Corrigendum:A telomere-related signature for predicting prognosis and immune microenvironment of osteosarcoma

Shihao Li¹Lina Zhang¹Haiyang Zhang^2*

• ¹Department of Orthopedics, Zibo Central Hospital West Campus, Zibo, China
• ²Department of Hand and Foot Surgery, Zibo Central Hospital, Shandong, China

1. Introduction Osteosarcoma is a primary malignant bone tumor that arises from interosseous leaf cells. Osteosarcoma most often occurs near the metaphyseal plates of the long bones. The most common sites are the femur (42%), tibia (19%) and humerus (10%). Other prevalent sites of occurrence include the craniofacial bones (8%) and the pelvis (8%), with primary osteosarcoma of extragnathic craniofacial bones accounting for 2% of all osteosarcomas (Cho et al., 2014). It is characterized by the rapid proliferation of tumor cells, which directly give rise to immature bone or bone-like tissue (Xu et al., 2019). As the most prevalent kind of primary malignant bone tumor in children and adolescents (Kebudi et al., 2016; Pastorino et al., 2023), osteosarcoma accounts for 2.4% of all juvenile cancers and approximately 20% of all primary bone tumors (Ottaviani and Jaffe, 2009). Osteosarcoma is very prone to metastasis in its early stages, with the lungs being the most common metastatic location, followed by the distal bones and lymph nodes (Cersosimo et al., 2020). In recent years, reactive oxygen species(ROS)-mediated therapy, such as photodynamic therapy (PDT), has emerged as a new promising treatment method for OS (Tan et al., 2022). Nanoparticles (CI@HSA NPs) encapsulating capsaicin (CAP) and the photosensitizer IR780 promote ferroptosis in osteosarcoma and improve the hypoxic microenvironment by releasing capsaicin, thereby improving the efficacy of photodynamic therapy (PDT) in osteosarcoma (Wang et al., 2024). However, in the early stage of osteosarcoma, there is lack of diagnostic markers and therapeutic targets. Thus, the osteosarcoma patients had a poor prognosis (Sugito and Kamal, 2022).Telomeres are specialized structures that form at the ends of chromosomes, serving to protect them from nucleolytic degradation and to distinguish them from DNA double-strand breaks (Fajkus et al., 2005). Previous studies have indicated that telomere anomalies may lead to a variety of disorders, including cancer, cardiovascular disease, myelodysplastic syndrome, dyskeratosis congenita, and Revesz syndrome (Maciejowski and de Lange, 2017; Savage, 2018; Armanios, 2022). Ennour-Idrissi and colleagues have demonstrated that longer telomeres are associated with better outcomes of breast cancer patients (Ennour-Idrissi et al., 2017). Mirabello have found that females with short telomeres may be predisposed to osteosarcoma (Mirabello et al., 2011). In telomerase-negative osteosarcoma U2OS cells, the deletion of the telomere-binding protein TPP1 can lead to telomere shortening, increased apoptosis and heightened sensitivity to radiation (Qiang et al., 2014). Furthermore, in both osteosarcoma Saos-2 and U2OS cell lines, the knockdown of pinX1, an intrinsic telomerase inhibitor and a putative tumor suppressor gene, results in telomere shortening, increased apoptosis and enhanced sensitivity to ionizing radiation (Li et al., 2015). Notably, oxidative stress caused by excess ROS accelerates telomere shortening (Fouquerel et al., 2019). The photosensitizer TMPipEOPP selectively binds to telomeric DNA G-quadruplex and cleaves it upon photo-irradiation through ROS production, leading to cancer cell death (Zhu et al., 2016). Therefore, further exploration of telomere-related genes in osteosarcoma is essential for early diagnosis and improved treatment strategies.Previous research has primarily focused on telomere length in cancer and its role in the prognosis of osteosarcoma. However, the impact of telomere-related genes on the progression of osteosarcoma has not been explored. Consequently, we developed a telomere-related signature based on these genes to predict outcomes and to investigate the potential implications of this signature in the immune microenvironment and drug sensitivity of osteosarcoma.2. Methods2.1 Data extractionIn the TARGET database (https://ocg.cancer.gov/programs/target), a total of 88 osteosarcoma samples with transcriptome data and clinical information were downloaded. After removing samples with incomplete clinical information, 84 samples with complete survival data (including overall survival and last survival status) were selected as the training set. The data were downloaded in matrix form, allowing for direct analysis without the need for preprocessing.In the Gene Expression Omnibus database (GEO, https://www.ncbi.nlm.nih.gov/geo/), the GSE21257, GSE152048, GSE16088, GSE28424, GSE39055, and GSE16091 datasets were downloaded. GSE21257 dataset consisted of 53 osteosarcoma samples, the data set were downloaded in matrix form. The probes can be converted into GeneSymbol corresponding to the platform information and can be used directly for analysis. GSE16088 comprised 14 osteosarcoma samples and 6 normal samples, while GSE28424 included 19 osteosarcoma samples and 4 normal samples. GSE39055 contained 37 osteosarcoma samples along with complete clinical data, and GSE16091 included 34 osteosarcoma samples with comprehensive clinical data. A total of 2086 telomere-related genes were collected from the TelNet (http://www.cancertelsys.org/telnet/, Table S1) (Li et al., 2022b). GSE152048 contained single-cell sequencing data of seven patients with osteosarcoma, from whom two samples, BC16, BC21, were selected for subsequent analysis. The raw data of each sample of GSE152048 were processed by the software Cell Ranger (https://www.10xgenomics.com/support/software/cell-ranger/, version6.0.2) to obtain the three files: barcodes.tsv, genes.tsv and matrix.tsv. We download these 3 files for each sample and used them directly for subsequent analysis without preprocessing.2.2 Lasso Cox regression analysisUnivariate Cox regression analysis was performed on telomere-related genes, and genes significantly correlated with the prognosis of osteosarcoma patients were selected using a threshold of P < 0.01. Subsequently, LASSO Cox regression analysis was performed using the glmnet package (version 4.1.7) (Friedman et al., 2010) in R language to further optimize the prognosis-related genes in osteosarcoma patients. Based on the filtered genes, the TELscore for each sample was calculated using the following formula:Coefi: The risk coefficient calculated by the lasso Cox model for each factor, Xi: the expression value of gene. Next, the R packages survival, survminer and two-sided log rank tests were used to determine the value of TELscore, and patients were divided into low TELscore and high TELscore groups according to the median of TELscore.2.3 Survival analysisBased on the Kaplan-Meier method, the overall survival rates of different groups were estimated using the “survival” package and “survminer” package (https://CRAN. R-project. org/package=survival) in the R language. The significance of differences in survival between the groups was assessed using the log-rank test. The Receiver Operating Characteristic (ROC) curves were generated using the “timeroc” package (version 0.4) in R (Blanche et al., 2013), and the area under the curve (AUC) values were subsequently calculated.2.4 Differentially expressed genes (DEGs)Based on the function package “limma” (version 3.52.4) (Ritchie et al., 2015) in R language. The DEGs between low TELscore and high TELscore groups were screened by |log2FC| > 0.5 and FDR < 0.05.2.5 Functional enrichment and immune cell infiltration analysesThe DEGs were subsequently submitted for enrichment analysis using Gene Ontology (GO), which included biological processes, molecular functions, and cellular component) and the Kyoto Encyclopedia of Genes and Genomes (KEGG) using the “ClusterProfiler” function package (version 4.7.1.2) (Yu et al., 2012) in R language. The significantly enriched pathways were screened by P < 0.05.The GSVA were performed using the c2.cp.kegg.v2023.1.Hs.symbols and c5.go.v2023.1.Hs.symbols gene sets from Molecular Signature Database (https://www.gsea-msigdb.org/gsea/msigdb). The significantly enriched pathways were screened by |logFC| < 0.5 and P < 0.05. The ssGSEA algorithm of the R package “GSVA” (version 1.46.0) (Charoentong et al., 2017) was used to calculate the abundance of 28 specific immune cell types. The stromal and immune cell scores of samples were calculated by the “estimate” function package (https://R-Forge.R-project.org/projects/estimate/). 2.6 Single cell data analysisSingle cell data processing was performed using the R language “Seurat” package (version 4.3.0). Single cell data were filtered to remove low-quality single cells with characteristic gene counts less than 200 and containing greater than 5% mitochondrial count. The scRNA-seq data were normalized by the function “normalizeddata”. Principal component analysis (PCA) analysis on the data was performed by the “RunPCA” function. Unsupervised clustering of major cell subtypes was performed by the findclusters function in Seurat and visualized by tSNE. Cell-type annotation was performed using the singleR software package and refined with manual annotation. DEGs between the two TELscore groups of osteosarcoma cells were calculated using “findmarkers” function. The “monocle” function package (version 2.26.0) was used for pseudotime analysis. The R package “Italk” (https://github. com/Coolgenome/iTALK, version 0.1.0) was applied to explore the cell communication atlas.2.7 Drug sensitivity analysisDrug sensitivity analysis was performed using the oncopredict package (Maeser et al., 2021) in R language. The GDSC2_Expr expression matrix, which contains data for 17,419 genes across 805 cell lines, along with the GDSC2_Res drug sensitivity dataset, which provides IC50 values for 198 drugs in these cell lines, were employed as the training set. the calcPhenotype function was applied to compute the predicted IC50 value for the sample based on its expression matrix. 2.8 Survival analysisThe “survival” package and “survminer” package in R were utilized to estimate the overall survival rates of various groups using the Kaplan-Meier method. The log-rank test was employed to assess the significance of differences in survival rates among these groups. Additionally, the R language`s pROC package (Robin et al., 2011) was used to generate the ROC curve and calculate the AUC value. Furthermore, the timeROC package (Blanche et al., 2013) was utilized to generate the timeROC curve and compute its AUC value.2.9 Cell culture and transfectionThe osteoblast cell line, hFOB1.19 was purchased from the Cell Bank of the Committee on Type Culture Collection of the Chinese Academy of Sciences (Shanghai, China). Two osteosarcoma cell lines (HOS, U-2OS) were purchased from the Procell Life Science & Technology Co., Ltd. (Wuhan, China), and osteosarcoma cell line, SAOS-2 was purchased from the iCell Bioscience Inc (Shanghai, China). hFOB1.19 cells Vero cell was cultured in DMEM medium (PM150270, Procell, Wuhan, China) supplemented with 10% FBS and 1% PS at 37°C in 95% air/5% CO2. HOS cells were maintained in MEM (PM150410, Procell, Wuhan, China) with 10% FBS and 1% P/S, in 37 °C, 5% CO2 cell culture incubator. U-2OS and SAOS-2 cells were cultured in McCoy’s 5A medium (PM150710, Procell, Wuhan, China) with 10% FBS and 1% P/S at 37˚C in a humidified atmosphere of 5% CO2 and 95% air.The SAOS-2 cells were transfected with OE-NC and OE-MAP7, respectively, using Lipocat2000C transfection reagent (AQ11669, Beijing Aoqing Biotechnology Co., Ltd, China) according to the manufacturer’s protocol. 2.10 qRT-PCR assayTriQuick Reagent (R1100, Solarbio, China) was utilized to extract total RNA from the cells. Reverse transcription was carried out using the Evo M-MLV Reverse Transcription Reagent Master Mix (AG11706-S, Accurate Biology, Changsha, China). Subsequently, the qRT-PCR assay was performed with the SuperStar Universal SYBR Master Mix (CW3360M, Jiangsu Cowin Biotech Co., Ltd., Jiangsu, China) on a real-time fluorescence quantitative PCR instrument. The primer sequences are listed in Table 1. The thermal cycling program was as follows: pre-denaturation at 95°C for 30 seconds, followed by 40 cycles of denaturation at 95°C for 10 seconds and annealing/extension at 60°C for 30 seconds. GADPH was used as the internal reference. mRNA expression levels were quantified using the 2-ΔΔCT method. 2.11 Western blotting (WB) analysisThe protein was extracted using RIPA buffer (R0010, Solarbio, Beijing, China) from cells. The lysates were centrifuged (12,000 rpm, 10 min, 4 °C) to collecte supernatant. The protein was separated by SDS-PAGE and then transferred onto PVDF membrane with a constant current of 250mA. The membranes were blocked in 5% nonfat dry milk for 2 h and incubated with primary antibody. Subsequently, the membranes were washed with 1×TBST and incubated for 1 h with secondary antibody. The primary antibodies used in this study was MAP7 Polyclonal antibody (1: 1000, 13446-1-AP, Proteintech) and GAPDH Monoclonal antibody (1:10000, 60004-1-Ig, Proteintech). The secondary antibody was Horseradish enzyme-labeled goat anti-mouse IgG (H+L) (1:10000, ZB-2301, Beijing Zhongshan Jinqiao Biotechnology Co., Ltd., China). Finally, the protein bands were detected using a fully automated chemiluminescence image analysis system (Chemi6000, Clinx, Shanghai, China). 2.12 Cell Counting Kit-8 (CCK-8) assayThe CCK-8 assay was performed using the CCK-8 reagent (C0037, Beyotime Biotechnology). SAOS-2 cells transfected with OE-NC and OE-MAP7, along with control cells, were seeded into 96-well plates at a density of 5,000 cells per well. After 24, 48, and 72 hours, the cells were incubated with CCK-8 reagent at 37°C for 2 hours. Subsequently, the absorbance of each well was measured at 450 nm using a microplate reader.2.13 Cell invasion and migrationThe Transwell assay (Costa, 3422) was utilized to assess the migration and invasion capabilities of the cells. A total of 100 µL of cell suspension (5×105 cells/mL) was added to the top chamber, which was either coated with Matrigel (356234, Corning) or left uncoated. Subsequently, 600 µL of culture medium supplemented with 20% serum was added to the lower chamber, and the cells were incubated for 48 hours. After incubation, the cells were stained with a 0.1% crystal violet solution (Q/12GF, Tianjin Guangfu Chemical Industry Institute, China) for 1 hour and then imaged using an inverted microscope (IMT-2, Olympus).2.14 Statistical analysisWilcoxon rank sum test was applied to compare gene expression differences among different groups. Single sample Kolmogorov-Smirnov test was used to test whether the data conforms to the normal distribution, and Pearson/spearman correlation analysis was performed using the R language “cor” function. Differences were considered statistically significant when p < 0.05. R software version 4.2.2 was used for all of the aforementioned statistical studies. Moreover, all experimental data were statistically analyzed using GraphPad Prism 9.5.0 software and were expressed as the mean ± standard deviation (SD). Each experiment was repeated at least three times with three replications per experiment. The unpaired two-tailed t-test was used for comparison between two groups, and the difference was statistically significant when p < 0.05.3. Results3.1 Construction of the telomere-related signature in osteosarcoma Firstly, in the training set TARGET cohort (including 84 osteosarcoma samples with complete survival information), we performed the univariate Cox regression analysis using the 2086 telomere-related genes as continuous variables, and calculated the hazard ratio (HR) of each gene. Among these, 85 telomere-related genes found to be significantly associated with the prognosis of osteosarcoma patients (Figure 1A, P < 0.01). Furthermore, LASSO Cox regression analysis identified 10 genes (SP110, HHAT, TUBB, MORC4, TERT, PPARG, MAP3K5, PAGE5, MAP7, CAMK1G) that exhibited a remarkable correlation with the prognosis of osteosarcoma patients (Figure 1B, minimum lambda value). Subsequently, we weighted the expression of these 10 genes with the regression coefficients from the lasso Cox regression analysis and established telomere-related signature to predict survival of patients. TELscore = (-0.28285616)×Express Value of SP110+(-0.01519659)×Express Value of HHAT+(-0.04605673)×Express Value of TUBB+(-0.19292665)×Express Value of MORC4+0.13670374×Express Value of TERT+(-0.03302422)×Express Value of PPARG+(-0.05545063)×Express Value of MAP3K5+(-0.05689784)×Express Value of PAGE5+(-0.04040167)× Express Value of MAP7+(-0.01099716)×Express Value of CAMK1G. In TARGET cohort, the high TELscore group was linked to inferior prognosis of osteosarcoma patients compared to low TELscore group (Figure 1C), which was also found in GSE21257 dataset (Figure 1D). Moreover, the ROC analysis showed that the AUC value of 1-, 3-, and 5-year overall survival in training set TARGET cohort were 0.81, 0.90 and 0.90, respectively (Figure 1E). In validation set GSE21257, the AUC value of 1-, 3-, and 5-year overall survival in training set TARGET cohort were 0.59, 0.56 and 0.70, respectively (Figure 1F). These findings suggested that the telomere-related signature could effectively predict the prognosis of osteosarcoma patients in TARGET cohort. 3.2 Prognostic value of the telomere-related signatureIn the TARGET cohort, we discovered that the TELscore was not significantly differential between the osteosarcoma patients with age ≧18 and osteosarcoma patients <18 years, female and male osteosarcoma patients (Figure 2A, 2B). Compared to non- metastatic osteosarcoma samples, the TELscore was remarkably increased in metastatic samples (Figure 2C). Moreover, we found that in osteosarcoma patients with age < 18 years, the high TELscore was related to worse prognosis of patients (Figure 2D). In the female and male groups, the osteosarcoma patients in high TELscore group exhibited inferior prognosis (Figure 2E, 2F). In metastatic and non-metastatic osteosarcoma patients, the high TELscore group was linked to a poor prognosis (Figure 2G, 2H). These findings suggested that the TELscore was related to metastasis and inferior prognosis of osteosarcoma.3.3 Functional annotation of the telomere-related signature In the TARGET cohort, we identified a total of 11,523 DEGs between the high and low TELscore group, including 43 up-regulated genes and 11,480 down-regulated genes (Figure 3A, 3B, high vs. low). The KEGG enrichment analysis revealed that the Glycosphingolipid biosynthesis-lacto and neolacto series, Carbohydrate digestion and absorption, and mineral absorption were significantly enriched in the high TELscore group, while a total of 158 signaling pathways were notably enriched in the Low TELscore group (Figure 3C, Table S2). The GO enrichment analysis indicated a total of 168 signaling pathways were markedly enriched in the high TELscore group, whereas 2,829 signaling pathways were dramatically enriched in the low TELscore group (Figure 3D, Table S2). Moreover, GSVA demonstrated that the score of 43 signaling pathways were significantly different between the high and low TELscore groups (Figure 3E, Table S2). 3.4 Predictive value of the telomere-related signature in the immune microenvironment of osteosarcomaIn the TARGET cohort, we calculated the correlation between TELscore and the abundance of 28 specific immune cell types. Our analysis revealed that TELscore exhibited a significantly negative correlation with six immune cells, including Central memory CD8+T cells and myeloid-derived suppressor cells. Conversely, TELscore showed a notably positive association with five immune cells, such as eosinophils and type 17 T helper Cells (Figure 4A, P < 0.05). Additionally, in the Xcell database (https://xcell.ucsf.edu/), TELscore was positively correlated with the relative proportions of four immune cell infiltrations (CD4 Tcm, MEP, plasma cells, and Pro B cells) and negatively associated with the relative proportions of seven immune cell infiltrations, including M2 Macrophages and Macrophages (Figure 4B, Table S3).Furthermore, we calculated the immune scores for high TELscore and low TELscore groups within the TARGET cohort. The ESTIMATE Score, Immune Score, and Stromal Score levels were lower in the high TELscore group compared to the low TELscore group (Figure 4C). Moreover, TELscore was negatively correlated with the ESTIMATE Score, Immune Score, and Stromal Score (Figure 4D)3.5 Single cell sequencing analysis for the telomere-related signature in osteosarcomaWe obtained a gene expression map from the single cell sequencing data set GSE152048. We performed PCA analysis to reduce dimensionality using 5,000 variable genes and identified 24 cell clusters using Seurat (Figure 5A). The results of the cell annotation were presented in Figure 5B. The telomere-related signature score levels in the identified cells were presented in Figure 5C. We identified DEGs between the high- and the low TELscore groups in the GSE152048 dataset and performed the GO and KEGG enrichment analyses to obtain more function information. The KEGG enrichment analysis showed that 11 and 23 signaling pathways were remarkably enriched in the high TELscore group and the low TELscore group, respectively (Figure 5D, Table S4). The GO enrichment analysis showed that 233 and 429 signaling pathways were observably enriched in the high TELscore group and the low TELscore group, respectively (Figure 5E, Table S4). Furthermore, we performed pseudotime analysis on osteosarcoma cells and identified five cell states (Figure 5F, left). As pseudotime increased (Figure 5F, middle), the osteosarcoma cells exhibited a tendency to increase TELscore (Figure 5F, right)3.6 Cell communication pattern of the telomere-related signatureSubsequently, we analyzed the different cellular signaling pathways of checkpoints, cytokines, growth factors, others between the the high TELscore and the low TELscore groups of osteosarcoma cells and microenvironment cells. In low TELscore group, TNFSF9, CD24 were the most active signaling pathways of checkpoints in osteosarcoma cells (Figure 6A, 6B). In the low TELscore group, ITGB1 and CCL3L1 were the most active signaling pathways of cytokines in osteosarcoma cells (Figure 6C, 6D). In the high TELscore group, TGFB was the most active signaling pathways of growth factors in osteosarcoma cells (Figure 6E, 6F). In the high TELscore group, LRP1, ITGB1, IBSP, COL1A1 and COL1A2 were the most active signaling pathways of others in osteosarcoma cells (Figure 6G, 6H).3.7 The predictive value of telomere-related signature in the drug sensitivity of osteosarcomaIn the TARGET cohort, we analyzed the correlation between TELscore and drug-sensitivity using R language “oncopredict” (version 0.2). The TELscore exhibited a prominent positive correlation with the IC50 values of 154 drugs, such as AZ960_1250, AMG.319_2045, Ruxolitinib_1507 and XAV939_1268 (Figure 7A). Conversely, it demonstrated a significant negative association with the IC50 values of SB505124_1194, ABT737_1910, BI.2536_1086 and Vorinostat_1012 (Figure 7B, Table S5, P < 0.05). Additionally, we analyzed the inference score of 10 genes using the CTD database (http://ctdbase.org/). All 10 genes were identified as targets for osteosarcoma (Figure 7C-7L), with MAP7, MAP3K5, PPARG, TERT, MORC4, TUBB, SP110, and HHAT demonstrating significant relevance to the development and progression of osteosarcoma.3.8 Low MAP7 expression was correlated with a poor prognosis of osteosarcoma patientsSubsequently, we analyzed the prognostic role of MAP7 in osteosarcoma patients and found that the expression of MAP7 was downregulated in osteosarcoma samples compared to normal samples in the GSE16088 (Figure 8A) and GSE28424 datasets (Figure 8B). In the TARGET and GSE39055 cohort, we divided osteosarcoma patients into MAP7high and MAP7low groups according to the median expression value of MAP7. The Kaplan-Meier survival analysis showed that the MAP7low group had a worse prognosis than the MAP7high groups in the TARGET (Figure 8C) and GSE39055 cohorts (Figure 8D). Furthermore, we discovered that the AUC values of MAP7 at 1-, 3-, and 5-years were 0.857, 0.682, and 0.654 in the TARGET cohort, respectively (Figure 8E). The AUC values in the GSE16091 cohort were 0.714, 0.67, and 0.6 for 1, 3, and 5 years, respectively. (Figure 8F). In the GSE39055 cohort, the AUC values were 0.644 (1-year), 0.767 (3-years), and 0.619 (5-years) (Figure 8G). In the GSE21275 cohort, the AUC values were 0.790 (3-years) and 0.771 (5-years) (Figure 8H). These results suggested that MAP7 was closely correlated with prognosis of osteosarcoma patients. 3.9 Overexpression of MAP7 might inhibit tumor proliferation, migration and invasion of osteosarcoma cellsFinally, to explore the role of MAP7 in the progression of osteosarcoma, we first confirmed the expression of MAP7 mRNA in several osteosarcoma cell lines. Compared to hFOB1.19 cells, the level of MAP7 mRNA expression was significantly decreased in HOS, U-2OS, and SAOS-2 cells (Figure 9A). Furthermore, we constructed MAP7 overexpression model in SAOS-2 cells, and found that MAP7 expression was remarkably increased in OE-MAP7 group compared to ON-NC group (Figure 9B, 9C). Moreover, overexpression of MAP7 significantly reduced cell proliferation (Figure 9D), and the ability of cell migration and invasion in SAOS‐2 cells (Figure 9E). These finding suggested that overexpression of MAP7 might inhibit tumor proliferation, migration and invasion of osteosarcoma cells.4. Discussion Numerous studies have been demonstrated the pivotal role of telomeres in the intricate processes of cancer development and progression. It has been reported that telomere shortening may act as a tumor suppressor by inhibiting cell growth, while also leading to widespread genomic instability, which promotes cancer development (Okamoto and Seimiya, 2019; Li et al., 2022c). Herein, a comprehensive understanding of telomeres in cancer is essential for the development of innovative and effective therapeutic strategies. Although many studies have focused on telomere length in cancer and its implications for osteosarcoma prognosis, there has been a lack of comprehensive examination regarding the involvement of telomere-related genes in the prognosis of osteosarcoma. In the present study, we developed a telomere-related signature for osteosarcoma patients based on ten telomere-related genes (SP110, HHAT, TUBB, MORC4, TERT, PPARG, MAP3K5, PAGE5, MAP7, CAMK1G). The TUBB protein was found to be down-regulated in osteosarcoma samples (Li et al., 2010). TERT was highly expressed in cisplatin-resistant osteosarcoma cells, where it shuttles from the nucleus to the mitochondrion in response to cisplatin therapy, thereby suppressing cisplatin-induced apoptosis in osteosarcoma cells. This mechanism may be particularly crucial in combating drug resistance (Zhang et al., 2017). Xie et al. indicated that the knockdown of TERT could decrease proliferation and inhibit the invasion and migration of osteosarcoma 143B and U2OS cells (Xie et al., 2023). PPARG has been identified as a hub gene in the onset of osteosarcoma (Li et al., 2022a), and the knockout of PPARG significantly represses the growth of osteosarcoma cells (Yuan et al., 2023). The expression of MAP3K5 was reported to be inversely correlated with the survival risk of osteosarcoma (Li et al., 2020). Compared to normal human osteoblast hFOB1.19 cells, MAP3K5 expression was significantly decreased in osteosarcoma cells, and its knockdown promoted the migration of these cells (Yang et al., 2023). Accordingly, in osteosarcoma, TUBB and MAP3K5 were believed to function as tumor suppressors, actively inhibiting tumor growth and progression, while TERT and PPARG appear to act as cancer promoters, exacerbating the development of osteosarcoma.Although the roles of SP110, HHAT, MORC4, PAGE5, MAP7, and CAMK1G in osteosarcoma have rarely been reported, their involvement in other types of cancer has been revealed. In hepatitis B virus-induced hepatocellular carcinoma, high SP110 expression was correlated with shorter survival rate of patients (Sengupta et al., 2022). HHAT knockdown suppresses Shh pathway activation and the development of non-small cell lung cancer cells in vitro, as well as tumor growth in vivo in mice xenografts (Rodriguez-Blanco et al., 2013). MORC4 was found to be overexpressed in breast cancer tissues, its downregulation by miR-193b-3p, influencing the proliferation of breast cancer cell proliferation via regulating apoptosis (Yang et al., 2019). PAGE5 was reported to be as mediators of chemoresistance in human small cell lung cancer (Wollenzien et al., 2023). CAMK1G has been reported to be significantly predictive of favorable overall survival of patients with prostate adenocarcinoma (Zhao et al., 2020). MAP7 has been shown to be highly in breast cancer and cervical cancer, and it can enhance the invasion and migration of breast cancer and cervical cancer cells (Zhang et al., 2020; Wang et al., 2022; Wang et al., 2023). MAP7 (ensconsin, E-MAP-115) is a ubiquitous microtubule-associated protein that organizes the microtubule cytoskeleton in mitosis and neuronal branching (Chaudhary et al., 2019). It has been reported that dysregulation of microtubule dynamics is associated with cancer development (Tala et al., 2014; Verma et al., 2023), suggesting that microtubule-targeting drugs are the basis of cancer chemotherapy. Lin et al. have demonstrated that the high expression of abnormal spindle microtubule assembly in osteosarcoma promotes proliferation of osteosarcoma cells (Lin et al., 2020). Thus, we hypothesized that MAP7 might also play a critical role in the progression of osteosarcoma. Accordingly, to investigate the role of MAP7 in osteosarcoma progression, we established a MAP7-overexpressing osteosarcoma cell model in the SAOS-2 cell. We discovered that MAP7 expression was significantly reduced in osteosarcoma cells, and low expression of MAP7 was associated with poor prognosis of osteosarcoma patients. Moreover, overexpression of MAP7 significantly reduced cell proliferation and the ability of cell migration and invasion in osteosarcoma cells. These results indicated that MAP7 might play an important role in the progression and prognosis of osteosarcoma.In both the training and validation sets, the high TELscore group was correlated with a poor prognosis of osteosarcoma patients. The AUCs for 1-, 3-, and 5-year overall survival in the training set TARGET cohort were 0.81, 0.90 and 0.90, respectively. The development of osteosarcoma has been observed to correlate with the age and gender of patients (Sadykova et al., 2020; Lee et al., 2021). Therefore, we analyzed the correlation of telomere-related signature with age and gender. We discovered that the TELscore was not significantly differential between the osteosarcoma patients with age ≧18 and osteosarcoma patients <18 years, female and male osteosarcoma patients. In addition, osteosarcoma has the property of systemic metastasis, especially pulmonary metastasis. Pulmonary metastasis remains the primary cause of death in osteosarcoma cases, over 90% of patients succumb to the effects of these lung metastases (Weeden et al., 2001; Kim et al., 2010). Accordingly, we explored the relationship between telomere-related signature and metastasis of osteosarcoma. The TELscore was remarkably increased in metastatic osteosarcoma patients compared to non-metastatic patients. Furthermore, in osteosarcoma patients with age < 18 years, female, male, metastatic and non- metastatic osteosarcoma patients, the high TELscore was correlated with worse prognosis. These findings indicated that telomere-related signature was not related to age and gender, and was correlated with metastasis of osteosarcoma patients, and telomere-related signature could effectively predict the prognosis of osteosarcoma patients.The immune microenvironment in osteosarcoma represents a highly intricate system, characterized by its complexity, remarkable plasticity, and a close association with immune escape, uncontrolled cell proliferation, and the metastasis of osteosarcoma cells (Tian et al., 2023). Dendritic cells (DCs), which are common antigen-presenting cells (APCs) originating from the bone marrow (Collin and Ginhoux, 2019), play a crucial role as effective antigen-presenting cells capable of stimulating immature T cells and generating specific immunological responses (Geissmann et al., 2010). DCs are present in all forms of bone sarcomas, and their quantity generally correlates with the extent of tumor-associated macrophage (TAM) invasion. In chondrosarcoma and chordoma, DCs are either absent or present in low numbers (Inagaki et al., 2016). In osteosarcoma, the infiltration of DCs is closely associated with autophagy (Zhang et al., 2021). Furthermore, osteosarcoma cells can develop variations that tolerate DCs and phagocytosis during the progression of the disease, leading to reduced DC activation and, ultimately, immune escape (Le et al., 2021). The infiltration of T lymphocytes and macrophages plays a significant role in osteosarcoma immunotherapy. In a study involving 16 primary osteosarcoma patients, infiltrating T lymphocytes were identified in biopsy tumor tissue and peripheral blood samples, with a higher level of positive T cells observed in tumor tissue compared to blood circulation (Han et al., 2016). It has been demonstrated that, within the tumor microenvironment, TAMs can interact with other immune cells to participate in local inflammatory regulation and contribute to drug resistance in osteosarcoma (Han et al., 2019).In the present study, the TELscore was positively associated with plasmacytoid dendritic cell, Type 17 T helper cell, and was negatively correlated with central memory CD8+T cell, M2 Macrophages, and regulatory T cell. And TELscore was negatively correlated with ESTIMATEScore, ImmuneScore and StromalScore. Accordingly, the telomere-related signature could regulate the immune microenvironment to affect the occurrence and progression of osteosarcoma. Moreover, we also found that the TELscore was positively correlated with the IC50 values of 154 drugs and negatively associated with 4 drugs. Considering the role of drug sensitivity in immunotherapy, we suggest that telomere related signatures may shed new light on optimizing immunotherapeutic strategies for osteosarcoma.A telomere-related signature was constructed in osteosarcoma patients based on the SP110, HHAT, TUBB, MORC4, TERT, PPARG, MAP3K5, PAGE5, MAP7, CAMK1G. The telomere-related signature had predictive values in the prognosis, immune microenvironment and drug sensitivity of osteosarcoma. Furthermore, MAP7 might be a prognostic marker for osteosarcoma patients.