Mature Osteoblasts Provide a Protective Niche Against Multiple Myeloma Growth and Survival within the Tumor Microenvironment

Multiple myeloma (MM), a tumor of B-lymphocyte lineage cells, originates in the bone marrow (BM) and is highly influenced by the BM tumor microenvironment (TME). Among cells of the TME, osteoblasts are the most versatile regulators of many hematopoietic lineage cells through either direct cell-cell communication or secreted factors. Specifically relevant to MM, G-protein coupled receptor signaling in pre-osteoblasts is essential for the differentiation, maturation, and egress of B-cells (Panaroni et al., 2015). Despite these key roles, the contribution of osteoblasts to the initiation and progression of MM is not well understood. MM is characterized by osteolytic bone lesions partly due to decreased numbers of osteoblasts. Here, we hypothesize that osteoblasts provide niche support to maintain myeloma cells in a quiescent stage and that the loss of the osteoblastic niche leads to the progression of MM. As a proof of concept, we previously showed that increasing osteoblastogenesis by inhibiting Activin A led to inhibition of MM growth in an in vivo humanized myeloma model (Vallet et al., 2010).

We generated mice in which mature osteoblasts could be postnatally deleted in an inducible and reversible manner. Diphtheria toxin receptor floxed mice were mated with mice expressing Cre-recombinase driven by the osteocalcin promoter to generate OC-Cre/iDTR mice. Littermates heterozygous for DTR but lacking the OC-Cre expression were used as controls. The OC-Cre/iDTR mice were indistinguishable from the controls until treated with diphtheria toxin (DT). To induce postnatal deletion of mature osteoblasts, the OC-Cre/iDTR and control mice both were treated with 50 µg/Kg DT beginning at 8-weeks of age. Osteocalcin immunohistochemistry of trabecular bone showed that the DT treated OC-Cre/iDTR mice were completely devoid of endosteal osteoblasts and young osteocytes. Consequently, serum levels of sclerostin were also significantly reduced in OC-Cre/iDTR mice compared to the controls.

To study MM engraftment and progression, 3x10⁶ 5TGM1-Luciferase MM cells were inject into tibia of OC-Cre/iDTR and control mice followed by a weekly injection of DT for 8-weeks. Bioluminescence imaging (BLI) was used to assess tumor progression. Both the control and OC-Cre/iDTR mice started with similar BLI signal at 1-week. Interestingly, 4-weeks onwards only the OC-iDTR mice continued to express and increase the BLI signal indicating that the MM cells engrafted and continued to proliferate only in the OC-Cre/iDTR mice. This data suggests that under physiological conditions mature osteoblasts actively suppress MM engraftment and progression.

To determine the direct effects of osteoblasts on MM cells, we established co-cultures of osteoblasts with MM cells. FACS sorting was used to isolate pre-osteoblasts from the long bones of Osterix-GFP+ (Osx+) mice, committed osteoblasts from Collagen 2.3-GFP+ (Col2.3+) mice, and mature osteoblasts from Osteocalcin-YFP+ (OCN+) mice. 5TGM1 MM cells were co-cultured along with 10,000 cells from each of the osteoblast populations. Although all three populations of osteoblasts significantly suppressed MM proliferation, the OCN+ mature osteoblasts suppressed MM proliferation the most.

The mature osteoblastic niche may regulate MM cells via a) direct cell-to-cell contact, b) secreted factors, and/or c) other intermediary cells. To identify these possibilities, the BM TME was examined at 2-weeks following the 5TGM1 intratibia injection into the OC-Cre/iDTR and control mice. Quantitative protein antibody arrays analysis on the BM supernatant identified numerous key factors involved in cell-cell communication and immunomodulation in MM. These factors included BAFF-R, TACI, IL-33, IL-3, IL-21, and IL-17F. Gene expression analysis of the sorted 5TGM1 cells from the injected tibia indicated increased expression of genes involved in mitochondrial metabolism.

Taken together, using in vivo and in vitro models, we show that mature osteoblasts offer specialized niches for MM cells where tumor cells are maintained in quiescence. The loss of the niche support allows the reactivation and progression of MM through the loss of cell-cell communication or through activation of immunomodulatory intermediary cells. Normalizing osteoblasts, such as through Activin A treatment, could provide novel avenues to reduce disease burden and long term tumor control.