Osteosarcoma is a type of highly aggressive bone tumor arising from primitive cells of mesenchymal origin in adults and is associated with a high rate of tumor relapse (1, 2). The main treatments for this malignancy are surgical resection combined with multiagent chemotherapy. Unfortunately, the incidence of osteosarcoma is around 5/1,000,000, and patients with osteosarcoma had a 5-year overall survival rate of ~15% (3–5). To our best knowledge, the pathophysiology of osteosarcoma remains unclear. The pathogenesis of osteosarcoma is complicated by the tumor microenvironment (TME) including immune cells, malignant mesenchymal tumor cells, vascular networks, and fibroblasts (6, 7). Therefore, it is of great importance to examine the molecular mechanisms underlying the pathophysiology of osteosarcoma and develop novel therapies for the treatment of osteosarcoma.

The conventional transcriptomic profiling is performed on mixed cell populations, which has an insufficient resolution for detecting specific cellular types and fails to assess the complexity of intratumoral heterogeneity (8, 9). Up to date, single-cell RNA sequencing (scRNA-seq) plays an important role in identifying the intratumor heterogeneity of different types of cancers and the cellular interaction within the TME. Davidson et al. performed scRNA-seq to examine the stromal compartment in murine melanoma and draining lymph nodes at points across tumor development and revealed a dynamic stromal niche that promoted tumor growth (10). Maynard et al. demonstrated that biological features revealed by scRNA-seq were biomarkers of clinical outcomes in human lung cancer, which highlighted how therapy-induced adaptation of the multicellular ecosystem of metastatic cancer shaped clinical outcomes (11). Chung et al. performed scRNA-seq and demonstrated that breast cancer transcriptome exhibited a wide range of intratumoral heterogeneity, which was shaped by the tumor cells and immune cells in the surrounding microenvironment (12). To our best knowledge, various studies have performed scRNA-seq in osteosarcoma under different experimental settings. For example, Liu et al. showed that single-cell transcriptomics elucidated the complexity of the tumor microenvironment of treatment-naive osteosarcoma (13). Zhou et al. identified 11 major cell clusters based on unbiased clustering of gene expression profiles and canonical markers via RN sequencing of 100,987 individual cells from 7 primary, 2 recurrent, and 2 lung metastatic osteosarcoma lesions (14). These studies highlighted the important applications of scRNA-seq in deciphering the molecular mechanisms underlying osteosarcoma pathophysiology.

Recently, reanalysis of high-throughput datasets including the scRNA-seq data in cancer studies has revealed many important findings. In the present study, we further explored the scRNA-seq data from the GEO database (GSE152048) and investigated the novel signaling pathways that may contribute to osteosarcoma metastasis.

Accumulating evidence has implied that the biological behaviors of tumor cells are heavily affected by the tumor microenvironment. Network interactions among various types of tumor cells and the TME have been shown to promote tumor progression at multiple levels (1, 7, 17). Therefore, further understanding the underlying interactions should give rise to a novel therapeutic approach. Our results showed that 11 clustered cell types were identified across 11 osteosarcoma tissues, with cell types including “osteoblastic”, “myeloid”, “osteoblastic_proli”, “osteoclast”, and “tumor-infiltrating lymphocytes (TILs)” being the main types. The DEGs between different cell types from primary, metastatic, and recurrent osteosarcomas were mainly enriched in the GO terms including “negative regulation of hydrolase activity”, “regulation of peptidase activity”, “regulation of binding”, “negative regulation of proteolysis”, and “negative regulation of peptidase activity” and in the KEGG pathways including “transcriptional misregulation in cancer”, “cellular senescence”, “apoptosis”, “FoxO signaling pathway”, “cell cycle”, “NF-kappa B signaling pathway”, “p53 signaling pathway”, “pentose phosphate pathway”, and “protein export”. For the cell–cell communication network analysis, the different interaction profiles between cell types were detected among primary, metastatic, and recurrent osteosarcomas. Further exploration of the KEGG pathway revealed that these ligand/receptor interactions may be associated with the NF-κB signaling pathway and its interacted mediators. In conclusion, the present study for the first time explored the scRNA-seq dataset in osteosarcoma, and our results revealed the 11 clustered cell types and demonstrated the novel cell–cell interactions among different cell types in primary, metastatic, and recurrent osteosarcomas. The present study may provide new insights into understanding the molecular mechanisms of osteosarcoma pathophysiology.

Based on the study of Zhou et al., the researchers characterized the transcriptomic properties, regulators, and dynamics of osteosarcoma malignant cells together with their TME, particularly stromal and immune cells (14). In addition, the study revealed that the proinflammatory fatty acid-binding protein 4+ macrophage infiltration was identified in lung metastatic osteosarcoma lesions. Lower osteoclast infiltration was detected in chondroblastic, recurrent, and lung metastatic osteosarcoma lesions compared with primary osteoblastic osteosarcoma lesions. They found that TIGIT blockade enhanced the cytotoxicity effects of primary CD3+ T cells with a high proportion of TIGIT+ cells against osteosarcoma (14). In our analysis, the main cell types include “osteoblastic”, “myeloid”, “osteoblastic_proli”, and “osteoclast”. In metastatic osteosarcoma, there is a large proportion of TIL, MSC, and osteoblastic and mesenchymal stem cells (MSCs); in recurrent osteosarcoma, there is a large proportion of chondroblastic cells.

Several studies have proposed the importance of TIL in osteosarcoma. TILs were recruited to the tumor site; however, tumor cells had the capability of escaping the immune response (18). Studies found that the neutrophil–lymphocyte ratio was strongly correlated with the overall survival and progression-free survival of patients with osteosarcoma (18). Recent studies showed that higher infiltration of immune cells was correlated with better clinical outcomes in osteosarcoma (18). Sundara et al. showed that PD-L1 and T-cell infiltration were increased in the presence of HLA class I expression in metastatic high-grade osteosarcoma (19). The role of MSCs has also been reported in metastatic osteosarcoma. Tsukamoto et al. showed that mesenchymal stem cells promoted tumor engraftment and metastatic colonization in a rat osteosarcoma model (20); mechanistic studies showed that MSCs under stress increased osteosarcoma migration and apoptosis resistance via extracellular vesicle-mediated communication (21). Cortini et al. found that tumor-activated MSCs promoted osteosarcoma stemness and migratory potential via interleukin-6 secretion (22). Chondroblastic osteosarcoma ranks as the second commonly diagnosed osteosarcoma in young adults. The cellular origin of chondroblastic osteosarcoma is still unclear. In recurrent and metastatic osteosarcomas, the strong interaction between chondroblastic and osteoblastic cells was detected in our study, suggesting the transdifferentiation of malignant osteoblastic cells from malignant chondroblastic cells.

Based on the analysis of signaling pathways, we noticed that the NF-κB signaling pathway is key in the pathophysiology of osteosarcoma. NF-κB is involved in the early stress response to cellular DNA damage, and tumor necrosis factor α (TNFα) is an activator of the classical NF-κB pathway (23). After TNFα stimulation, IκB kinases (IKKs), such as IKKα and IKKβ, are first activated in cells. IKKs degrade intracellular IκB, thus releasing NF-κB molecules such as RelA, which are phosphorylated at S536, and this allows RelA to enter the nucleus (23), bind to genes at corresponding binding sites, and initiate the transcription of downstream genes, such as CCND1 and Bcl-2, to regulate cell cycle progression, proliferation, and survival (23). Nishimura et al. showed that transfection of NF-κB decoy oligodeoxynucleotide suppressed pulmonary metastasis by murine osteosarcoma (24). Londhe et al. showed that classical NF-κB metabolically reprogramed sarcoma cells through regulation of hexokinase 2 (25). Activation of the NF-κB axis could enhance CRL4B DCAF11 E3 ligase activity and regulate cell cycle progression in human osteosarcoma cells (26). Inhibition of NF-κB signaling could attenuate osteosarcoma progression (26, 27). In the present study, the costimulatory analysis revealed that ligands and receptors were correlated with NF-κB signaling. In future studies, we may perform functional studies to examine if disturbance of the ligands/receptors could lead to modulation of the NF-κB signaling pathway, which may eventually regulate the osteosarcoma progression.

The present was only focused on the bioinformatics analysis, which may limit the significance of the current findings. Thus, further studies especially validating the potential mediators detected in this study by using experimental assays should be considered. Due to the limited source of the scRNA-seq dataset for osteosarcoma, the present study only analyzed one dataset, and future studies should consider exploring other available scRNA-seq datasets to confirm the present findings. Moreover, our analysis only focused on the NF-κB signaling pathway and its interacted mediators in osteosarcoma, and other signaling pathways may be considered for exploration in our future studies.

In conclusion, the present study for the first time explored the scRNA-seq dataset in osteosarcoma, and our results revealed the 11 clustered cell types and demonstrated the novel cell–cell interactions among different cell types in primary, metastatic, and recurrent osteosarcomas. The NF-κB signaling pathway may play a key role in regulating the TME of osteosarcoma. The present study may provide new insights into understanding the molecular mechanisms of osteosarcoma pathophysiology.

The original contributions presented in the study are included in the article/ Supplementary Material . Further inquiries can be directed to the corresponding author.

JM and RW designed and supervised the whole project. RW, XD, HaL, ZS, and HeL processed the datasets and analyzed the data. JM wrote the manuscript. YS and WW revised the drafted manuscript. All authors contributed to the article and approved the submitted version.

This study was supported by the Basic Public Welfare Research Program of Zhejiang (LGF19H060002).

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.