Abstract
In phases of myeloma dormancy such as MGUS (monoclonal gammopathy of uncertain significance) or post chemotherapy remission, certain characteristics of the bone marrow niche promote quiescence of the tumor. Myeloma cell dormancy has been proposed to be induced by contact with 'bone-lining' endosteal niche cells, which are destroyed by active disease. Preservation of dormancy could prevent disease relapse or MGUS progression to myeloma, however the molecular mechanisms and signaling pathways that maintain this bone-lining niche are unknown.
We addressed this in vivo by sorting endosteal niche components (osteoblasts, mesenchymal stem cells (MSCs), endothelial cells and tumor) from myeloma-bearing and control mice, followed by RNA-Seq transcriptome profiling and gene set enrichment analysis (GSEA). Endosteal MSCs showed greatest transcriptome differences between myeloma-bearing and control groups. MSCs from myeloma-bearing mice showed positive enrichment (p = 0.004) for a 200-gene bone remodeling gene set. The leading edge (highest contributors to enrichment) of this gene set contained 24% BMP pathway genes, far higher than the next nearest pathway represented (TGFb signaling and AP-1 complex, both 6.5%). Whereas various other signaling pathways identified in the leading edge are known to be of high importance in myeloma bone disease (e.g. RANKL, TGFb, Wnt, hedgehog), BMP signaling has not hitherto been reported as deregulated.
To assess the role of BMP signaling in myeloma in vitro and in vivo, we used the BMP pathway inhibitor LDN-193189 (LDN), which has high affinity for type I BMP receptors Alk2, Alk3 and Alk6. In the myeloma-bearing KaLwRij/5TGM1 mouse model, LDN significantly improved trabecular (p=0.02) and cortical bone volume (p=0.004) and reduced serum TRAP levels (p=0.003). Histomorphometric analysis demonstrated that LDN reduced osteoclast numbers (p<0.0001) and increased osteoblast numbers (p=0.018) in myeloma-bearing mice. As BMPs are well known to be osteoinductive, the mechanism by which BMP inhibition might cause increased bone mass in this model was investigated. In vitro, expression of rankl was significantly decreased by LDN in osteoblasts cultured from myeloma-bearing mice (p=0.03). Transcriptome profiling of endosteal MSCs sorted from myeloma mice treated with LDN or vehicle, subjected to GSEA (MSigDB Hallmark gene set database), showed significant enrichment in gene sets representing epithelial-mesenchymal-transition (EMT) and apical junction formation with LDN use. The leading edge genes included those involved in matrix deposition and osteoblast differentiation, indicating LDN may lead to reversal of certain aspects of the osteoblast differentiation block seen in MSCs in myeloma.
LDN had no effect on overall tumour burden in vivo, however altered the niche-preference (endosteal vs central marrow) of myeloma cells to favor the endosteal niche (ratio endosteal: central marrow myeloma distribution 0.23 vehicle group, 0.37 LDN group, p=0.034). LDN also significantly reduced expression in liver of the iron regulatory hormone hepcidin (p=0.0003). Increased hepcidin has been described in multiple myeloma and likely contributes to inflammatory anemia, of clinical relevance in many myeloma patients.
The analysis of cell-type specific in vivo gene expression changes in the myeloma endosteal niche led to investigation of a pathway not previously recognized as deregulated in myeloma bone disease. This technique has not previously been applied to the myeloma niche or, to our knowledge, any non-myeloid tumor invading bone marrow. We demonstrate its use in highlighting type 1 BMP receptor signaling as a novel therapeutic target in myeloma bone disease. BMP inhibition enhances osteoblast differentiation and reduces osteoclast activity in this model, suggesting anabolic and anti-resorptive benefits. The resulting preservation of the endosteal niche may have potential to increase the dormant tumor fraction. In addition to amelioration of bone disease, we hypothesize other clinical benefits, including improvement of inflammatory anemia and prolongation of quiescent phases, e.g. MGUS and post-treatment remission.
Disclosures
Ramasamy: Celgene: Honoraria, Research Funding.
Somatic SMAD3-activating mutations cause melorheostosis by up-regulating the TGF-β/SMAD pathway
