring in adolescents and significantly impacting the health of children. Since the establishment of treatment standards for OS in the 1980s, the long-term survival rate for localized OS patients has exceeded 60%. However, there has been limited progress in the field of OS treatment over the years[ 2 , 33 ]. Improving the life quality for OS patients and enhancing the survival rate of those who have already relapsed or metastasized remains a challenging issue in the field of OS treatment. Therefore, further exploration of the mechanisms underlying the development of OS remains of significant importance. Disulfidptosis is a newly discovered form of programmed cell death[ 9 ]. It occurs as a cell death process resulting from increased oxidative stress within tumor cells. The oxidative stress forced cancer cells overexpressing SLC7A11 to import more cystine to biosynthesis the glutathione. During the biosynthesis process, NADPH is rapidly depleted, and the thiol groups on the cellular cytoskeleton undergo oxidation to form disulfide bonds, leading to cytoskeletal contraction and ultimately cell death. The reduction expression of the pentose phosphate pathway (PPP), which is the main biosynthesis process of NADPH, plays a predominant role in initiating disulfideptosis[ 10 ]. Therefore, it’s a crucial step to analyze the function of PPP in cancer cells. Additionally, since cell death induced by disulfideptosis hinges on the formation of disulfide bonds and constriction of the cytoskeleton, actin polymerization may be a crucial step as it provides a primary target for disulfide binding and cytoskeleton constriction. Currently, more experimental validation is needed regarding the role of disulfidptosis in the development of OS cells. In our study, utilizing the ssGSEA, we calculated the disulfidptosis-relation score based on the 24 marker genes. The difference between PPP and cytoskeleton regulation expression was analyzed, resulting in a significant low expression in PPP and high expression in cytoskeleton regulation in disulfidptosis high related group, alluding high-relation group were more potential to initiate the disulfidptosis. The DEGs were then identified and further studied by WGCNA and machine study technique such as univariate cox regression analysis, log-rank test and LASSO algorithm, resulting in 5 DRGs which figure a great significance for both the OS prognosis and disulfidptosis in OS cells. We furtherly calculated the risk-scores based on the 5 DRGs. 85 OS patients were classified into high-risk and low-risk groups based on the score. After validating the model, the function analysis was conducted and we observed significant differences between the two groups in processes such as cytoskeleton regulation, JAK-STAT signaling pathway, and immune related process. Particularly, the high-risk group exhibited significantly higher expression in cytoskeleton regulation and a lower expression immune related process. But the PPP pathway showed no significant differences between risk groups. What’s more, nicotinate and nicotinamide metabolism was overexpressed in high risk groups, suggesting a lower generation rate of NADPH. However, NADPH is an important reducing agent for maintaining the redox balance pool in cancer cells which is responsible to reduce the ROS and maintain the GSH level in cytoplasm and lower NADPH level in tumor promoted the disulfidptosis and typically indicated a worse proliferation of tumor cells and better prognosis[ 34 , 35 ]. In the previous study, it was found that neutrophil granulocytes, macrophages, B lymphocytes, and T lymphocytes have the ability to release reactive oxygen species (ROS) into the tumor microenvironment, potentially consuming NADPH to maintain redox balance in cancer cells[ 30 ]. Then we particularly analyzed the infiltration of each immune cells, finding that almost every immune cells are low infiltrated in high risk group. With the results above, it is likely that cancer cells in the high-risk cohort might experience less oxidative pressure from immune cells and be less prone to inducing disulfidptosis, leading to a worse prognosis, which provided a potential insight on the association between disulfidptosis and immune cells infiltration. Additionally, regulation of the cytoskeleton also played a crucial role in both the prognosis of OS and disulfidptosis. We conducted correlation analysis between the risk score and gene sets related to cytoskeleton regulation. Interestingly, CYFIP2 emerged as the gene most positively correlated with the risk score, while CYFIP1, another member of the CYFIP family, was identified as a promoter of disulfidptosis in Liu et al.'s research, suggesting a potential antagonistic relationship between CYFIP2 and CYFIP1. The underlying mechanism remains explored. Finally, we conducted a comparison of drug sensitivity between the risk groups. Among them, 8 drugs exhibited higher sensitivity in the high-risk groups. This finding sheds light on potential treatment strategies for patients with poor prognosis osteosarcoma. Our study has certain limitations. The immune influence on NADPH in OS cells and the relationship between CYFIP2 and CYFIP1 remains experimental validation. To avoid more select bias, we preserved the 85 samples data as much as possible to develop the prognostic model. One of the samples lacked clinical data and was omitted from the cohorts when constructed the clinical associated nomogram, which may result in some degree of impact on the nomogram. And the treatment of the patients in the database might variate, which could influence patients’ prognosis and the accuracy of risk group classification. Conclusion In summary, we identified the disulfidptosis related genes in OS and developed a reliable prognosis prediction model. The immune characteristics between the risk groups were explored, which provided an insight on the relationship between disulfidptosis and immune cells in OS cells. Declarations Author contributions Conceptualization, Yonghui Zhao; Methodology, Yonghui Zhao; Software, Yonghui Zhao, Hao Ru; Validation, Xiaochen Su, Yulong Zhang; Formal Analysis, Yonghui Zhao, Xiaochen Su; Investigation, Yonghui Zhao; Resources, Yilei Zhang, Yingang Zhang; Data Curation, Yonghui Zhao; Writing – Original Draft Preparation, Yonghui Zhao; Writing – Review & Editing, Xiaochen Su, Menghao Teng; Visualization, Yonghui Zhao, Ziliang Lu; Supervision, Yingang Zhang; Project Administration, Yingang Zhang; Funding Acquisition, Yingang Zhang Funding The Education Foundation of Xi'an Jiaotong University (TQ202212) Informed Consent Statement Not applicable. Data Availability Statement The transcriptome data in the research have been annoted with their sources. Acknowledgements Not applicable. Conflicts of interests The author declare no conflict of interest. References Bielack, Stefan S., Beate Kempf-Bielack, Günter Delling, G. Ulrich Exner, Silke Flege, Knut Helmke, Rainer Kotz, et al. “Prognostic Factors in High-Grade Osteosarcoma of theExtremities or Trunk: An Analysis of 1,702 Patients Treatedon Neoadjuvant Cooperative Osteosarcoma Study GroupProtocols.” Journal of Clinical Oncology: Official Journal of the American Society of Clinical Oncology 41, no. 27 (September 20, 2023): 4323–37. https://doi.org/10.1200/JCO.22.02767. Kansara, Maya, Michele W. Teng, Mark J. Smyth, and David M. Thomas. “Translational Biology of Osteosarcoma.” Nature Reviews. Cancer 14, no. 11 (November 2014): 722–35. https://doi.org/10.1038/nrc3838. Ritter, J., and S. S. Bielack. “osteosarcoma.” Annals of Oncology: Official Journal of the European Society for Medical Oncology 21 Suppl 7 (October 2010): vii320-325. https://doi.org/10.1093/annonc/mdq276. Chen, Chenglong, Lu Xie, Tingting Ren, Yi Huang, Jie Xu, and Wei Guo. “Immunotherapy for Osteosarcoma: Fundamental Mechanism, Rationale, and Recent Breakthroughs.” Cancer Letters 500 (March 1, 2021): 1–10. https://doi.org/10.1016/j.canlet.2020.12.024. Corre, Isabelle, Franck Verrecchia, Vincent Crenn, Francoise Redini, and Valérie Trichet. “The osteosarcoma Microenvironment: A Complex but Targetable Ecosystem.” Cells 9, no. 4 (April 15, 2020): 976. https://doi.org/10.3390/cells9040976. Wedekind, Mary F., Lars M. Wagner, and Timothy P. Cripe. “Immunotherapy for osteosarcoma: Where Do We Go from Here?” Pediatric Blood & Cancer 65, no. 9 (2018): e27227. https://doi.org/10.1002/pbc.27227. Wen, Yang, Fan Tang, Chongqi Tu, Francis Hornicek, Zhenfeng Duan, and Li Min. “Immune Checkpoints in Osteosarcoma: Recent Advances and Therapeutic Potential.” Cancer Letters 547 (October 28, 2022): 215887. https://doi.org/10.1016/j.canlet.2022.215887. Koppula, Pranavi, Yilei Zhang, Jiejun Shi, Wei Li, and Boyi Gan. “The Glutamate/Cystine Antiporter SLC7A11/xCT Enhances Cancer Cell Dependency on Glucose by Exporting Glutamate.” The Journal of Biological Chemistry 292, no. 34 (August 25, 2017): 14240–49. https://doi.org/10.1074/jbc.M117.798405. Liu, Xiaoguang, Li Zhuang, and Boyi Gan. “Disulfidptosis: Disulfide Stress–Induced Cell Death.” Trends in Cell Biology, August 2023, S0962892423001411. https://doi.org/10.1016/j.tcb.2023.07.009. Liu, Xiaoguang, Litong Nie, Yilei Zhang, Yuelong Yan, Chao Wang, Medina Colic, Kellen Olszewski, et al. “Actin Cytoskeleton Vulnerability to Disulfide Stress Mediates Disulfidptosis.” Nature Cell Biology 25, no. 3 (March 2023): 404–14. https://doi.org/10.1038/s41556-023-01091-2. Koppula, Pranavi, Yilei Zhang, Li Zhuang, and Boyi Gan. “Amino Acid Transporter SLC7A11/xCT at the Crossroads of Regulating Redox Homeostasis and Nutrient Dependency of Cancer.” Cancer Communications (London, England) 38, no. 1 (April 25, 2018): 12. https://doi.org/10.1186/s40880-018-0288-x Lewerenz, Jan, Sandra J. Hewett, Ying Huang, Maria Lambros, Peter W. Gout, Peter W. Kalivas, Ann Massie, et al. “The Cystine/Glutamate Antiporter System x(c)(-) in Health and Disease: From Molecular Mechanisms to Novel Therapeutic Opportunities.” Antioxidants & Redox Signaling 18, no. 5 (February 10, 2013): 522–55. https://doi.org/10.1089/ars.2011.4391 Liu, Xiaoguang, Kellen Olszewski, Yilei Zhang, Esther W. Lim, Jiejun Shi, Xiaoshan Zhang, Jie Zhang, et al. “Cystine Transporter Regulation of Pentose Phosphate Pathway Dependency and Disulfide Stress Exposes a Targetable Metabolic Vulnerability in Cancer.” Nature Cell Biology 22, no. 4 (April 2020): 476–86. https://doi.org/10.1038/s41556-020-0496-x. Buddingh EP, Kuijjer ML, Duim RA, Bürger H et al. Tumor-infiltrating macrophages are associated with metastasis suppression in high-grade osteosarcoma: a rationale for treatment with macrophage activating agents. Clin Cancer Res 2011 Apr 15;17(8):2110-9. PMID: 21372215 Hänzelmann, Sonja, Robert Castelo, and Justin Guinney. “GSVA: Gene Set Variation Analysis for Microarray and RNA-Seq Data.” BMC Bioinformatics 14, no. 1 (January 16, 2013): 7. https://doi.org/10.1186/1471-2105-14-7. Ritchie, Matthew E., Belinda Phipson, Di Wu, Yifang Hu, Charity W. Law, Wei Shi, and Gordon K. Smyth. “Limma Powers Differential Expression Analyses for RNA-Sequencing and Microarray Studies.” Nucleic Acids Research 43, no. 7 (April 20, 2015): e47. https://doi.org/10.1093/nar/gkv007. “Ggplot2: Elegant Graphics for Data Analysis | Journal of the Royal Statistical Society Series A: Statistics in Society | Oxford Academic.” Accessed May 6, 2024. https://academic.oup.com/jrsssa/article/174/1/245/7077675. Langfelder P, Horvath S (2008). “WGCNA: an R package for weighted correlation network analysis.” _BMC Bioinformatics_, 559.. Tibshirani, Robert. “Regression Shrinkage and Selection Via the Lasso.” Journal of the Royal Statistical Society: Series B (Methodological) 58, no. 1 (1996): 267–88.https://doi.org/10.1111/j.2517-6161.1996.tb02080.x. Friedman J, Tibshirani R, Hastie T (2010). “Regularization Paths for Generalized Linear Models via Coordinate Descent.” _Journal of Statistical Software_, *33*(1), 1-22. doi:10.18637/jss.v033.i01 . Kassambara A, Kosinski M, Biecek P (2021). _survminer: Drawing Survival Curves using 'ggplot2'_. R package version 0.4.9,. Paul Blanche, Jean-Francois Dartigues, Helene Jacqmin-Gadda (2013). Estimating and Comparing time-dependent areas under receiver operating characteristic curves for censored event times with competing risks. Statistics in Medicine,32(30), 5381-5397. URLhttp://onlinelibrary.wiley.com/doi/10.1002/sim.5958/full . Harrell Jr FE (2023). _rms: Regression Modeling Strategies_. R package version 6.7-1, . Therneau T (2023). _A Package for Survival Analysis in R_. R package version 3.5-7, . Terry M. Therneau, Patricia M. Grambsch (2000). _Modeling Survival Data: Extending the Cox Model_. Springer, New York. ISBN 0-387-98784-3. Xu, Liwen, Chunyu Deng, Bo Pang, Xinxin Zhang, Wei Liu, Gaoming Liao, Huating Yuan, et al. “TIP: A Web Server for Resolving Tumor Immunophenotype Profiling.” Cancer Research 78, no. 23 (December 1, 2018): 6575–80. https://doi.org/10.1158/0008-5472.CAN-18-0689. Gibbs D (2024). _ImmuneSubtypeClassifier: An R package for classification of immune subtypes, in cancer, using gene expression data._. R package version 0.1.0. Geeleher P, Cox NJ, Huang RS. Clinical drug response can be predicted using baseline gene expression levels and in vitro drug sensitivity in cell lines. Genome Biol. 2014 Mar 3;15(3):R47. doi: 10.1186/gb-2014-15-3-r47. PMID: 24580837; PMCID: PMC4054092. Burridge, Keith, and Krister Wennerberg. “Rho and Rac Take Center Stage.” Cell 116, no. 2 (January 2004): 167–79. https://doi.org/10.1016/S0092-8674(04)00003-0. Clark, R. A. “Activation of the Neutrophil Respiratory Burst Oxidase.” The Journal of Infectious Diseases 179 Suppl 2 (March 1999): S309-317. https://doi.org/10.1086/513849. Forman, Henry Jay, and Martine Torres. “Reactive Oxygen Species and Cell Signaling: Respiratory Burst in Macrophage Signaling.” American Journal of Respiratory and Critical Care Medicine 166, no. supplement_1 (December 15, 2002): S4–8. https://doi.org/10.1164/rccm.2206007. Bedard, Karen, and Karl-Heinz Krause. “The NOX Family of ROS-Generating NADPH Oxidases: Physiology and Pathophysiology.” Physiological Reviews 87, no. 1 (January 2007): 245–313. https://doi.org/10.1152/physrev.00044.2005. Gill, Jonathan, and Richard Gorlick. “Advancing Therapy for Osteosarcoma.” Nature Reviews. Clinical Oncology 18, no. 10 (October 2021): 609–24. https://doi.org/10.1038/s41571-021-00519-8. Ju, Huai-Qiang, Jin-Fei Lin, Tian Tian, Dan Xie, and Rui-Hua Xu. “NADPH Homeostasis in Cancer: Functions, Mechanisms and Therapeutic Implications.” Signal Transduction and Targeted Therapy 5, no. 1 (October 7, 2020): 231. https://doi.org/10.1038/s41392-020-00326-0. Wang, Xiaole, Kunfeng Chen, and Zhijian Zhao. “LncRNA OR3A4 Regulated the Growth of Osteosarcoma Cells by Modulating the miR-1207-5p/G6PD Signaling.” OncoTargets and Therapy Volume 13 (April 2020): 3117–28. https://doi.org/10.2147/OTT.S234514. Additional Declarations No competing interests reported. Supplementary Files SupplementaryMaterial.xlsx Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-4426108","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":312174008,"identity":"238ee237-4705-4b30-8c2b-89dc027a443b","order_by":0,"name":"Yonghui Zhao","email":"","orcid":"","institution":"The First Affiliated Hospital of Xi'an Jiaotong University","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Yonghui","middleName":"","lastName":"Zhao","suffix":""},{"id":312174009,"identity":"fbeb363f-7dba-4fd8-93d5-1cbb7b7cbde5","order_by":1,"name":"Xiaochen Su","email":"","orcid":"","institution":"The First Affiliated Hospital of Xi'an Jiaotong University","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Xiaochen","middleName":"","lastName":"Su","suffix":""},{"id":312174010,"identity":"ed099e6a-e6cf-4d67-999f-255796d22d39","order_by":2,"name":"Menghao Teng","email":"","orcid":"","institution":"The First Affiliated Hospital of Xi'an Jiaotong University","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Menghao","middleName":"","lastName":"Teng","suffix":""},{"id":312174011,"identity":"ed4c8021-4760-4507-83fe-8ea29f30032d","order_by":3,"name":"Hao Ru","email":"","orcid":"","institution":"The First Affiliated Hospital of Xi'an Jiaotong University","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Hao","middleName":"","lastName":"Ru","suffix":""},{"id":312174012,"identity":"4699f3c0-6503-41bd-9ef6-0276d8512ec8","order_by":4,"name":"Ziliang lu","email":"","orcid":"","institution":"The First Affiliated Hospital of Xi'an Jiaotong University","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Ziliang","middleName":"","lastName":"lu","suffix":""},{"id":312174013,"identity":"435b5900-4fd6-4c82-b2b8-7cf87817d424","order_by":5,"name":"Yulong Zhang","email":"","orcid":"","institution":"The First Affiliated Hospital of Xi'an Jiaotong University","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Yulong","middleName":"","lastName":"Zhang","suffix":""},{"id":312174014,"identity":"3734bd5e-780a-41bd-b0f7-f0b7a51a0565","order_by":6,"name":"Yilei Zhang","email":"","orcid":"","institution":"The First Affiliated Hospital of Xi'an Jiaotong University, Xi'an Jiaotong University","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Yilei","middleName":"","lastName":"Zhang","suffix":""},{"id":312174015,"identity":"3e9151f9-fc39-427b-9015-2d439dd78525","order_by":7,"name":"Yingang Zhang","email":"data:image/png;base64,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","orcid":"","institution":"The First Affiliated Hospital of Xi'an Jiaotong University","correspondingAuthor":true,"submittingAuthor":false,"prefix":"","firstName":"Yingang","middleName":"","lastName":"Zhang","suffix":""}],"badges":[],"createdAt":"2024-05-15 15:16:45","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-4426108/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-4426108/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":58225598,"identity":"497c895a-6053-4a4c-be7a-4dfcfe9322be","added_by":"auto","created_at":"2024-06-12 17:47:29","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":575213,"visible":true,"origin":"","legend":"\u003cp\u003eDevelopment and establishment of risk groups based on the DRGs. (A) Disulfidptosis relation evaluation and classification of the training cohort. (B) Volcano plot of DEGs between the high related group and low related group. Pentose phosphorate pathway and cytoskeleton metabolism analysis and comparison between the groups. (C) The bar plot indicates in high risk group, PPP showed higher expression while cytoskeleton regulation showed lower expression. (D - G) WGCNA of the DEGs between the two groups. (D, E) The best soft threshold is 6. (F, G) Dynamic trees cut plot and module-trait relationships plot showing the non-grew modules. (H - I) LASSO algorithm analysis of the DEGs. (J) Forest plot of the 5 DRGs. (K, L) Best cut-off analysis of risk score both in the training cohort and validation cohort.\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-4426108/v1/a049f40890a9e6459c105a6b.png"},{"id":58225599,"identity":"50b47a13-7000-4b2c-a49f-2e427e0ce8d6","added_by":"auto","created_at":"2024-06-12 17:47:30","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":426442,"visible":true,"origin":"","legend":"\u003cp\u003eValidation of the prognostic prediction model and establishment of nomogram\u003c/p\u003e\n\u003cp\u003eValidation of the prognostic model and nomogram. (A, B) Survival development and DRGs expression with the riskscore increasing both in the training cohort and validation cohort. (C, D) Receiver Operating Characteristic (ROC) curve analysis of the training cohort and validation cohort to evaluate the prognosis accuracy of riskscore. (E) Forest plot for the riskscore and clinical factors. (F) Nomogram integrating the clinical factors and riskscore for the prediction of 1-, 3-, 5-year survival probability. Establishment of calibration curves (G) and decision curve (H) to assessing the reliability of the nomogram.\u003c/p\u003e","description":"","filename":"2.png","url":"https://assets-eu.researchsquare.com/files/rs-4426108/v1/705c56360439bf0e0eee07a1.png"},{"id":58224541,"identity":"5ceee878-1391-4a74-b099-62b146c24500","added_by":"auto","created_at":"2024-06-12 17:39:29","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":370409,"visible":true,"origin":"","legend":"\u003cp\u003eSurvival analysis for the DRGs and risk groups. (A - E) Construction of KM plot for DRGs. Samples were classified into high expression group and low expression group based on the median expression value of each DRGs. (F, J) Survival comparison between the high and low risk groups performing Kaplan Meier plot. High risks group have a significant poor prognosis in both the training cohort and validation cohort.\u003c/p\u003e","description":"","filename":"3.png","url":"https://assets-eu.researchsquare.com/files/rs-4426108/v1/0860b4ae9d72b232a1d037ee.png"},{"id":58224543,"identity":"a3fa7a1e-990c-411e-8e5c-fda0313d5b45","added_by":"auto","created_at":"2024-06-12 17:39:29","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":609396,"visible":true,"origin":"","legend":"\u003cp\u003eDeregulate genes of the groups\u003c/p\u003e\n\u003cp\u003eExploration of the different functions between the risk groups. (A, B) Gene Set Variation Analysis (GSVA) for KEGG and GO analysis heatmap based on DEGs between the high and low risk group. (C - H) Function analysis and comparison in different groups based on the ssGSEA Functions such as TIP (A), glucose metabolism (B), cytoskeleton regulation (C), lipid metabolism (D), ferroptosis (E) and amino acid metabolism (F) are individually analyzed and compared between the high-/low- groups.\u003c/p\u003e","description":"","filename":"4.png","url":"https://assets-eu.researchsquare.com/files/rs-4426108/v1/d1f45fee95c9d6fbcd855d47.png"},{"id":58224542,"identity":"01c4607e-0efb-4ad0-8761-ee14886e6ec3","added_by":"auto","created_at":"2024-06-12 17:39:29","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":1170635,"visible":true,"origin":"","legend":"\u003cp\u003eImmune infiltration, regulation of cytoskeleton characteristics along with drug sensitivity analysis among the risk groups\u003c/p\u003e\n\u003cp\u003eImmune, cytoskeleton and drug sensitivity analysis based on the risk groups. (A) Immune cell infiltration analysis bar-plot in different groups based on the ssGSEA method. (B) Immune subgroup analysis in high and low groups individually, with 6 subgroups included. (C, D) Gene expression comparison heatmap of TIP (immune cell cycle process associated gene set) and the regulation of cytoskeleton. (E) Correlation analysis of spearman and correlation plot, showing the correlation between the riskscores and expression level of genes in the regulation of cytoskeleton pathway. (F - M) Drug sensitivity analysis comparison bar-plot between the risk groups.\u003c/p\u003e","description":"","filename":"5.png","url":"https://assets-eu.researchsquare.com/files/rs-4426108/v1/431441ea5160cd824f0b1ae6.png"},{"id":76265594,"identity":"0c8ebd7f-ae00-4b69-a08d-d541e4db98fe","added_by":"auto","created_at":"2025-02-14 07:23:40","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":4257102,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-4426108/v1/3a0f9e07-6198-4ef6-86bf-30dab85a99b3.pdf"},{"id":58224546,"identity":"27bd208c-7cef-4c8f-953f-4d7dd60aa98d","added_by":"auto","created_at":"2024-06-12 17:39:29","extension":"xlsx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":130406,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryMaterial.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-4426108/v1/beef04b79eaa08ae046d6975.xlsx"}],"financialInterests":"No competing interests reported.","formattedTitle":"Identification and validation of disulfidptosis-related signature to evaluate clinical outcomes, immune infiltration and drug sensitivity in osteosarcoma","fulltext":[{"header":"Introduction","content":"\u003cp\u003eOsteosarcoma is a malignant tumor that significantly impacts the health of children, adolescents, and individuals over 60 years old[\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e]. It is most prevalent in adolescents, followed by the elderly in terms of incidence. It typically occurs at the metaphysis of long bones, especially around the knees, and rarely affects the skull, mandible, and pelvis[\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e]. For localized OS patients, the standard treatment involves preoperative (neoadjuvant) and postoperative (adjuvant) polychemotherapy in conjunction with surgical resection[\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e]. Luckily, 68% of the OS patients are localized and the 5-year survival rate for localized OS patients can exceed 60%. However, for the remaining 20%-30% of OS patients experiencing metastasis or recurrence, the 5-year survival rate is only around 20%[\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e, \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e]. Due to the heterogeneity within OS and inter-individual variability[\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e], the progress in both developing and applying new molecular targeted drugs has been slow[\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e]. Evidence proving the efficacy of immune therapy for OS remains few[\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e]. Therefore, further research into the mechanisms of progression and metastasis in OS remains crucial.\u003c/p\u003e \u003cp\u003eDisulfidoptosis is a recently identified form of programmed cell death initiated in Solute carrier family 7 member 11 (SLC7A11)-overexpressing cells[\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e]. Due to the high oxidative stress induced by the mutation of chromosomes and metabolic reprograming of cancer cells, GSH is severely needed to protect against oxidative-dependent cell death[\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e]. For GSH synthesis, cysteine is the rate-limiting precursor, and intracellular cysteine is mainly supplied by SLC7A11-mediated cystine uptake[\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e]. Therefore, to counterpart the excessive ROS, most cancer cells typically overexpress SLC7A11[\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e]. SLC7A11, also known as xCT, is one of the catalytic subunits of system xc\u0026ndash;[\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e]. The system xc\u0026ndash; is located on the cell membrane and facilitates a 1:1 exchange of extracellular cystine with intracellular glutamate, providing cysteine for GSH synthesis[\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e]. And the other subunit of system xc\u0026ndash; is the regulatory subunit solute carrier family 3 member 2 (SLC3A2; also known as 4F2hc or CD98), which acts as a chaperone for SLC7A11[\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eDue to the toxicity of cystine, cancer cells are compelled to reduce it to cysteine upon uptake. This process depletes the cellular NADPH pool. To counteract this NADPH consumption, cancer cells upregulate the glucose pentose phosphate pathway (PPP), resulting in a dependence on glucose[\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e]. Consequently, when glucose is limited, high cystine uptake leads to NADPH depletion, causing aberrant disulfide bonding in the actin cytoskeleton. This results in collapse of the actin network and ultimately programmed cell death[\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e]. However, the precise mechanism of how disulfide bond-mediated cell death, termed disulfidptosis, functions in osteosarcoma (OS) cells remains unclear.\u003c/p\u003e \u003cp\u003eIn our study, we identified 24 marked disulfidptosis genes from previous research[\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e]. We gathered publicly accessible information on OS patients from the TARGET and GEO datasets (GSE21257). Subsequently, we developed an innoative risk model based on these disulfidptosis-related genes. This model identified 5 Disulfidptosis-Related Genes (DRGs) strongly correlated with both disulfidptosis and OS prognosis. Furthermore, we analyzed the differences in immune characteristics and actin metabolism between the risk groups to elucidate the underlying mechanisms of our risk model.\u003c/p\u003e"},{"header":"Methods","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003eData collection\u003c/h2\u003e \u003cp\u003eWe obtained the expression profile and the clinical data of OS patients from the TARGETs dataset (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.cancer.gov/ccg/research/genome-sequencing/tcga\u003c/span\u003e\u003cspan address=\"https://www.cancer.gov/ccg/research/genome-sequencing/tcga\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e) and the GEO dataset (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.ncbi.nlm.nih.gov/geo/\u003c/span\u003e\u003cspan address=\"https://www.ncbi.nlm.nih.gov/geo/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e) [\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e]. After excluding samples without expression profile or clinical data, we subsequently got 85 samples from TARGETs dataset for training cohort and 53 samples from GEO dataset for validation cohort. All expression profiles had been transformed with the following formula:\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec4\" class=\"Section2\"\u003e \u003ch2\u003elog2(exp\u003csub\u003ei\u003c/sub\u003e+1)\u003c/h2\u003e \u003cp\u003e, where the \u003cem\u003eexp\u003c/em\u003e\u003csub\u003e\u003cem\u003ei\u003c/em\u003e\u003c/sub\u003e is the expression value of each gene. 24 marked disulfidptosis genes were collected from Liu et al.\u0026rsquo;s study[\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e].\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec5\" class=\"Section2\"\u003e \u003ch2\u003eCalculation of the disulfidptosis score on the training cohort and DEGs analysis\u003c/h2\u003e \u003cp\u003eUtilizing the \u0026ldquo;GSVA\u0026rdquo; R package, we performed the single-sample gene set enrichment analysis (ssGSEA) technique to calculate a score of each sample that assessed the relevance of the disulfidptosis in training cohort[\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e]. The gene set for ssGSEA analysis was the 24 marked disulfidptosis genes we collected before (supplementary table \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). Then the cohort was classified into high-relevance class and low relevance class.\u003c/p\u003e \u003cp\u003eWith the classification of the training cohort, Differentially Expressed Genes (DEGs) between the two classes were analyzed by \u0026ldquo;limma\u0026rdquo; R package[\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e]. Volcano plot was made by \u0026ldquo;ggplot2\u0026rdquo; R package[\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eFurtherly, utilizing the ssGSEA technique[\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e], the expression of pentose phosphate pathway and regulation of cytoskeleton pathway were assessed with ssGSEA score and individually compared in disulfidposis-high/low-related groups to validate the differenet disulfidptosis procession between the two groups.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec6\" class=\"Section2\"\u003e \u003ch2\u003eWeighted Gene Co-expression Network Analysis (WGCNA) of the DEGs\u003c/h2\u003e \u003cp\u003eWGCNA (Weighted Gene Co-expression Network Analysis) is a technique utilized to identify co-expressed genes and evaluate the significance of individual genes and co-expressed gene modules concerning the clinical phenotypes being studied. To delve into the relationship between the Differentially Expressed Genes (DEGs) and disulfidptosis scores within the training cohort, we employed the WGCNA method to partition the DEGs into distinct modules and evaluated the correlation between each module and disulfidptosis scores using the 'WGCNA' R package[\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e]. Specifically, we set the minimum number of genes in each module to 50 and the deepSplit parameter to 2. Modules with similarity coefficients below 0.25 were merged. Subsequently, we determined the correlation between gene modules and disulfidptosis scores. The two modules displaying correlation coefficient values greater than 0.5 were chosen as the gene sets for further investigation.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec7\" class=\"Section2\"\u003e \u003ch2\u003eEstablishment and validation of prognostic model for DRGs\u003c/h2\u003e \u003cp\u003eThe key process of our study is to establish a prognostic model. Univariate Cox proportional hazards regression analysis, log-rank test and least absolute shrinkage and selection operator (LASSO) was used to select the most significant disulfidptosis related genes for prognosis (DRGs) by the genes modules mentioned above[\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e]. \u0026ldquo;glmnet\u0026rdquo; R package was utilized[\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e]. We overlapped the outcomes of univariate Cox proportional hazards regression analysis and the log-rank test, followed by further analysis using the LASSO algorithm. Particularly, we set penalty parameter tuning through 10-fold crossvalidation for the LASSO algorithm. After DRGs being selected, the following formula were performed to calculate the risk score:\u003c/p\u003e \u003cp\u003e \u003cem\u003eRisk score\u003c/em\u003e = \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\({\\sum }_{i=1}^{n}\\left({coef}_{{i}^{ }}*{exp}_{{i}^{ }}\\right)\\)\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e \u003cp\u003e, where the \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\({exp}_{{i}^{ }}\\)\u003c/span\u003e\u003c/span\u003e is the gene expression level of the DRGs, and \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\({coef}_{{i}^{ }}\\)\u003c/span\u003e\u003c/span\u003e is the corresponding regression coefficient from the calculation of the LASSO algorithm. The \u0026ldquo;survminer\u0026rdquo; R package was used to calculate the best cutoff value and divide the samples into high risk group and low risk group based on the risk score[\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e]. Additionally, Receiver Operating Characteristic (ROC) curve analysis was performed to assess the precision of the risk model using the \u0026ldquo;timeROC\u0026rdquo; R package[\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e]. The same method was used on another independent cohort (GSE21257) to validate the strength of the prognostic model.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003eDevelopment and validation of a nomogram\u003c/h2\u003e \u003cp\u003eThe nomogram was constructed based on the risk score and clinical factors, with statistical significance at P\u0026thinsp;\u0026lt;\u0026thinsp;0.05. To assess and refine the nomogram, calibration curves and decision curve analysis were performed, utilizing the 'rms' R package[\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e].\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec9\" class=\"Section2\"\u003e \u003ch2\u003eSurvival analysis\u003c/h2\u003e \u003cp\u003eTo compare the prognosis differences between high risk group and low risk group, Kaplan-Meier survival analysis was utilized using \u0026ldquo;survival\u0026rdquo; R package and \u0026ldquo;survminer\u0026rdquo; R package[\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e, \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e]. Besides, to furtherly analyze the underling molecular mechanism and identify the independent predictors of OS, we conducted Kaplan-Meier survival analysis for each of the disulfidptosis related gens (DRGs) based on their expression levels.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec10\" class=\"Section2\"\u003e \u003ch2\u003essGSEA and Gene Set Variation Analysis\u003c/h2\u003e \u003cp\u003eTo gain a comprehensive understanding of the functional disparities between the high-risk and low-risk groups, we employed Gene Set Variation Analysis (GSVA). This analysis offered an overarching perspective on the differences in biological processes between the two groups, aiding in the identification of key pathways associated with disulfidptosis in osteosarcoma.\u003c/p\u003e \u003cp\u003eFurthermore, we acquired six gene sets from GENE CARD, TIPs (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://biocc.hrbmu.edu.cn/TIP/\u003c/span\u003e\u003cspan address=\"http://biocc.hrbmu.edu.cn/TIP/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e), and FerrDb (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://www.zhounan.org/ferrdb/current/\u003c/span\u003e\u003cspan address=\"http://www.zhounan.org/ferrdb/current/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e), covering various biological processes such as glucose metabolism, metabolic metabolism, amino acid metabolism, cancer-immunity cycle, and cytoskeleton metabolism[\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e]. Subsequently, we performed single-sample Gene Set Enrichment Analysis (ssGSEA) on the cohort using the 'GSVA' R package[\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e]. This analysis enabled us to calculate scores for each biological process, providing insights into their activity levels within individual samples. We then conducted an assessment and comparison of these scores between the high-risk and low-risk groups using six bar plots. These plots visually depicted the differences in activity levels of the respective biological processes, facilitating a clear comparison between the two groups.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec11\" class=\"Section2\"\u003e \u003ch2\u003eImmune infiltration, regulation of cytoskeleton characteristics along with drug sensitivity analysis\u003c/h2\u003e \u003cp\u003eDrawing from the outcomes of our analysis, immune infiltration analysis based on the ssGSEA method was conducted to further explore the immune characteristics in the OS cohorts. We then performed immunophenotyping analysis using the 'ImmuneSubtypeClassifier' R package in R[\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e]. This analysis allowed us to evaluate the immune cell composition within the tumor microenvironment and assess any differences between the high-risk and low-risk groups. Overall, these comprehensive analyses provided valuable insights into the molecular and immune landscape associated with disulfidptosis in osteosarcoma.\u003c/p\u003e \u003cp\u003eAdditionally, we generated gene heatmaps to scrutinize the variances in the Cancer-Immunity Cycle and the regulation of the cytoskeleton pathway between the high-risk and low-risk groups. These heatmaps provided a visual representation of the expression patterns of genes involved in these processes, and significant different genes between the groups was selected. Furthermore, correlation analysis was conducted to explore the relationship between the risk scores and genes in the cytoskeleton regulation pathway. This analysis aimed to identify any potential associations between the risk of disulfidptosis and the expression levels of genes involved in cytoskeleton regulation. Lastly, we conducted drug sensitivity analysis between the groups using the \"oncoPredict\" R package. Drug sensitivity data was obtained from the Genomics of Drug Sensitivity in Cancer 2 (GDSC2) database (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.cancerrxgene.org/\u003c/span\u003e\u003cspan address=\"https://www.cancerrxgene.org/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e), allowing us to assess potential differences in drug responses between the high-risk and low-risk groups of osteosarcoma patients[\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e].\u003c/p\u003e \u003c/div\u003e"},{"header":"Results","content":"\u003cdiv id=\"Sec13\" class=\"Section2\"\u003e \u003ch2\u003eDisulfidptosis associated classification of the cohorts \u0026amp; different genes analysis\u003c/h2\u003e \u003cp\u003eUsing ssGSEA technique, we scored each of the sample in training cohort based on the 24 marked disulfidptosis genes (supplementary table \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003e1\u003c/span\u003e), and the training cohort was divided into high relation group and low relation group, where the high relation group indicates a high relevance to the disulfidptosis (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e1\u003c/span\u003eA, supplementary table2). Then, different expression genes analysis was conducted for the two groups, resulting in the identification of DEGs (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e1\u003c/span\u003eB, supplementary table 3). As previous study shows, the occurrence of disulfidptosis relies on the impairment of the PPP pathway and depletion of NADPH[\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e]. The regulation of the cytoskeleton pathway promotes the synthesis of the cellular framework, which also plays a role in accelerating the collapse and contraction of the cytoskeleton during disulfidptosis-induced programmed cell death, thereby protecting the cell[\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e, \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e]. Accordingly, we conducted further analysis of the expression differences between the two groups and found a significant decrease in the expression of PPP and increase in the expression of the regulation of cytoskeleton pathway in disulfidptosis-high-related groups. This suggested OS cells in high-related groups were more likely to suffer from the disulfidptosis (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e1\u003c/span\u003eC). With the examination of WGCNA, when the soft threshold was 6, the distribution of the expression profile best fit a power-law distribution, and the connections between networks were more stable (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e1\u003c/span\u003eD, Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e1\u003c/span\u003eE), which ensured the accuracy of the subsequent analysis. After modules with similarity coefficients bellowing 0.25 being merged, 2 major non-grey genes modules were left (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e1\u003c/span\u003eF), and the correlation coefficient of both the modules were greater than 0.5, presenting that both the modules were significantly connecting with disulfidptosis (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e1\u003c/span\u003eG). With the further analysis of STRING and Cytoscope, we got several genes as the disulfidptosis associated genes set (supplementary table 4,5).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec14\" class=\"Section2\"\u003e \u003ch2\u003ePrognostic prediction model establishment and risk score related classification\u003c/h2\u003e \u003cp\u003eLog-rank test, Univariate Cox regression analysis, along with LASSO algorithm was performed to assess the significance of the genes in the modules to prognosis and subsequently selected 5 DRGs (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e1\u003c/span\u003eH, I). As the random forest plot show, the C index of the 5 DRGs mentioned above was 0.84, indicating a high level of predictive accuracy and a strong consistency between the model\u0026rsquo;s predictions and actual observations. What\u0026rsquo;s more, each of the DRGs presented an independent protective factor related to disulfidptosis (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e1\u003c/span\u003eJ).\u003c/p\u003e \u003cp\u003eBased on the expression levels of the 5 DRGs and the corresponding coefficient values, we calculated the risk scores for each sample and determined the optimal cutoff values for the samples based on these risk scores. The training cohort and validation cohort were then divided into high-risk and low-risk groups accordingly (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e1\u003c/span\u003eK, L). By considering the relevance to disulfidptosis and patients\u0026rsquo; prognosis, we established a prognostic model for disulfidptosis in OS.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec15\" class=\"Section2\"\u003e \u003ch2\u003eValidation of the prognostic prediction model and establishment of nomogram\u003c/h2\u003e \u003cp\u003eThe classification based on survival and the DRGs expression profile of both the training cohort and the validation cohort were shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e2\u003c/span\u003eA, B. By calculating the area under the curve (AUC) for the prediction model at 1 year, 3 years, and 5 years in the training cohort, we obtained values of 0.87, 0.85, and 0.88, respectively, indicating excellent performance of the prognostic model in the training set (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e2\u003c/span\u003eC). The same steps were performed in the validation set to obtain the AUC\u003c/p\u003e \u003cp\u003evalue, which were 0.84, 0.68, and 0.71, demonstrating good performance of the model in the validation set and confirming its reliability (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e2\u003c/span\u003eD).\u003c/p\u003e \u003cp\u003eFollowing the validation of the model's reliability, we analyzed the hazard ratio of clinical factors related to prognosis in OS patients. The P values for metastasis status and risk score were all less than 0.05, with hazard ratio of 11.13(4.60\u0026ndash;26.9), 2.98(1.34\u0026ndash;6.6), respectively, indicating a negative correlation with prognosis. And the P values for the gender has a marginally significant effect with hazard ratio of 0.43(0.18-1.0) (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e2\u003c/span\u003eE) (One of the samples in the training cohorts lacked the clinical data and was omitted from the cohorts). Subsequently, combining these clinical factors, we constructed a random forest plot of the model to accurately estimate patients' prognoses, helping with the targeted formulation of treatment plans in clinical practice (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e2\u003c/span\u003eF).\u003c/p\u003e \u003cp\u003eFurthermore, we conducted calibration and decision curve analyses to calibrate and validate the clinical prognostic model, comprehensively assessing its accuracy and reliability. The results demonstrated higher accuracy in patient prognosis evaluation compared to other individual clinical markers, confirming the reliability of the predictive model in forecasting patient outcomes (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e2\u003c/span\u003eG, Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e2\u003c/span\u003eH).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec16\" class=\"Section2\"\u003e \u003ch2\u003eSurvival analysis\u003c/h2\u003e \u003cp\u003eAfter comprehensively validating the model's reliability, we conducted Kaplan Meier analysis on the 5 genes among DRGs to obtain survival curves depicting the relationship between the high or low expression of each gene and patient survival. The results revealed a significant correlation between the expression levels of all the 5 DRGs and patient survival rates. Among them, CORT and RASGRP2 showed a negative correlation with patient survival time. Others showed a positive correlation. (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e3\u003c/span\u003eA-E). In conjunction with the earlier study, RASGRP2, CORT, CAMK4 and GAB3 had been demonstrated to have an independent relationship with patient prognosis as well as a large extent of difference in both LASSO regression analysis and univariate Cox regression analysis, suggesting a potentially significant role for the 4 genes in the occurrence and progression of OS.\u003c/p\u003e \u003cp\u003eSubsequently, we plotted survival curves between the high and low-risk groups in both the training and validation cohorts, based on the previous classification. The curves presented that the high-risk group had significantly inferior survival outcomes compared to the low-risk group in both the training and validation cohorts, further confirming the accuracy of the predictive model (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e3\u003c/span\u003eF, G).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec17\" class=\"Section2\"\u003e \u003ch2\u003eGSVA and drug sensitivity analysis\u003c/h2\u003e \u003cp\u003eGSVA was employed on the risk groups to uncover the underlying mechanisms associated with the risk score. The GSVA-KEGG results unveiled notable disparities in biological pathways, including nicotinate and nicotinamide metabolism, regulation of actin cytoskeleton, and several immune cell-associated pathways between the two groups (Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e4\u003c/span\u003eA). Notably, the initiation steps of cell death induced by disulfidptosis rely on cytoskeleton contraction, which significantly correlates with the heightened expression of the regulation of actin cytoskeleton pathway. The GSVA-GO analysis results also showcased significant distinctions in biological processes, such as cysteine catabolic process, alpha ketoglutarate transport, and alanine catabolic process between the two groups (Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e4\u003c/span\u003eA, B). However, both the KEGG and GO analyses for GSVA revealed no significant differences in PPP between the two groups. Additionally, further analysis of lipid, glucose, amino acid metabolism, and TIP, ferroptosis, and cytoskeleton biological processes using ssGSEA confirmed that there were no significant differences in nutrient metabolism between the two groups (Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e4\u003c/span\u003eC-H).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec18\" class=\"Section2\"\u003e \u003ch2\u003eImmune infiltration, regulation of cytoskeleton characteristics along with drug sensitivity analysis among the risk groups\u003c/h2\u003e \u003cp\u003eAs indicated by the results, cytoskeleton metabolism was significantly underexpressed in the high-risk group, suggesting a resistance to disulfidptosis. However, no significant difference was observed in the pentose phosphate pathway (PPP) between the two groups, and nicotinamide metabolism, responsible for generating NADPH precursor substances.\u003c/p\u003e \u003cp\u003eTo delve into the underlying mechanism of disulfidptosis in osteosarcoma and its influence on outcomes, we also analysed other different biological processes between the risk groups. Findings from the KEGG-GSVA analysis and TIPs highlighted notable variances in immune biological processes between the two groups, especially in pathways associated with immunity like T, NOD, TOLL-like receptor signaling pathways, and the immune cell cycle. Immune cells such as neutrophil granulocytes, macrophages, B lymphocytes, and T lymphocytes can generate ROS through NADPH-dependent ROS-generating oxidases and release ROS into the microenvironment of tumor cells[\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e, \u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e, \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e]. We proceeded to analyze the immune phenotype. Moreover, upon comparing immune infiltration between risk groups (Fig.\u0026nbsp;\u003cspan refid=\"Fig10\" class=\"InternalRef\"\u003e5\u003c/span\u003eA), we observed a significantly higher infiltration of most immune cells in the high-risk group.\u003c/p\u003e \u003cp\u003eWe also grouped the training and validation cohorts by immune subtype, with the C4 and C6 subtypes having the worst prognosis, the C1 and C2 subtypes having a relatively better prognosis, and the C3 subtype having the best prognosis. We found that regardless of the high or low-risk group, the C4 subtype accounted for a considerable proportion, with a small proportion of the C2 subtype present. The training group also included some C1 subtypes and a small amount of C6 and C3 subtypes. The proportion of the C4 subtype in the high-risk group was significantly higher than that in the low-risk group, while the proportion of the C2 subtype was significantly lower in the high-risk group. In the training set, the proportion of the C1 subtype in the high-risk group was significantly lower than that in the low-risk group. The low-risk group in the training set included small amounts of the C3 and C6 subtypes, which align with the predicted poorer prognosis for the high-risk group, further demonstrating the reliability of the prognostic model (Fig.\u0026nbsp;\u003cspan refid=\"Fig10\" class=\"InternalRef\"\u003e5\u003c/span\u003eB). Subsequently, to explore the molecular mechanism, we further plotted heatmaps of gene sets regulating cell cytoskeleton structure and TIPs gene sets between the high and low-risk groups in the training set, presenting a significant difference between the genes regulating the biological process (Fig.\u0026nbsp;\u003cspan refid=\"Fig10\" class=\"InternalRef\"\u003e5\u003c/span\u003eC, D). Correlation analysis was also conducted. The result showed that some of the genes showed a positive correlation with the risk-score, indicating a negative correlation with the disulfidptosis. Among them, CYFIP2, which is the subunits of WRC complex, showed the most positive correlation with risk score. According to Liu et al.'s study, CYFIP1, another member of CYFIPs family, demonstrated significant potential to promote disulfidptosis[\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e], indicating a potential antagonistic relationship between CYFIP2 and CYFIP1 in disulfidptosis (Fig.\u0026nbsp;\u003cspan refid=\"Fig10\" class=\"InternalRef\"\u003e5\u003c/span\u003eE).\u003c/p\u003e \u003cp\u003eThe drug sensitivity analysis revealed that compounds such as AZ960_1250, XAV939_1268, SB212763_1025, Ruxolitinib_1507, NU7441_1038, MK-8776_2046, JAK_8517_1739, and KU-55933_1030 exhibited greater sensitivity in patients classified in the high-risk group (Fig.\u0026nbsp;\u003cspan refid=\"Fig10\" class=\"InternalRef\"\u003e5\u003c/span\u003eF-M). This discovery provides valuable insight into potential clinical treatments for patients with poor prognoses.\u003c/p\u003e \u003c/div\u003e"},{"header":"Discussion","content":"\u003cp\u003eOS is the most common primary tumor of the bone, typically occurring in adolescents and significantly impacting the health of children. Since the establishment of treatment standards for OS in the 1980s, the long-term survival rate for localized OS patients has exceeded 60%. However, there has been limited progress in the field of OS treatment over the years[\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e, \u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e]. Improving the life quality for OS patients and enhancing the survival rate of those who have already relapsed or metastasized remains a challenging issue in the field of OS treatment. Therefore, further exploration of the mechanisms underlying the development of OS remains of significant importance.\u003c/p\u003e \u003cp\u003eDisulfidptosis is a newly discovered form of programmed cell death[\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e]. It occurs as a cell death process resulting from increased oxidative stress within tumor cells. The oxidative stress forced cancer cells overexpressing SLC7A11 to import more cystine to biosynthesis the glutathione. During the biosynthesis process, NADPH is rapidly depleted, and the thiol groups on the cellular cytoskeleton undergo oxidation to form disulfide bonds, leading to cytoskeletal contraction and ultimately cell death. The reduction expression of the pentose phosphate pathway (PPP), which is the main biosynthesis process of NADPH, plays a predominant role in initiating disulfideptosis[\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e]. Therefore, it\u0026rsquo;s a crucial step to analyze the function of PPP in cancer cells. Additionally, since cell death induced by disulfideptosis hinges on the formation of disulfide bonds and constriction of the cytoskeleton, actin polymerization may be a crucial step as it provides a primary target for disulfide binding and cytoskeleton constriction. Currently, more experimental validation is needed regarding the role of disulfidptosis in the development of OS cells.\u003c/p\u003e \u003cp\u003eIn our study, utilizing the ssGSEA, we calculated the disulfidptosis-relation score based on the 24 marker genes. The difference between PPP and cytoskeleton regulation expression was analyzed, resulting in a significant low expression in PPP and high expression in cytoskeleton regulation in disulfidptosis high related group, alluding high-relation group were more potential to initiate the disulfidptosis. The DEGs were then identified and further studied by WGCNA and machine study technique such as univariate cox regression analysis, log-rank test and LASSO algorithm, resulting in 5 DRGs which figure a great significance for both the OS prognosis and disulfidptosis in OS cells. We furtherly calculated the risk-scores based on the 5 DRGs. 85 OS patients were classified into high-risk and low-risk groups based on the score. After validating the model, the function analysis was conducted and we observed significant differences between the two groups in processes such as cytoskeleton regulation, JAK-STAT signaling pathway, and immune related process. Particularly, the high-risk group exhibited significantly higher expression in cytoskeleton regulation and a lower expression immune related process. But the PPP pathway showed no significant differences between risk groups. What\u0026rsquo;s more, nicotinate and nicotinamide metabolism was overexpressed in high risk groups, suggesting a lower generation rate of NADPH. However, NADPH is an important reducing agent for maintaining the redox balance pool in cancer cells which is responsible to reduce the ROS and maintain the GSH level in cytoplasm and lower NADPH level in tumor promoted the disulfidptosis and typically indicated a worse proliferation of tumor cells and better prognosis[\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e, \u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e]. In the previous study, it was found that neutrophil granulocytes, macrophages, B lymphocytes, and T lymphocytes have the ability to release reactive oxygen species (ROS) into the tumor microenvironment, potentially consuming NADPH to maintain redox balance in cancer cells[\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e]. Then we particularly analyzed the infiltration of each immune cells, finding that almost every immune cells are low infiltrated in high risk group. With the results above, it is likely that cancer cells in the high-risk cohort might experience less oxidative pressure from immune cells and be less prone to inducing disulfidptosis, leading to a worse prognosis, which provided a potential insight on the association between disulfidptosis and immune cells infiltration.\u003c/p\u003e \u003cp\u003eAdditionally, regulation of the cytoskeleton also played a crucial role in both the prognosis of OS and disulfidptosis. We conducted correlation analysis between the risk score and gene sets related to cytoskeleton regulation. Interestingly, CYFIP2 emerged as the gene most positively correlated with the risk score, while CYFIP1, another member of the CYFIP family, was identified as a promoter of disulfidptosis in Liu et al.'s research, suggesting a potential antagonistic relationship between CYFIP2 and CYFIP1. The underlying mechanism remains explored.\u003c/p\u003e \u003cp\u003eFinally, we conducted a comparison of drug sensitivity between the risk groups. Among them, 8 drugs exhibited higher sensitivity in the high-risk groups. This finding sheds light on potential treatment strategies for patients with poor prognosis osteosarcoma.\u003c/p\u003e \u003cp\u003eOur study has certain limitations. The immune influence on NADPH in OS cells and the relationship between CYFIP2 and CYFIP1 remains experimental validation. To avoid more select bias, we preserved the 85 samples data as much as possible to develop the prognostic model. One of the samples lacked clinical data and was omitted from the cohorts when constructed the clinical associated nomogram, which may result in some degree of impact on the nomogram. And the treatment of the patients in the database might variate, which could influence patients\u0026rsquo; prognosis and the accuracy of risk group classification.\u003c/p\u003e"},{"header":"Conclusion","content":"\u003cp\u003eIn summary, we identified the disulfidptosis related genes in OS and developed a reliable prognosis prediction model. The immune characteristics between the risk groups were explored, which provided an insight on the relationship between disulfidptosis and immune cells in OS cells.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eAuthor contributions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eConceptualization, Yonghui Zhao; Methodology, Yonghui Zhao; Software, Yonghui Zhao, Hao Ru; Validation, Xiaochen Su, Yulong Zhang; Formal Analysis, Yonghui Zhao, Xiaochen Su; Investigation, Yonghui Zhao; Resources, Yilei Zhang, Yingang Zhang; Data Curation, Yonghui Zhao; Writing – Original Draft Preparation, Yonghui Zhao; Writing – Review \u0026amp; Editing, Xiaochen Su, Menghao Teng; Visualization, Yonghui Zhao, Ziliang Lu; Supervision, Yingang Zhang; Project Administration, Yingang Zhang; Funding Acquisition, Yingang Zhang\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe Education Foundation of Xi'an Jiaotong University (TQ202212)\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eInformed Consent Statement\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNot applicable.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eData Availability Statement\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe transcriptome data in the research have been annoted with their sources.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAcknowledgements\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNot applicable.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConflicts of interests\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe author declare no conflict of interest.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eBielack, Stefan S., Beate Kempf-Bielack, G\u0026uuml;nter Delling, G. Ulrich Exner, Silke Flege, Knut Helmke, Rainer Kotz, et al. \u0026ldquo;Prognostic Factors in High-Grade Osteosarcoma of theExtremities or Trunk: An Analysis of 1,702 Patients Treatedon Neoadjuvant Cooperative Osteosarcoma Study GroupProtocols.\u0026rdquo; Journal of Clinical Oncology: Official Journal of the American Society of Clinical Oncology 41, no. 27 (September 20, 2023): 4323\u0026ndash;37. https://doi.org/10.1200/JCO.22.02767.\u003c/li\u003e\n\u003cli\u003eKansara, Maya, Michele W. Teng, Mark J. Smyth, and David M. Thomas. \u0026ldquo;Translational Biology of Osteosarcoma.\u0026rdquo; Nature Reviews. Cancer 14, no. 11 (November 2014): 722\u0026ndash;35. https://doi.org/10.1038/nrc3838.\u003c/li\u003e\n\u003cli\u003eRitter, J., and S. S. Bielack. \u0026ldquo;osteosarcoma.\u0026rdquo; Annals of Oncology: Official Journal of the European Society for Medical Oncology 21 Suppl 7 (October 2010): vii320-325. https://doi.org/10.1093/annonc/mdq276.\u003c/li\u003e\n\u003cli\u003eChen, Chenglong, Lu Xie, Tingting Ren, Yi Huang, Jie Xu, and Wei Guo. \u0026ldquo;Immunotherapy for Osteosarcoma: Fundamental Mechanism, Rationale, and Recent Breakthroughs.\u0026rdquo; Cancer Letters 500 (March 1, 2021): 1\u0026ndash;10. https://doi.org/10.1016/j.canlet.2020.12.024.\u003c/li\u003e\n\u003cli\u003eCorre, Isabelle, Franck Verrecchia, Vincent Crenn, Francoise Redini, and Val\u0026eacute;rie Trichet. \u0026ldquo;The osteosarcoma Microenvironment: A Complex but Targetable Ecosystem.\u0026rdquo; Cells 9, no. 4 (April 15, 2020): 976. https://doi.org/10.3390/cells9040976.\u003c/li\u003e\n\u003cli\u003eWedekind, Mary F., Lars M. Wagner, and Timothy P. Cripe. \u0026ldquo;Immunotherapy for osteosarcoma: Where Do We Go from Here?\u0026rdquo; Pediatric Blood \u0026amp; Cancer 65, no. 9 (2018): e27227. https://doi.org/10.1002/pbc.27227.\u003c/li\u003e\n\u003cli\u003eWen, Yang, Fan Tang, Chongqi Tu, Francis Hornicek, Zhenfeng Duan, and Li Min. \u0026ldquo;Immune Checkpoints in Osteosarcoma: Recent Advances and Therapeutic Potential.\u0026rdquo; Cancer Letters 547 (October 28, 2022): 215887. https://doi.org/10.1016/j.canlet.2022.215887.\u003c/li\u003e\n\u003cli\u003eKoppula, Pranavi, Yilei Zhang, Jiejun Shi, Wei Li, and Boyi Gan. \u0026ldquo;The Glutamate/Cystine Antiporter SLC7A11/xCT Enhances Cancer Cell Dependency on Glucose by Exporting Glutamate.\u0026rdquo; The Journal of Biological Chemistry 292, no. 34 (August 25, 2017): 14240\u0026ndash;49. https://doi.org/10.1074/jbc.M117.798405.\u003c/li\u003e\n\u003cli\u003eLiu, Xiaoguang, Li Zhuang, and Boyi Gan. \u0026ldquo;Disulfidptosis: Disulfide Stress\u0026ndash;Induced Cell Death.\u0026rdquo; Trends in Cell Biology, August 2023, S0962892423001411. https://doi.org/10.1016/j.tcb.2023.07.009.\u003c/li\u003e\n\u003cli\u003eLiu, Xiaoguang, Litong Nie, Yilei Zhang, Yuelong Yan, Chao Wang, Medina Colic, Kellen Olszewski, et al. \u0026ldquo;Actin Cytoskeleton Vulnerability to Disulfide Stress Mediates Disulfidptosis.\u0026rdquo; Nature Cell Biology 25, no. 3 (March 2023): 404\u0026ndash;14. https://doi.org/10.1038/s41556-023-01091-2.\u003c/li\u003e\n\u003cli\u003eKoppula, Pranavi, Yilei Zhang, Li Zhuang, and Boyi Gan. \u0026ldquo;Amino Acid Transporter SLC7A11/xCT at the Crossroads of Regulating Redox Homeostasis and Nutrient Dependency of Cancer.\u0026rdquo; Cancer Communications (London, England) 38, no. 1 (April 25, 2018): 12. https://doi.org/10.1186/s40880-018-0288-x\u003c/li\u003e\n\u003cli\u003eLewerenz, Jan, Sandra J. Hewett, Ying Huang, Maria Lambros, Peter W. Gout, Peter W. Kalivas, Ann Massie, et al. \u0026ldquo;The Cystine/Glutamate Antiporter System x(c)(-) in Health and Disease: From Molecular Mechanisms to Novel Therapeutic Opportunities.\u0026rdquo; Antioxidants \u0026amp; Redox Signaling 18, no. 5 (February 10, 2013): 522\u0026ndash;55. https://doi.org/10.1089/ars.2011.4391\u003c/li\u003e\n\u003cli\u003eLiu, Xiaoguang, Kellen Olszewski, Yilei Zhang, Esther W. Lim, Jiejun Shi, Xiaoshan Zhang, Jie Zhang, et al. \u0026ldquo;Cystine Transporter Regulation of Pentose Phosphate Pathway Dependency and Disulfide Stress Exposes a Targetable Metabolic Vulnerability in Cancer.\u0026rdquo; Nature Cell Biology 22, no. 4 (April 2020): 476\u0026ndash;86. https://doi.org/10.1038/s41556-020-0496-x.\u003c/li\u003e\n\u003cli\u003eBuddingh EP, Kuijjer ML, Duim RA, B\u0026uuml;rger H et al. Tumor-infiltrating macrophages are associated with metastasis suppression in high-grade osteosarcoma: a rationale for treatment with macrophage activating agents. Clin Cancer Res 2011 Apr 15;17(8):2110-9. PMID: 21372215\u003c/li\u003e\n\u003cli\u003eH\u0026auml;nzelmann, Sonja, Robert Castelo, and Justin Guinney. \u0026ldquo;GSVA: Gene Set Variation Analysis for Microarray and RNA-Seq Data.\u0026rdquo; BMC Bioinformatics 14, no. 1 (January 16, 2013): 7. https://doi.org/10.1186/1471-2105-14-7.\u003c/li\u003e\n\u003cli\u003eRitchie, Matthew E., Belinda Phipson, Di Wu, Yifang Hu, Charity W. Law, Wei Shi, and Gordon K. Smyth. \u0026ldquo;Limma Powers Differential Expression Analyses for RNA-Sequencing and Microarray Studies.\u0026rdquo; Nucleic Acids Research 43, no. 7 (April 20, 2015): e47. https://doi.org/10.1093/nar/gkv007.\u003c/li\u003e\n\u003cli\u003e\u0026ldquo;Ggplot2: Elegant Graphics for Data Analysis | Journal of the Royal Statistical Society Series A: Statistics in Society | Oxford Academic.\u0026rdquo; Accessed May 6, 2024. https://academic.oup.com/jrsssa/article/174/1/245/7077675.\u003c/li\u003e\n\u003cli\u003eLangfelder P, Horvath S (2008). \u0026ldquo;WGCNA: an R package for weighted correlation network analysis.\u0026rdquo; _BMC Bioinformatics_, 559.\u0026lt;https://bmcbioinformatics.biomedcentral.com/articles/10.1186/1471-2105-9-559\u0026gt;.\u003c/li\u003e\n\u003cli\u003eTibshirani, Robert. \u0026ldquo;Regression Shrinkage and Selection Via the Lasso.\u0026rdquo; Journal of the Royal Statistical Society: Series B (Methodological) 58, no. 1 (1996): 267\u0026ndash;88.https://doi.org/10.1111/j.2517-6161.1996.tb02080.x.\u003c/li\u003e\n\u003cli\u003eFriedman J, Tibshirani R, Hastie T (2010). \u0026ldquo;Regularization Paths for Generalized Linear Models via Coordinate Descent.\u0026rdquo; _Journal of Statistical Software_, *33*(1), 1-22. doi:10.18637/jss.v033.i01 \u0026lt;https://doi.org/10.18637/jss.v033.i01\u0026gt;.\u003c/li\u003e\n\u003cli\u003eKassambara A, Kosinski M, Biecek P (2021). _survminer: Drawing Survival Curves using \u0026apos;ggplot2\u0026apos;_. R package version 0.4.9,\u0026lt;https://CRAN.R-project.org/package=survminer\u0026gt;.\u003c/li\u003e\n\u003cli\u003ePaul Blanche, Jean-Francois Dartigues, Helene Jacqmin-Gadda (2013). Estimating and Comparing time-dependent areas under receiver operating characteristic curves for censored event times with competing risks. Statistics in Medicine,32(30), 5381-5397. URLhttp://onlinelibrary.wiley.com/doi/10.1002/sim.5958/full .\u003c/li\u003e\n\u003cli\u003eHarrell Jr FE (2023). _rms: Regression Modeling Strategies_. R package version 6.7-1, \u0026lt;https://CRAN.R-project.org/package=rms\u0026gt;.\u003c/li\u003e\n\u003cli\u003eTherneau T (2023). _A Package for Survival Analysis in R_. R package version 3.5-7, \u0026lt;https://CRAN.R-project.org/package=survival\u0026gt;.\u003c/li\u003e\n\u003cli\u003eTerry M. Therneau, Patricia M. Grambsch (2000). _Modeling Survival Data: Extending the Cox Model_. Springer, New York. ISBN 0-387-98784-3.\u003c/li\u003e\n\u003cli\u003eXu, Liwen, Chunyu Deng, Bo Pang, Xinxin Zhang, Wei Liu, Gaoming Liao, Huating Yuan, et al. \u0026ldquo;TIP: A Web Server for Resolving Tumor Immunophenotype Profiling.\u0026rdquo; Cancer Research 78, no. 23 (December 1, 2018): 6575\u0026ndash;80. https://doi.org/10.1158/0008-5472.CAN-18-0689.\u003c/li\u003e\n\u003cli\u003eGibbs D (2024). _ImmuneSubtypeClassifier: An R package for classification of immune subtypes, in cancer, using gene expression data._. R package version 0.1.0.\u003c/li\u003e\n\u003cli\u003eGeeleher P, Cox NJ, Huang RS. Clinical drug response can be predicted using baseline gene expression levels and in vitro drug sensitivity in cell lines. Genome Biol. 2014 Mar 3;15(3):R47. doi: 10.1186/gb-2014-15-3-r47. PMID: 24580837; PMCID: PMC4054092.\u003c/li\u003e\n\u003cli\u003eBurridge, Keith, and Krister Wennerberg. \u0026ldquo;Rho and Rac Take Center Stage.\u0026rdquo; Cell 116, no. 2 (January 2004): 167\u0026ndash;79. https://doi.org/10.1016/S0092-8674(04)00003-0.\u003c/li\u003e\n\u003cli\u003eClark, R. A. \u0026ldquo;Activation of the Neutrophil Respiratory Burst Oxidase.\u0026rdquo; \u003cem\u003eThe Journal of Infectious Diseases\u003c/em\u003e 179 Suppl 2 (March 1999): S309-317. https://doi.org/10.1086/513849.\u003c/li\u003e\n\u003cli\u003eForman, Henry Jay, and Martine Torres. \u0026ldquo;Reactive Oxygen Species and Cell Signaling: Respiratory Burst in Macrophage Signaling.\u0026rdquo; \u003cem\u003eAmerican Journal of Respiratory and Critical Care Medicine\u003c/em\u003e 166, no. supplement_1 (December 15, 2002): S4\u0026ndash;8. https://doi.org/10.1164/rccm.2206007.\u003c/li\u003e\n\u003cli\u003eBedard, Karen, and Karl-Heinz Krause. \u0026ldquo;The NOX Family of ROS-Generating NADPH Oxidases: Physiology and Pathophysiology.\u0026rdquo; \u003cem\u003ePhysiological Reviews\u003c/em\u003e 87, no. 1 (January 2007): 245\u0026ndash;313. https://doi.org/10.1152/physrev.00044.2005.\u003c/li\u003e\n\u003cli\u003eGill, Jonathan, and Richard Gorlick. \u0026ldquo;Advancing Therapy for Osteosarcoma.\u0026rdquo; Nature Reviews. Clinical Oncology 18, no. 10 (October 2021): 609\u0026ndash;24. https://doi.org/10.1038/s41571-021-00519-8.\u003c/li\u003e\n\u003cli\u003eJu, Huai-Qiang, Jin-Fei Lin, Tian Tian, Dan Xie, and Rui-Hua Xu. \u0026ldquo;NADPH Homeostasis in Cancer: Functions, Mechanisms and Therapeutic Implications.\u0026rdquo; \u003cem\u003eSignal Transduction and Targeted Therapy\u003c/em\u003e 5, no. 1 (October 7, 2020): 231. https://doi.org/10.1038/s41392-020-00326-0.\u003c/li\u003e\n\u003cli\u003eWang, Xiaole, Kunfeng Chen, and Zhijian Zhao. \u0026ldquo;LncRNA OR3A4 Regulated the Growth of Osteosarcoma Cells by Modulating the miR-1207-5p/G6PD Signaling.\u0026rdquo; \u003cem\u003eOncoTargets and Therapy\u003c/em\u003e Volume 13 (April 2020): 3117\u0026ndash;28. https://doi.org/10.2147/OTT.S234514.\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"osteosarcoma, disulfidptosis, immunotherapy, prognostic model","lastPublishedDoi":"10.21203/rs.3.rs-4426108/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-4426108/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eDisulfidptosis is a novel form of programmed cell death discovered by Liu et al. It's initiated in cells highly expressing SLC7A11, especially in cancers. Our principal aim is to establish and validate a prognostic prediction model for osteosarcoma patients, potentially providing a fresh perspective on the characteristics of disulfidptosis in osteosarcoma and its treatment. Osteosarcoma cohorts obtained from the TARGET and GEO databases were classified into disulfidptosis-high/low-related groups to analyze the Differentially Expressed Genes (DEGs) using the ssGSEA method. DEGs were subsequently analyzed by the Weighted Gene Co-expression Network Analysis (WGCNA) method. Various machine learning algorithms, including the log-rank test, univariate Cox analysis, and LASSO algorithm, were employed, yielding 5 Disulfidptosis-Related Genes (DRGs). GSVA and ssGSEA, were also conducted to investigate the underlying mechanisms of disulfidptosis in osteosarcoma. We established a reliable disulfidptosis-related classification, aand our subsequent analysis has suggested intriguing disparities in the expression of the pentose phosphate pathway (PPP) and cytoskeleton regulation among the groups, indicating that the high-related group was more susceptible to disulfidptosis. 5 disulfidptosis-related genes were selected from the differentially expressed genes (DEGs) , and samples in the cohorts were divided into high-/low-risk groups based on the risk score. Functional analysis demonstrated significantly higher expression of the regulation of the cytoskeleton pathway in the high-risk group. Additionally, immune cell-associated pathways such as the T cell receptor signaling pathway and NOD/TOLL-like receptor signaling pathway showed significant decreases in the high-risk group. We then analyzed the infiltration of immune cells in the tumor microenvironment, revealing lower infiltration of almost every immune cell in the high-risk group. To gain insights into the clinical treatment of osteosarcoma patients, we also analyzed the differences in drug sensitivity between the risk groups, identifying 8 drugs that were more sensitive in the high-risk group.\u003c/p\u003e","manuscriptTitle":"Identification and validation of disulfidptosis-related signature to evaluate clinical outcomes, immune infiltration and drug sensitivity in osteosarcoma","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2024-06-12 17:39:25","doi":"10.21203/rs.3.rs-4426108/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"5a85ff2c-1492-4300-904d-1c400183548f","owner":[],"postedDate":"June 12th, 2024","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[{"id":32996364,"name":"Biological sciences/Biotechnology"},{"id":32996365,"name":"Biological sciences/Cancer"},{"id":32996366,"name":"Biological sciences/Computational biology and bioinformatics"},{"id":32996367,"name":"Biological sciences/Genetics"},{"id":32996368,"name":"Biological sciences/Immunology"}],"tags":[],"updatedAt":"2025-02-14T07:23:14+00:00","versionOfRecord":[],"versionCreatedAt":"2024-06-12 17:39:25","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-4426108","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-4426108","identity":"rs-4426108","version":["v1"]},"buildId":"7rjqhiLT3MXkJMwkYKINL","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}