背景

骨肉瘤（Osteosarcoma, OS）是一种最常见的骨恶性肿瘤，起源于能够产生骨基质的原始细胞，以形成不成熟的骨样组织为特征。骨肉瘤作为一种罕见肿瘤，发病率大约为3-4/100万人，主要影响青少年和60岁以上的老年人，男性略多于女性，发病部位常见于四肢长骨干骺端。目前常用的标准治疗结合了所有可见肿瘤的完整切除和多药物联合化疗方案，使用如多柔比星、大剂量甲氨蝶呤、顺铂和异环磷酰胺等强效抗骨肉瘤药物。这种新辅助化疗-手术-辅助化疗综合治疗模式显著提高了骨肉瘤病人的5年生存率。尽管如此，近40年来骨肉瘤治疗的进展仍显缓慢，未再获得进一步的突破，长期生存率停滞不前。主要原因是骨肉瘤的综合治疗过程中产生的耐药性及高转移性。骨肉瘤耐药的复杂性在于其多样且难以捉摸的分子机制。传统的群体细胞分析掩盖了个体细胞间的异质性，限制了对耐药机制的深入理解。单细胞转录组测序技术通过分析单个细胞的基因表达谱，揭示了肿瘤内细胞的异质性和特定亚群的转录特征，识别出驱动耐药性的关键基因和信号通路，并追踪细胞在化疗压力下的动态变化，从而更精准地解释骨肉瘤的耐药机制。

目的

本研究旨在通过单细胞及空间转录组测序技术，对人骨肉瘤的肿瘤微环境进行全面分析，刻画骨肉瘤的单细胞分子图谱，以此建立骨肉瘤单细胞注释模型，并为反卷积骨肉瘤空间转录组学及Bulk转录组提供参考在分子水平上揭示骨肉瘤差异的生物行为的潜在机制，包括复发、转移、耐药；探究骨肉瘤肿瘤微环境中不同细胞群体相互作用的具体机制，希望能够为未来骨肉瘤的治疗提供一定的科学依据和思路。

方法

共收集了10例骨肉瘤病人的组织样本进行了单细胞转录组测序，其中包括6例未经治疗的骨肉瘤穿刺样本和4例接受新辅助治疗后的手术切除样本。将这些数据与已公开发表的骨肉瘤单细胞转录组数据集GSE162454和GSE152048进行了整合，构建了迄今为止规模最大的骨肉瘤单细胞转录组样本分析队列，并对骨肉瘤肿瘤微环境的细胞亚群组成进行了详尽分析。还收集了1例新辅助化疗后骨肉瘤样本进行了空间转录组测序。通过对骨肉瘤单细胞转录组数据的深入分析及细胞图谱的建立，本研究为空间转录组以及Bulk转录组数据建立了分子图谱进行去卷积分析以解析骨肉瘤肿瘤微环境的空间生态位。本研究进一步探讨了高表达SPP1的骨肉瘤细胞亚群在将正常成纤维细胞激活为肿瘤相关成纤维细胞过程中的作用。

结果

在骨肉瘤单细胞转录组分析队列中骨肉瘤细胞和髓系细胞构成了肿瘤微环境的主要部分。新辅助治疗后，肿瘤微环境中髓系免疫细胞及淋巴系免疫细胞的比例减少，且其功能状态从抗肿瘤向促肿瘤转变，导致免疫抑制状态的进一步加剧。同时，肿瘤细胞、间充质细胞、内皮细胞的比例相对增加，这反映了它们在促进化疗耐药中的潜在作用。通过单细胞转录组分析建立的注释模型，反卷积骨肉瘤空间转录组数据成功发现了cluster 6空间生态位位于肿瘤坏死区域前沿，高表达 COL4A1、COL4A2和COL8A1等细胞外基质基因，提示这些基因可能与化疗耐药相关。通过进一步深入分析骨肉瘤肿瘤微环境中不同细胞群体的相互作用，本研究发现了骨肉瘤细胞能够通过分泌SPP1与正常成纤维细胞的整合素受体αVβ3结合，进而激活其为促进肿瘤生长的肿瘤相关成纤维细胞，表明SPP1有可能成为骨肉瘤治疗的潜在靶点。

结论

这些研究结果不仅揭示了单细胞转录组在揭示骨肉瘤的肿瘤异质性和新辅助化疗对肿瘤微环境的复杂影响等方面的重要作用，还部分阐释了骨肉瘤肿瘤微环境中肿瘤细胞与肿瘤相关成纤维细胞之间的相互作用，为深入理解骨肉瘤的发展、耐药机制及其转移等机制提供了重要依据。

Background

Osteosarcoma (OS) is the most common malignant bone tumor, originating from primitive cells capable of producing bone matrix, characterized by the formation of immature bone-like tissue. As a rare tumor, the incidence of osteosarcoma is approximately 3-4 per million people, primarily affecting adolescents and individuals over 60 years old, with a slightly higher prevalence in males than females. The disease commonly occurs at the metaphysis of long bones in the extremities. Current standard treatments combine complete tumor resection with multidrug chemotherapy regimens, including potent anti-osteosarcoma drugs such as doxorubicin, high-dose methotrexate, cisplatin, and ifosfamide. This treatment model of "neoadjuvant chemotherapy + surgery + adjuvant chemotherapy" has significantly improved the five-year survival rate of osteosarcoma patients. Nevertheless, the progress in osteosarcoma treatment over the past 40 years has not been improved, with long-term survival rates remaining stagnant. The primary reasons are the development of drug resistance and the high metastatic potential observed during the comprehensive treatment of osteosarcoma. The complexity of osteosarcoma drug resistance lies in its diverse and elusive molecular mechanisms. Traditional bulk cell analyses obscure the heterogeneity between individual cells, limiting our understanding of resistance mechanisms. Single-cell RNA sequencing (scRNA-seq) technology, by analyzing the gene expression profiles of individual cells, unveils the cellular heterogeneity within tumors and the transcriptional characteristics of specific subpopulations. This approach identifies key genes and signaling pathways driving drug resistance and tracks the dynamic changes of cells under chemotherapy pressure, thereby providing a more precise explanation of the resistance mechanisms in osteosarcoma.

Purpose

This study aims to conduct a comprehensive analysis of the tumor microenvironment of human osteosarcoma using single-cell and spatial transcriptomics technologies to reveal the biological behavior of osteosarcoma at the molecular level and explore potential mechanisms of tumor progression and metastasis, hoping to provide some scientific basis and insights for the future treatment of osteosarcoma.

Methods

In this study, 10 samples from osteosarcoma patients were collected for single-cell transcriptomic sequencing, including 6 pre-chemotherapy samples and 4 post-chemotherapy samples, which were integrated with publicly available single-cell datasets GSE162454 and GSE152048 to construct the largest-scale analysis cohort of osteosarcoma single-cell transcriptomic samples to date, detailing the composition of the osteosarcoma tumor microenvironment. Additionally, one sample from post-chemotherapy osteosarcoma was collected for spatial transcriptomic sequencing. Analysis of osteosarcoma single-cell transcriptomic data was used to deconvolute spatial transcriptomic data and establish molecular maps for bulk transcriptomic data from the TARGET database. Furthermore, we explored the role of osteosarcoma cells highly expressing SPP1 in activating normal fibroblasts into cancer associated fibroblasts (CAFs).

Results

We found that osteosarcoma cells and myeloid cells are major components of the osteosarcoma tumor microenvironment. Post-neoadjuvant treatment led to a reduction in the proportion of myeloid and lymphoid immune cells in the tumor microenvironment and a shift from anti-tumor to pro-tumor functional states, further exacerbating immune suppression. The relative proportions of tumor cells, stromal cells, and endothelial cells increased, reflecting their potential role in promoting chemotherapy resistance. Further analysis revealed that osteosarcoma cells could activate normal fibroblasts into pro-tumorigenic CAFs by binding SPP1 to the integrin receptors αVβ3 of normal fibroblasts.

Conclusions

These results reveal the heterogeneity of osteosarcoma, the complex effects of neoadjuvant chemotherapy on osteosarcoma treatment, and the interactions between tumor cells and fibroblasts, providing a basis for further elucidating the development and drug resistance mechanisms of osteosarcoma.