Background:​

G protein-coupled receptor (GPCR) signaling mediates a wide spectrum of physiological functions, including bone development and remodeling. Fibrous dysplasia (FD) is a common skeletal dysplasia where normal bone and bone marrow are replaced by fibrous tissue and expansile trabecular bone lesions. The craniofacial bones are often involved, leading to pain and facial deformities. FD is a mosaic disease caused by a somatic mutation in the GNAS gene encoding the G-protein alpha subunit (G_sα) that leads to constitutive activation of the G_s signaling pathway via increased cyclic AMP (cAMP) levels. Unfortunately, FD has no effective medical treatments.

Major challenges have hampered the development of pharmacologic strategies that specifically target GNAS or the G_sα protein. We previously developed the ColI(2.3)⁺/Rs1⁺ mouse model (Rs1) in which the G_s signaling pathway is activated specifically in bone by an engineered GPCR protein. These mice showed increased trabecular bone formation with loss of marrow space and cortical bone, which strongly resembles human FD. There was also a dramatic increase in the number of immature osteoblasts present in the FD lesions, suggesting that activation of G_s signaling caused an accumulation of these cells. Our prior studies showed that the FD-like lesions had increased Wnt signaling, which may be a major driver of the phenotype. Furthermore, blocking the G_s signaling could reverse the abnormal bone phenotype, providing proof-of-concept for finding drugs that could reverse the phenotype.

Methods:​

Long bones from 2 wildtype and 2 ColI(2.3)⁺/Rs1⁺ 9-week-old male mice were used to isolate stromal cells for scRNAseq analyses. Differentially expressed genes between clusters and between experimental groups within clusters were ascertained using Seurat. The cell types in each cluster were identified using SingleR to match gene expression in each cluster to existing databases. To test the role of Wnt signaling, we performed global Wnt inhibition on the ColI(2.3)⁺/Rs1⁺mouse model using the porcupine inhibitor LGK974. The mice were evaluated by histology and micro computed tomography (micro-CT). Finally, drug repositioning analysis was performed by comparing the molecular disease signature of FD we created from our single cell gene expression data. Our disease signature was queried against pharmacological agents in Broad Institute’s Connectivity Map database (CMap).

Results:​

The cellular compositions between WT control and ColI(2.3)⁺/Rs1⁺ bones as identified from our scRNAseq data were similar, except that there was a significant increase in the total number of cells with an expansion of osteoblastic lineage cells. The osteoblast cluster also had the highest number of differential expression (DE) genes. Expression of Gi coupled receptors were increased, potentially as a compensatory mechanism for the strong activation of G_s-GPCR pathway induced by Rs1 expression. In addition, the scRNAseq data revealed activation of the GH/IGF1 signaling pathway in osteoblastic cells in the FD-like bone lesions. The scRNAseq data also identified broad expression of many Wnt ligands within the bone cells, including within the osteoblast cluster. Treatment with the broad Wnt production inhibitor of porcupine, LGK974, showed resorption of the abnormal FD bone;​ however, the fibrocellular infiltrate in the ColI(2.3)⁺/Rs1⁺ mice was still present. Drugs from the Connectivity Map database that have opposite gene expression patterns to FD were identified through our computational analysis and are candidates for testing as potential treatment of FD.

Conclusions:​

This study uses a translational approach combining a unique mouse model of FD to elucidate the cellular interactions in FD-like lesions and create a novel gene expression disease signature of FD. We found that broad Wnt inhibition can lead to decreased fibrous dysplastic bone, but the fibrocellular infiltrate does not appear to fully reverse. These results provide new insight into understanding interactions between the Wnt and G_s signaling pathways in FD pathogenesis and bone formation. Together, this work expands our understanding of FD pathogenesis and helps identify potential strategies for treatment, laying a strong foundation for future research on GPCR signaling and bone development.