Contribution of the Microenvironment in Bone Resident Cancer: the Role of Thrombospondin-1 in Bone and Risk Loci Contributing to Multiple Myeloma

Cancer growth in bone, a defining characteristic of multiple myeloma and common in many solid tumors including breast and prostate cancer, is characterized by the active participation of host microenvironment cells. Tumor cells stimulate the vicious cycle through stimulation of bone remodeling by osteoclasts and osteoblasts, resulting in increased release of growth factors and cytokines to further promote local tumor growth. Notably, cancer cells are unable to remodel bone independently, and require the metabolically active bone cells for this support. A specific role for the bone microenvironment in mediating tumor growth was first discussed in 1889 with the introduction of Paget\u27s seed-and-soil hypothesis that tumor cells will only grow in a favorable microenvironment. This dissertation is focused on investigating the properties of a congenial soil. In Chapter 2 we discuss the role of bone-resident Thrombospondin-1 (TSP1), a glycoprotein with well-documented roles in promoting cancer growth. Chapter 3 is focused on the identification of genes conferring inherited susceptibility to multiple myeloma that participate both in malignant cells and in host supportive cells. TSP1 is a large matricellular protein that interacts β3 integrin, CD47, and other receptors to regulate cell migration, adhesion, and proliferation with effects on tumor angiogenesis, inflammation, and wound healing. We examined the role of TSP1 in the bone microenvironment. TSP1-/- mice had increased trabecular bone volume and increased cortical bone area and thickness compared to wild type controls. Surprisingly, TSP1-/- mouse bones did not resist bending as much as anticipated, and we confirmed a bone materials defect, indicating that TSP1 was important for maintaining bone quality. In addition to the mechanical defects, TSP1-/- mice had decreased serum CTX compared to controls, indicating an osteoclast defect. Primary osteoclast cultures require TSP1 at early stages of osteoclastogenesis, though exogenous TSP1 is dispensable at later stages of maturation. Interestingly, we found that TSP1-deplete osteoclast cultures had a significant and dramatic increase in inducible nitric oxide synthase (iNOS) expression. Moreover, upon administration of a NOS inhibitor to mice, the TSP1-/- bone resorption defect was rescued to wild type levels. To test whether osteoblast-deposited bone-resident TSP1 inhibits early osteoclast NO signaling, we plated osteoclasts on either wild type or TSP1-/- bone and found that only wild type bone was sufficient to block iNOS expression. Thus, we conclude that TSP1 is a paracrine signaling molecule that couples OB activity to OC formation and targeting of the TSP1 signaling pathway, currently under investigation for treating cancer and modulating blood pressure, may have additional effects on bone. The second part of this dissertation focuses on host susceptibility to multiple myeloma (MM) and its requisite precursor, monoclonal gammopathy of undetermined significance (MGUS). The C57Bl/KaLwRij (KaLwRij) mouse strain has increased susceptibility to benign idiopathic paraproteinemia (BIP), analogous to human MGUS, and is a common mouse model for the disease. Following immunization, KaLwRij mice developed a sustained M-spike at a higher frequency than 12 other genetically diverse mouse strains, most notably, the highly related C57Bl/6J (B6) strain. To query the genetic basis for BIP susceptibility in KaLwRij mice, we completed SNP analysis to identify variation between KaLwRij and B6. We found that KaLwRij is a unique inbred strain, completely distinct from the closely related B6. We identified several thousand variants between KaLwRij and B6 and compiled a candidate gene list of 419 genes. To ensure that we would pursue genes applicable to human disease, we completed whole genome association analysis between MM patients and healthy controls and identified SNPs located in 180 gene loci. Combining these two gene sets resulted in a candidate gene list of 5 genes, one of which, Samsn1, is annotated as a negative regulator of B-cell activation. Upon closer examination of the Samsn1 locus, we found that KaLwRij harbored a germline deletion of all coding exons of the gene, resulting in absent RNA and protein expression. Consistent with the published role of SAMSN1 in genetically deficient mice, KaLwRij mice had enhanced B-cell function as evaluated by B-cell proliferation in primary culture and in vivo IgG2b response. To investigate the role of Samsn1 in host cells, we profiled Samsn1 expression in bone microenvironment cells and found that macrophages were high expressers. KaLwRij macrophages had enhanced expression of polarization markers, suggesting that they are pre-activated under homeostatic conditions. Further, we found that KaLwRij macrophages stimulated to a pro-tumorigenic M2 polarization promoted 5TG myeloma tumor growth more than B6 control macrophages in mice. In conclusion, we identified candidate gene lists underlying murine and human susceptibility to MGUS and myeloma. Specifically, we found that Samsn1 contributes to KaLwRij susceptibility to BIP, and likely to human MGUS, through effects in pre-malignant B-cells and in macrophages