Understanding the cell heterogeneity and tumor progression of giant cell tumor of bone at single-cell resolution

Abstract

Giant cell tumor of bone (GCTB) is a rare primary bone tumor. Although mainly classified as a benign tumor, GCTB exhibits a local recurrence rate as high as 50% following surgery and can induce bone and soft tissue destruction resulting in severe morbidity for the patient. Histologically, GCTB consists of multinucleated osteoclast-like giant cells, mononuclear monocytes, and spindle-shaped mesenchymal-like stromal cells. Although bone resorption is contributed by the osteoclast-like giant cells, the actual neoplastic component of GCTB is the mesenchymal-like stromal cells. Intriguingly, molecular genetic analysis has shown that more than 90% of GCTB harbors the H3.3G34W mutation and is restricted to mesenchymal-like cell component, suggesting that the H3.3G34W mutation plays a pivotal role in the tumorigenesis of GCTB. However, the molecular relationship between H3.3G34W mutation and tumor progression in GCTB has not been well investigated. To better understand the molecular processes of GCTB, this study sequenced 45,695 single cells from six human GCTB samples. It demonstrated that the cellular composition of GCTB is more complex than once thought, involving interactions between various myeloid cell types (giant cells, monocyte, macrophage), several subtypes of mesenchymal-like stroma cells, endothelial cells, and T cells. Among the mesenchymal-like stromal cells, five distinct cellular subpopulations were identified. Trajectory analysis revealed that most neoplastic mesenchymal-stromal cells from a collagen-rich subpopulation differentiated toward a subpopulation of osteoblastic progenitors where matrix degradation, osteoclast differentiation, and Hippo pathway as molecular features associated with stromal cell trajectory progression. Moreover, cells with or without the H3.3 G34W mutation were also identified at single-cell resolution. Genes involved in the Hippo pathway were found to be highly enriched in the stromal cell subsets with a higher proportion of H3.3 G34W mutation-harboring cells. Immunostaining results further confirmed that H3.3G34W mutation was strongly correlated with YAP (the downstream effector of the Hippo pathway) nuclear translocation in GCTB patient samples. Functional assays indicated that YAP inhibition reduced tumor cell proliferation. Overall, these findings illustrate the intratumor heterogeneity of human GCTB, underlying the neoplastic cellular progression, and reveal that YAP inhibitors could potentially function as new treatment strategies for patients suffering from GCTB.