Abstract Chondrosarcomas (CSs) are resistant to conventional chemotherapy and radiotherapy. Therefore, new therapeutic approaches are needed. The aim of this study was to validate the use of adenosine analogs as a new therapeutic strategy for the treatment of CS. Five adenosine analogs (aristeromycin, cladribine, clofarabine, formycin, and pentostatin) were evaluated in vitro on CS cell lines via both (two-dimensional) 2D cultures and three-dimensional (3D) alginate bead models. Cell viability was assessed by cell counting or ATP assays. Apoptosis was measured and cell cycle analyzed. The most promising compounds were further tested in vivo using a xenograft CS model in nude mice. Four analogs significantly reduced the viability of CSs. Among these, cladribine and clofarabine demonstrated potent efficacy in both 2D and 3D models by inducing apoptosis. Cladribine was further found to induce cell-cycle arrest, leading to apoptosis-mediated cell death. In vivo, both cladribine and clofarabine exhibited substantial antitumor effects in a xenograft model. In conclusion, cladribine and clofarabine, which have already been approved for clinical use in leukemia and multiple sclerosis, are promising candidates for the treatment of CS. Their efficacy in preclinical models suggests that these molecules could be repurposed for phase 2 clinical trials in patients with CS. Keywords: MT: Oligonucleotides: Therapies and Applications, adenosine analogs, cladribine, clofarabine, chondrosarcoma, apoptosis, cell death, tumor xenograft, proteomics, cell cycle Graphical abstract graphic file with name fx1.jpg [35]Open in a new tab __________________________________________________________________ Chondrosarcomas are resistant to conventional therapies. This study identifies the adenosine analogs cladribine and clofarabine as potent inducers of apoptosis and cell-cycle arrest in preclinical chondrosarcoma models. These clinically approved drugs exhibit significant in vitro and in vivo efficacy, supporting their potential repurposing for future clinical trials in chondrosarcoma. Introduction Chondrosarcomas (CSs) are primary malignant bone tumors characterized by the production of a cartilaginous matrix. They rank as the third most common primary malignant tumor after myelomas and osteosarcomas, accounting for 20%–27% of all malignant bone tumors in adults. Annually, they represent 2.1 to 4 new cases per million people.[36]^1^,[37]^2 According to the European Society for Medical Oncology guidelines, the gold standard for treating conventional CS, which constitutes approximately 85% of all CS, is surgical intervention. This may involve either bone curettage (a minimally invasive technique) or extensive ablation, which can include the amputation of bone and surrounding soft tissues, leading to increased morbidity. Furthermore, the feasibility of surgery depends on the location and stage of the tumor; in some cases, surgical options may be limited. Conventional CSs are notably resistant to radiotherapy and chemotherapy; however, while radiotherapy can serve as a palliative measure, the overall evidence for its effectiveness in patient studies is sparse because of the rarity of the disease. Consequently, ongoing research aims to discover new therapeutic approaches and enhance the efficacy of existing chemotherapies to provide better patient care. In this context, scientists are focusing on developing innovative anticancer therapies, particularly through the design of new or repositioned drugs, with purine analogs representing a significant category. Purines are essential components of deoxyribonucleic acid (DNA), ribonucleic acid (RNA), and other biomolecules, including adenosine triphosphate (ATP) and S-adenosyl-methionine (SAM). Due to their critical roles in various cellular processes, such as energy production, signaling, DNA synthesis, and repair, purines and their metabolites have been implicated in cancer development and in the mechanisms of radioresistance.[38]^3^,[39]^4 Therefore, targeting purine metabolism represents a promising therapeutic avenue for treating cancers, particularly bone cancers as well as other malignancies.[40]^5^,[41]^6 Since the 1950s, purine antimetabolites, including various purine analogs, have been developed to inhibit purine synthesis. These compounds share structural similarities with purines and can be incorporated into purine nucleotides and DNA, thus competing for essential cellular processes. They comprise a diverse group of molecules with varying mechanisms of action, indications, and pharmacokinetics. Notably, adenosine analogs have been developed over time to enhance their antineoplastic efficacy.[42]^7 The discovery of 3-deazaneplanocin A (DZNep) in the 1980s marked a significant advancement in this field. Numerous studies have demonstrated its anticancer effects across a wide range of cancers including CSs, as evidenced by our own in vitro and in vivo research.[43]^8^,[44]^9^,[45]^10 DZNep has also been shown to enhance chemosensitivity specifically in CSs in vitro.[46]^11 Given these findings, we hypothesized that other adenosine analogs may also exert therapeutic effects on CSs. To investigate this potential, we selected five additional adenosine analogs: two natural analogs, aristeromycin and formycin, along with three clinically used analogs for treating various leukemias: pentostatin, cladribine, and the cladribine derivative clofarabine.[47]^12^,[48]^13^,[49]^14^,[50]^15 Aristeromycin, a unique carbocyclic nucleoside antibiotic derived from Streptomyces citricolor, has been shown to reduce cell viability in prostate cancer cells and, like DZNep, inhibits S-adenosylhomocysteine hydrolase (SAHH), an enzyme critical to the transmethylation cycle and epigenetic processes.[51]^16 Formycin A, isolated from Streptomyces kaniharaensis, is known for its potent antibiotic properties and has also been recognized as a ribonucleoside analog with significant anti-HIV-1 activity without cytotoxic effects.[52]^17^,[53]^18 Pentostatin, cladribine, and clofarabine are established adenosine analogs used clinically for treating various leukemias.[54]^19^,[55]^20 Pentostatin acts as an adenosine deaminase (ADA) inhibitor with demonstrated antitumor activity.[56]^21^,[57]^22 Cladribine and clofarabine, in their active forms, inhibit ribonucleotide reductase, an enzyme crucial for synthesizing deoxyribonucleotide triphosphates (dNTPs). This inhibition results in reduced dNTP pools, impairing DNA synthesis and repair, ultimately leading to increased DNA damage and triggering programmed cell death. Clofarabine has also been shown to have antitumor effects on Ewing’s sarcoma and other sarcomas.[58]^23 Given these insights, we investigated the effects of these adenosine analogs on cell viability, cell death, and overall antitumor efficacy in CS. Results Cytotoxic effects of adenosine analogs in SW1353 CS cell line To evaluate the efficacy of various adenosine analogs, we first investigated their cytotoxic effects on the widely studied CS cell line SW1353. The cells were treated with increasing concentrations of aristeromycin, cladribine, formycin, or pentostatin for seven days. Microscopic examination of treated SW1353 cells revealed significant morphological changes, including reduced cell density and noticeable cell shrinkage at 1 μM concentrations of aristeromycin, cladribine, and formycin compared with those in control cells ([59]Figure 1A). Figure 1. [60]Figure 1 [61]Open in a new tab Cytotoxic effects of adenosine analogs in SW1353 chondrosarcoma cell line SW1353 chondrosarcoma cells and chondrocytes were treated with increasing doses of aristeromycin, cladribine, formycin, or pentostatin. After 7 days of treatment, microscopic observations were performed at 10× magnification for pentostatin and 4× magnification for the other three analogues (A), and cell viability was assessed by counting viable cells (B) or measuring ATP levels (C). The experiments were independently repeated at least 3 times for each cell line and treatment. The graphs represent the means of the experiments. Cell viability was further assessed using acridine orange (AO) and 4′,6-diamidino-2-phénylindole (DAPI) staining. This analysis demonstrated that aristeromycin, cladribine, and formycin decreased cell viability in a concentration-dependent manner in the SW1353 line ([62]Figure 1B). In contrast, primary chondrocytes were significantly less sensitive to these three analogs. ATP quantification using the CellTiterGlo reagent corroborated the cytotoxic effects of aristeromycin, cladribine, and formycin on CS cells ([63]Figure 1C). In contrast, pentostatin did not induce any cytotoxic effects in either CS cells or chondrocytes after seven days of treatment. Efficacy of adenosine analogs across multiple CS cell lines To better characterize tumor heterogeneity, we tested the three effective adenosine analogs (aristeromycin, cladribine, and formycin) on three additional human CS lines: JJ012, CH2879, and FS090. After 48 h of treatment, the cells were stained with AO and DAPI, and cell viability was quantified ([64]Figure 2). As anticipated, aristeromycin, cladribine, and formycin effectively reduced viability in SW1353 cells without affecting chondrocyte viability. Importantly, similar responses were observed in the other CS cell lines. Specifically, treatment with aristeromycin (10 μM), cladribine (1 μM), or formycin (1 μM) resulted in reduced viability across all four CS lines tested. Figure 2. [65]Figure 2 [66]Open in a new tab Efficacy of adenosine analogs across multiple chondrosarcoma cell lines Four chondrosarcoma cell lines (SW1353, JJ012, FS090, and CH2879) and chondrocytes were treated with increasing concentrations of aristeromycin, cladribine, or formycin. After 2 days of treatment, cell viability was assessed by counting viable cells. The results of three independent experiments are shown. Histograms represent the percentage of live cells after treatment compared to controls. The data are expressed as the means ± SEM. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, and ∗∗∗∗p < 0.0001. Apoptosis induced by aristeromycin and cladribine Given the comparable effects previously observed, we opted to focus subsequent analyses on the two most common CS lines: SW1353 and JJ012. To investigate whether the observed decrease in cell viability resulted from apoptotic cell death, we exposed the CS cell lines, as well as chondrocytes, to the three adenosine analogs for 48 h. Apoptosis was quantified using annexin V and propidium iodide staining. Both aristeromycin and cladribine significantly increased apoptosis in SW1353 and JJ012 cells ([67]Figures 3A and 3B). In contrast, formycin did not induce apoptosis in CS cell lines, suggesting that its cytotoxicity may be mediated through a different mechanism. Importantly, there was no increase in apoptosis observed in chondrocytes regardless of the analog tested. Figure 3. [68]Figure 3 [69]Open in a new tab Apoptosis induced by aristeromycin and cladribine and efficacy in 3D chondrosarcoma models SW1353 and JJ012 chondrosarcoma cells and chondrocytes were treated with aristeromycin, cladribine, or formycin. After 2 days of treatment, apoptotic cells were identified by annexin V and propidium iodide (PI) staining. The cytograms show representative results from three or four independent experiments (A). Histograms represent the percentage of apoptotic cells (early and late) obtained (B). Alginate beads containing JJ012 or SW1353 chondrosarcoma cells, or chondrocytes, were treated with aristeromycin, cladribine, or formycin 15 days after formation. After 7 days of treatment, cell viability was assessed via ATP measurement (C). The results of three independent experiments are shown. Data are expressed as the means ± SEM. ∗p < 0.05, ∗∗p < 0.01, and ∗∗∗∗p < 0.0001. Efficacy of cladribine in 3D CS models To increase the translational relevance of our findings, we evaluated the effects of aristeromycin, cladribine, and formycin in a three-dimensional (3D) alginate bead model that more closely mimics the tumor microenvironment and predicts the weak efficiency of cisplatin observed in vivo and in patients.[70]^10 After 14 days of culture, the alginate beads were treated with the adenosine analogs for an additional seven days ([71]Figure 3C). Cladribine was the only analog that significantly reduced viability in both CS cell lines without affecting chondrocytes. Aristeromycin caused a reduction in viability exclusively in the SW1353 line, whereas formycin reduced viability across all tested cell types, including chondrocytes. These findings prompted us to focus further investigations on cladribine. Proteomic analysis following cladribine treatment To elucidate the mechanisms underlying the effects of cladribine, we conducted a proteomic study. After 48 h of treatment, 36 proteins were found to be deregulated in the SW1353 cell line (false discovery rate [FDR] < 0.05), with 12 downregulated (33.3%) and 24 upregulated (66.7%) proteins identified ([72]Figure 4A; [73]Table S1). In the JJ012 cell line, a total of 1,275 proteins were deregulated after cladribine treatment (FDR < 0.05), of which 938 were highly significant (FDR < 0.001). This included 685 downregulated (53.8%) and 590 upregulated (46.2%) proteins ([74]Figure S1A; [75]Table S1). Figure 4. [76]Figure 4 [77]Open in a new tab Proteomic analysis following cladribine treatment in SW1353 Heatmap showing the effects of cladribine treatment on the SW1353 chondrosarcoma cell line after 48 h (A). Gene enrichment analysis was performed using ShinyGO. The bubble plot shows the top 15 significantly enriched terms for deregulated proteins in chondrosarcoma cells after 48 h of cladribine treatment. Gene Ontology (GO) terms are shown for Biological Process (BP) (B) and KEGG pathways (C). An FDR <0.05 was considered significant, and fold enrichment scores were used to evaluate the relevance of the GO terms and the enriched pathways. Analysis of these deregulated proteins revealed significant enrichment in pathways related to the cell cycle, purine metabolism, and overall metabolism, as identified through Gene Ontology (GO) Biological Process (BP), and Kyoto Encyclopedia of Genes and Genomes (KEGG) database analyses ([78]Figures 4B, 4C, [79]S1B, and S1C; [80]Tables S2 and [81]S3). Additional evaluations using GO terms Molecular Function, Reactome, and WikiPathways databases confirmed enrichment in cell cycle and metabolic pathways, including ribosomal and ribonucleoprotein activities, as well as pathways associated with DNA repair and the p53 signaling pathway, which are often linked to cell death ([82]Tables S4 and [83]S5). Collectively, these results suggest that cladribine may induce apoptosis by disrupting the cell cycle and influencing the p53 pathway. Alteration of cell cycle induced by cladribine Given that proteomic findings indicate a potential effect on the cell cycle, we examined the various phases of the cell cycle in SW1353 and JJ012 cell lines following 48 h of cladribine treatment. In SW1353 cells, cladribine treatment resulted in an increase in S-phase cells and a decrease in G2/M-phase cells, indicating S-phase cell-cycle arrest ([84]Figure 5A). Conversely, in JJ012, cladribine treatment led to an increase in G1-phase cells and a decrease in G2/M-phase cells, suggesting either a block in the transition from G1 to S phase or an acceleration of the transition from G2/M to G1 phase ([85]Figure 5B). Figure 5. [86]Figure 5 [87]Open in a new tab Changes in the cell cycle induced by cladribine After 24 h of treatment with cladribine, SW1353 or JJ012 cells were harvested, and the cell cycle distribution was analyzed via image-based cytometry (A and B, respectively). The cytograms show a representative result from three independent experiments. Histograms represent the distribution of cells across the different cell cycle phases. The data are expressed as the means ± SEM. ∗p < 0.05, ∗∗∗p < 0.001, and ∗∗∗∗p < 0.0001. In vivo efficacy of cladribine in tumor growth reduction To assess the in vivo efficacy of cladribine, we established heterotopic CS xenografts using JJ012 cells in nude mice. Once the tumors were palpable, cladribine (20 mg/kg) or vehicle (DMSO) was intraperitoneally administered three times per week for six weeks in the mice ([88]Figure 6A). The treatment was well tolerated, with no significant reductions in body weight observed ([89]Figure S2A) or signs of discomfort. Compared with vehicle, cladribine significantly slowed tumor growth, as evidenced by decreased tumor volume ([90]Figure 6B). Additionally, tumor weight was significantly reduced in the cladribine-treated mice ([91]Figure 6C). Figure 6. [92]Figure 6 [93]Open in a new tab In vivo efficacy of cladribine in tumor growth reduction A xenograft of JJ012 chondrosarcoma cells was subcutaneously implanted into nude mice, which were treated intraperitoneally with 20 mg/kg cladribine (n = 9) or vehicle (n = 4), three times per week for 42 days (A). The tumor volume was measured three times per week (B). After 42 days, tumors were collected and weighed (C). Data are expressed as the means ± SEM. ∗∗p <0.01 and ∗∗∗p <0.001. Evaluation of the antitumor effects of clofarabine, a cladribine derivative, in CSs Next, we were interested in a derivative of cladribine, the clofarabine, which has demonstrated enhanced efficacy in Ewing’s sarcoma mouse models.[94]^24 In vitro, clofarabine significantly decreased the viability of CS cells in both monolayer and 3D cultures without affecting the viability of chondrocytes. Furthermore, clofarabine induced apoptosis in CS cells ([95]Figure 7). Figure 7. [96]Figure 7 [97]Open in a new tab In vitro evaluation of antitumor effects of clofarabine, a cladribine derivative, in chondrosarcomas Chondrosarcoma cell lines and chondrocytes were treated with increasing doses of clofarabine. Cell viability was assessed after 7 days of treatment by ATP measurement using the CellTiter-Glo assay (A), or after 2 days of treatment by counting viable cells (B). Apoptosis was determined by annexin V and propidium iodide staining after 48 h of treatment. The cytograms show representative results from three independent experiments, and histograms represent the percentage of apoptotic cells (early and late) observed (C). Additionally, alginate beads containing chondrosarcoma cells or chondrocytes were treated with clofarabine, and cell viability was evaluated after 7 days of treatment via ATP measurement (D). The data are expressed as the means ± SEM. ∗p <0.05, ∗∗p <0.01, ∗∗∗p <0.001, and ∗∗∗∗p <0.0001. Furthermore, we assessed the antitumor effects of clofarabine in a CS xenograft model in nude mice. Following JJ012 cell implantation, the mice received either vehicle (DMSO) or clofarabine (20 mg/kg) intraperitoneally ([98]Figure 8A). The mice tolerated the treatment well, with no significant changes in body weight ([99]Figure S2B), and interestingly, clofarabine reduced tumor growth. The tumor volume in the clofarabine-treated mice remained stable throughout the 42-day treatment period and continued to be maintained for up to 10 days after treatment cessation ([100]Figure 8B). Moreover, compared with vehicle treatment, clofarabine tended to decrease tumor weight compared to vehicle treatment, with final weights recorded at 100.7 ± 34.8 mg for clofarabine-treated mice and 473 ± 181 mg for vehicle-treated mice (p = 0.07; [101]Figure 8C). Figure 8. [102]Figure 8 [103]Open in a new tab In vivo evaluation of antitumor effects of clofarabine JJ012 chondrosarcoma cells were implanted as subcutaneous xenografts into nude mice. Once tumors became palpable, intraperitoneal injections of 20 mg/kg clofarabine (n = 7) or vehicle (n = 4) were administered three times per week for 40 days (A). The tumor volume was measured with a caliper three times per week throughout the treatment period (B). On day 42, four mice treated with clofarabine were kept alive and monitored for an additional 10 days after treatment cessation to evaluate potential tumor recurrence. The remaining mice were sacrificed, and the tumors were weighed (C). ∗∗p <0.01, ∗∗∗p <0.001, and ∗∗∗∗p <0.0001. Discussion Prior studies have demonstrated that the adenosine analog DZNep effectively induces apoptosis both in vitro and in vivo in CS models. However, DZNep is not available for clinical use, posing significant challenges for human application in terms of time and rentability/cost. In this study, we identified two additional adenosine analogs—cladribine and clofarabine—that are currently utilized clinically for other malignancies. Both analogs significantly decreased CS viability in vitro and in vivo, promoted apoptotic activity, and induced cell-cycle arrest, positioning them as promising candidate drugs for the treatment of patients with CS. In our in vitro studies, cladribine and clofarabine emerged as adenosine analogs capable of effectively reducing cell viability and inducing apoptosis in all tested CS cell lines, while sparing normal chondrocytes. In contrast, the other analogs either failed to induce apoptosis across all the CS lines or exhibited toxicity toward chondrocytes. The variability in the efficacy of these adenosine analogs can be attributed to several factors. For example, it is noteworthy that pentostatin did not reduce survival in the SW1353 cell line, despite its proven effectiveness in hairy cell leukemia, where it is reported to have similar efficacy to cladribine.[104]^25^,[105]^26 Pentostatin specifically inhibits ADA, whereas cladribine is described as ADA-resistant. Nevertheless, both agents result in the accumulation of deoxyadenosine triphosphate (dATP), leading to DNA strand breaks and the activation of apoptotic pathways via p53 and cytochrome c release from mitochondria. The enhanced efficacy of cladribine can be attributed to its additional mechanism of action: its active triphosphorylated form (CdATP) integrates into DNA, inhibiting DNA polymerase β and disrupting DNA repair processes. This further DNA damage activates poly(ADP-ribose) polymerase (PARP), resulting in the depletion of cellular nicotinamide adenine dinucleotide and ATP.[106]^27 This PARP activation-induced necrosis could explain the superior effectiveness of cladribine in treating CS. While pentostatin-treated cells may evade cell cycle control, those exposed to cladribine experience halted cell cycle progression, rendering them incapable of DNA synthesis or repair, ultimately leading to apoptosis. Additionally, Uchiyama et al. reported that DZNep and its analog aristeromycin strongly inhibit SAHH in prostate cancer cells, resulting in reduced cell proliferation.[107]^16 However, unlike DZNep, aristeromycin did not consistently inhibit growth across all tested CS lines, particularly in the 3D models. This observation aligns with findings in prostate cancer cell lines, suggesting that the efficacy of aristeromycin may vary significantly.[108]^16 The structural similarity of aristeromycin to ATP, with a cyclopentyl modification replacing the furanose sugar, could contribute to its increased toxicity, particularly toward nontumor cells such as chondrocytes, as our results indicate. A similar toxicity profile was noted with formycin, which significantly compromised chondrocyte viability. Consequently, both aristeromycin and formycin may require the development of novel derivatives to reduce their adverse effects while maintaining therapeutic efficacy.[109]^21 In contrast to the previously mentioned analogs, cladribine and clofarabine demonstrated promising results in reducing cell viability by inducing S-phase arrest and blocking the G1/S transition, which in turn promoted apoptosis in CSs. This mechanism is consistent with the established effects of these agents in leukemia and Ewing’s sarcoma.[110]^5^,[111]^14^,[112]^23^,[113]^24^,[114]^28^,[115]^29 Importantly, both adenosine analogs exhibited selective toxicity, sparing nontumor cells. In vivo, both cladribine and clofarabine significantly reduced tumor volume. Notably, clofarabine showed superior efficacy. Tumors in clofarabine-treated mice did not increase in size throughout the 42-day treatment period, whereas tumors treated with cladribine, although they grew more slowly than those in the control groups, still exhibited some growth. Despite this, complete tumor regression was not achieved with clofarabine within the study’s dose and duration. Upon discontinuation of treatment, initial tumor regrowth was observed within ten days, indicating a need for continuous treatment or optimization of the administration protocol. Both cladribine and clofarabine were administered via intraperitoneal injection, a method that may not be optimal. Previous studies have shown that the oral administration of clofarabine can be more effective than the same dosage given intraperitoneally.[116]^24^,[117]^30 Additionally, higher doses of clofarabine might be necessary, as its effects on leukemia blast cells are dose-dependent. For example, higher doses resulted in sustained inhibition of DNA synthesis, whereas lower doses only allowed for partial recovery.[118]^31 In our study, we utilized a dose of 20 mg/kg, which may have been insufficient. Therefore, further efforts are needed to optimize the administration and dosing protocols for clofarabine in the treatment of CS. Additionally, investigating combinations of clofarabine with other cytotoxic agents could yield synergistic effects, enhancing therapeutic outcomes for patients with CS. Future studies should investigate whether these adenosine analogs can sensitize CS cells to radiotherapy or chemotherapy, which could be particularly relevant given the intrinsic resistance of CS to conventional treatments. Furthermore, the in vivo experiments were performed using the classical and widely employed xenograft model of CS, involving subcutaneous implantation in the flank or back of nude mice. While this model is simple and commonly used,[119]^32^,[120]^33 it presents several limitations. Notably, the subcutaneous injection of tumor cells does not accurately reflect the anatomical location of CSs in patients. Therefore, it would be relevant to repeat these experiments using more sophisticated models. For instance, patient-derived xenograft models could be particularly valuable, as they preserve the tumor microenvironment and better capture the biological heterogeneity and clinical characteristics of the original tumors.[121]^34 In addition, evaluating the efficacy of these compounds in immune-competent models would be essential to understand potential immunomodulatory effects and the involvement of the tumor immune microenvironment. In summary, we have identified three adenosine analogs—DZNep (in a previous report), cladribine, and clofarabine—that show efficacy against CS in vitro and in vivo. These agents hold promise as potential therapeutic options, as they effectively inhibit tumor growth with minimal toxicity. These findings suggest that cladribine, clofarabine and their derivatives could be valuable treatments for CS, a largely underexplored area in oncology. Further research is needed to clarify their molecular mechanisms and evaluate their therapeutic potential in clinical settings. Materials and methods Cell culture The human CS cell lines JJ012 (RRID:CVCL_D605) and FS090 were generously provided by Dr. J.A. Block from Rush University Medical Center.[122]^35 The CH2879 cell line (RRID:CVCL_9921) was kindly donated by Professor A. Llombart-Bosch from the University of Valencia, Spain.[123]^36 The CS cell line SW1353 (RRID:CVCL_0543) was obtained from the American Type Culture Collection (ATCC; Manassas, VA, USA). Human chondrocytes were isolated, as previously described,[124]^37 from the femoral heads of patients undergoing arthroplasty, following the acquisition of informed consent in accordance with local legislation and ethical guidelines established by the Comité de Protection des Personnes Nord Ouest III. All cells were cultured in Dulbecco’s modified Eagle’s medium (DMEM; LONZA, Levallois Perret, France), supplemented with 10% fetal bovine serum (FBS; Dutscher, Bernolsheim, France) and antibiotics (penicillin and streptomycin; LONZA). The cultures were maintained at 37°C in a humidified atmosphere with 5% CO[2]. Regular testing for mycoplasma contamination was conducted via polymerase chain reaction (PCR) techniques. For 3D cultures, the SW1353 and JJ012 cell lines were encapsulated in alginate beads, as previously described.[125]^10 Briefly, 2 million cells per mL were suspended in 1 mL of a 1.2% sodium alginate solution (Merck, Saint-Quentin-Fallavier, France). Beads were formed by dripping the cell suspension into a sterile 100 mM CaCl[2] solution (VWR, Fontenay-Sous-Bois, France). The beads were subsequently washed in 0.15 M NaCl solution (Sigma-Aldrich) and incubated in DMEM for two weeks prior to further treatments. Adenosine analog drugs Five molecules were selected for analogy with adenosine. Aristeromycin was obtained from Santa Cruz Biotechnology (Clinisciences, Nanterre, France). Pentostatin and formycin were provided by Sigma-Aldrich (Merck, Saint-Quentin-Fallavier, France). Cladribine and clofarabine were obtained from MedChem (Clinisciences). Aristeromycin, cladribine, and clofarabine were resuspended in dimethyl sulfoxide (DMSO, Dutscher). Pentostatin and formycin were dissolved in DMEM and Dulbecco’s phosphate-buffered saline with no calcium or magnesium (DPBS; LONZA) respectively. Assessment of cell viability Two methods were employed to measure cell viability: a cell staining assay using AO and DAPI (ChemoMetec, Denmark), and a CellTiter-Glo Luminescence Viability Assay Kit (Promega Corporation, Lyon, France). Cell staining assay The viability and cell count assays were conducted following the protocol recommended by the supplier (ChemoMetec). The cells were trypsinized to detach them from the culture surface and then resuspended in a suitable medium. A cell suspension was loaded into the Via2-Cassette tip for analysis. Inside the cassette, the cells were stained with two fluorescent dyes: AO, which stains all nucleated cells, and DAPI, which selectively stains nonviable cells. The samples were analyzed via an image-based cytometer, the NucleoCounter NC-3000 (ChemoMetec), utilizing the “Viability and Cell Count Assay” program. ATP measurement A CellTiter-Glo luminescence viability assay kit was used to quantify cell viability according to the manufacturer’s instructions. Briefly, the CellTiter-Glo reagent was added directly to the cell culture. This reagent lyses the cells and generates a luminescent signal through a reaction between luciferase and oxyluciferin, with the conversion to oxyluciferin being catalyzed by ATP produced by living cells. The resulting luminescent signal was captured and quantified to measure the amount of ATP present, which is directly correlated with the number of viable cells. Luminescence was measured using a Varioskan LUX (Thermo Fisher Scientific, Illkirch-Graffenstaden, France). The results are presented as dose-response inhibition curves, illustrating the relationship between treatment concentration and cell viability. Apoptosis assay To assess apoptosis, the cells were harvested and counted at a concentration of 2 × 10^5 cells per sample. Following centrifugation, the cells were resuspended and stained with 100 μL of annexin V binding buffer, 2 μL of Hoechst 33342, and 2 μL of annexin V-CF488A conjugate. After incubation (at 37°C for 15 min), the cells were washed twice to remove excess dye. The samples were subsequently stained again with 100 μL of annexin V binding buffer and 2 μL of propidium iodide to differentiate between viable and nonviable cells. The samples were analyzed using an image-based cytometer, the NucleoCounter NC-3000 (ChemoMetec), with the “Annexin V Assay” program. The analysis revealed the percentage of cells that underwent apoptosis, which included both the early and late apoptotic stages. The results were plotted to visualize the apoptotic response of the treated cells. CS xenograft model in nude mice Animal research was conducted in compliance with the current European Directive (2010/63/EU), which has been incorporated into national legislation (decree 87/848). All animal experiments adhered to internationally accepted ethical principles for laboratory animal use and care to minimize suffering, and all procedures were approved by the Ministry of Education and Research as well as the regional ethics committee (CENOMEXA, France; APAFIS number: 37292). Twenty-two NMRI nude mice (6 weeks old, males), obtained from Janvier Labs (Le Genest-Saint-Isle, France), were housed in ventilated cages under controlled conditions. The environment was specific pathogen-free, with a 12-h reversed light-dark cycle and a temperature maintained at 23 ± 2°C at the Centre Universitaire de Ressources Biologiques (CURB, Caen, France; accreditation number: C14118015). The mice had ad libitum access to food and water. Subcutaneous xenografts were established by injecting 1 million JJ012 cells suspended in 150 μL of Matrigel (Thermo Fisher Scientific) into the right flank of each mouse. When tumors were palpable (i.e., 19 days post-implantation), the mice were randomly divided into three groups for treatment: 5 mice treated with DMSO (vehicle control); 10 mice treated with cladribine at a dose of 20 mg/kg; and 7 mice treated with clofarabine at a dose of 20 mg/kg. Treatment was administered via intraperitoneal injection three times a week. The tumor dimensions were measured regularly via calipers, and the tumor volume was calculated using the following formula: (L × w^2)/2, where L is the length and w is the width of the tumor. When a cutoff point (tumor volume > 1,000 mm^3) was reached or after 42 days after the beginning of treatments, mice were euthanized by cervical dislocation following inhalation of 5% CO[2]. Four mice previously treated with clofarabine were kept alive for additional days post-treatment to monitor tumor regrowth. The tumors were subsequently collected, measured, photographed, and weighed. Two mice did not complete the experiment: one from the “DMSO” group, which reached the cutoff point on day 30, and one from the “cladribine” group, which died on day 16 from an unknown cause. Proteomic experiment Sample preparation and analysis Total protein extraction was performed using radioimmunoprecipitation assay buffer (RIPA) buffer as previously described.[126]^38 The buffer consisted of 50 mM Tris-HCl, 1% IGEPAL, 150 mM NaCl, 1 mM EGTA, 1 mM NaF, and 0.25% sodium deoxycholate, and was supplemented with protease inhibitors to prevent protein degradation during the extraction process. The samples were transferred to Proteogen platform (UNICAEN, Caen) for proteomic analysis. Five micrograms of each protein extract were prepared via a modified gel-aided sample preparation protocol.[127]^39 The samples were digested with trypsin/Lys-C overnight at 37°C. For nano-LC fragmentation, peptide samples were first desalted and concentrated onto a μC18 Omix (Agilent) before analysis. The chromatography step was performed on a NanoElute (Bruker Daltonics) ultrahigh-pressure nanoflow chromatography system. Approximately 100 ng of each peptide sample were concentrated onto a C18 pepmap 100 (5 mm × 300 μm i.d.) precolumn (Thermo Fisher Scientific) and separated at 50°C onto a reversed phase Reprosil column (25 cm × 75 μm i.d.) packed with 1.6 μm C18 coated porous silica beads (IonOpticks). The mobile phases consisted of 0.1% formic acid, 99.9% water (v/v) (A) and 0.1% formic acid in 99.9% acetonitrile (ACN;v/v) (B). The nanoflow rate was set at 250 nL/min, and the gradient profile was as follows: from 2 to 30% B within 70 min, followed by an increase to 37% B within 5 min and further to 85% B within 5 min and reequilibration. Mass spectrometry (MS) experiments were carried out on a trapped ion mobility spectrometry (TIMS) with a time-of-flight (TOF) Pro mass spectrometer (Bruker Daltonics) with a modified nanoelectrospray ion source (CaptiveSpray, Bruker Daltonics). A 1,400 spray voltage with a capillary temperature of 180°C was typically employed for ionization. MS spectra were acquired in positive mode in the mass range from 100 to 1700 m/z with a 0.60–1.60 1/k0 window. For the experiments described here, the mass spectrometer was operated in parallel accumulation serial fragmentation (PASEF) data-independent acquisition (DIA) mode with the exclusion of single-charged peptides. The DIA acquisition scheme consisted of 16 variable windows ranging from 300 to 1,300 m/z. Protein identification Database searching and LFQ quantification (using XIC) were performed using DIA-NN (version 1.8.2).[128]^40 An updated UniProt Homo sapiens database was used for library-free search/library generation. For RT prediction and extraction mass accuracy, the default parameter 0.0 was used, which means DIA-NN performed automatic mass and retention time (RT) correction. The top six fragments (ranked by their library intensities) were used for peptide identification and quantification. The variable modifications allowed were as follows: Nterm-acetylation and oxidation (M). In addition, C-propionoamide was used as a fixed modification. “Trypsin/P” was selected. Data were filtered according to an FDR of 1%. Cross-run normalization was performed via RT-dependent methods. Identification of differentially expressed proteins To quantify the relative levels of protein abundance between different groups, data from DIA-NN were analyzed via the DEP package from R. Briefly, proteins that are identified in 2 out of 3 replicates of at least one condition were filtered; missing data were imputed using a normal distribution from PERSEUS and differential enrichment analysis was based on linear models and the empirical Bayes statistic. A 1.2-fold increase in relative abundance and a 0.05 FDR were used to determine enriched proteins. ANOVA was performed with PERSEUS to determine the enriched proteins. Functional annotation and pathway enrichment analysis Pathway enrichment analysis and functional annotation were conducted using ShinyGO (v.0.8), an online bioinformatics tool available at [129]http://bioinformatics.sdstate.edu/go/.[130]^41 The analysis included GO functional annotation, including BP and Molecular Functions. Additionally, pathways were analyzed using Reactome, WikiPathways, and KEGG databases. Enrichment p values were calculated via the hypergeometric test, and to account for multiple testing, the FDR was computed using the Benjamini-Hochberg method. The data were analyzed by via the “sort by average ranks (FDR and fold enrichment)” option, which permits the initial filtering of pathways and GO terms by an FDR threshold below 0.05, and only those that were deemed significant were subsequently ranked. The top 15 pathways were selected for further analysis and discussion. Cell cycle analysis The cell cycle distribution was assessed via the NucleoCounter NC-3000 (ChemoMetec) in accordance with the manufacturer’s “Fixed Cell Cycle-DAPI Assay” protocol. A total of 18,000 cells per cm^2 were seeded and incubated for 24 h with the respective treatments. Following this incubation, both floating and adherent cells (which were harvested via trypsinization) were combined and washed with phosphate-buffered saline (PBS). Next, approximately 2 × 10^6 cells were fixed in ice-cold 70% ethanol for a minimum of 24 h at 4°C. After fixation, the cells were washed again with cold PBS. The cell pellet was then resuspended in 0.5 mL of DAPI solution (comprising 1 μg/mL DAPI and 0.1% Triton X-100 in PBS) and incubated for 5 min at 37°C. The percentages of cells arrested in the sub-G1, G0/G1, S, and G2/M phases of the cell cycle were quantified via the manufacturer’s software. Statistical analyses All the statistical analyses were conducted via Prism version 10.1.2. Data from both in vitro and in vivo experiments are presented as the means ± SEM. Statistical comparisons were made via two-way ANOVA, with Dunnett’s or Sidak’s multiple comparison tests as appropriate. For tumor weight analysis, a t test was employed. Statistical significance was defined as a p value of less than 0.05. Data availability The datasets supporting the conclusions of this article are included within the article and its additional files. Acknowledgments