CD166 Identifies Multiple Myeloma Cells That Preferentially Home to the Bone Marrow Microenvironment and Promote Disease Progression and Bone Lytic Disease

Multiple myeloma (MM) is an incurable malignancy characterized by the proliferation of neoplastic plasma cells in the bone marrow (BM) and by multiple lytic lesions throughout the skeleton. We previously reported that CD166 is a functional molecule on normal hematopoietic stem cells that plays a critical role in HSC homing to the BM and engraftment (Blood. 2014; 124(4):519-29.), suggesting that CD166 is involved in trafficking and lodgment of HSC. CD166, which is a member of the immunoglobulin superfamily capable of mediating both homophilic and heterophilic (CD6) interactions, has been shown to enhance metastasis and invasion in several tumors including breast cancer and melanoma. However, whether CD166 is involved in MM and whether it plays a role in MM progression or the bone lytic disease has not been addressed. H929-GFP MM cells were injected intravenously into NSG mice and GFP cells were recovered from the BM 14hr later. While only 29.9%±1.4% of total H929-GFP cells were CD166+, the frequency of CD166+ cells contained in BM-homed H929-GFP cells was significantly higher (80.0%±2.5%, n=9, p<0.01). Knocking-down (KD) CD166 expression on H929-GFP cells with shRNA reduced the homing of MM cells to the BM. The number of BM-homed GFP cells was significantly decreased for CD166KD cells (5658±904, n=6) compared to mock cells (8551±848, n=6; p<0.05). To examine the potential role of CD166 in osteolytic lesions, we used a novel Ex Vivo Organ Culture Assay (EVOCA) in which MM cells are co-cultured over calvariae from 10d-old pups for 7 days creating an in vitro 3D system for the interaction of MM cells with the bone microenvironment. EVOCA data from two MM cells lines, H929 and OPM2 as well as from three primary MM patients’ CD138+ BM cells demonstrated that bone osteolytic lesions were significantly reduced when CD166 was absent on either MM cells (sorted CD166- fraction) or osteoblast lineage cells (calvariae from CD166^-/- mice). Osteolytic lesions can be caused by decreased osteoblastogenesis or increased osteoclastogenesis. Co-culturing BM stromal cells (BMSC) from WT or CD166^-/- mice with mock or CD166KD H929 cells in osteoblastogenic media revealed that loss of CD166 on either MM cells or BMSC released the inhibition of mRNA expression of RUNX2, a critical transcriptional factor for early osteoblast differentiation, suggesting that interactions through CD166 suppresses osteoblastogenesis. Gene analysis of calvariae co-cultured with mock or CD166KD H929 cells revealed that the ratios of mRNA levels of RANKL to that of OPG, an important inhibitor of osteoclastogenesis, were elevated in cultures containing a CD166+ component suggesting that CD166 enhances osteoclastogenesis. Furthermore, co-cultures of BM monocytes (BMM) from WT or CD166^-/- mice with mock or CD166KD H929 cells in the presence of RANKL and MCSF confirmed that the absence of CD166 on either BMM or MM cells results in significantly reduced osteoclast differentiation as evidenced by TRAP staining. Given that Akt signaling is regulated by CD166 and that Akt pathway is important in osteoclast differentiation, we examined whether or not the Akt pathway is involved in CD166-induced osteoclastogenesis in MM. Western blot analysis of BMM from WT or CD166^-/- mice starved then stimulated with either mock or CD166 KD H929 cells revealed that phospho-Akt was significantly down-regulated in the absence of CD166 on either BMM or MM cells indicating that CD166 can activate the Akt pathway and promote MM-induced osteoclast differentiation. Finally, we compared in NSG mice the progression of MM initiated with control, mock or CD166KD H929-GFP cells (n=7-13/group, 2 independent experiments). Compared to control and mock groups, NSG mice transplanted with CD166KD cells showed delayed disease progression and prolonged survival as evidenced by serum levels of human IgA (kappa) and Kaplan-Meier survival curves (p<0.05), respectively. Meanwhile, the inhibition of osteocalcin mRNA expression in calvariae of NSG mice was significantly reduced in the presence of CD166KD cells compared to that with control or mock cells, suggesting again that expression of CD166 on MM cells inhibits osteoblastogenesis. These results suggest that CD166 directs MM cell homing and retention in the BM and promotes MM disease progression and propose that CD166 may induce osteoclastogenesis in MM through the Akt pathway. CD166 may serve as a therapeutic target in the treatment of MM.