The bone marrow (BM) microenvironment plays critical role in the progression of multiple myeloma (MM) and identification of dysregulated microenvironmental pathways is essential for development of effective interventions for this disease. We and others have shown that bone disease is both a consequence of and a necessity for MM progression and that restoring bone turnover helps control MM. Prooxidant environment has been shown to promote tumorigenesis and bone disease, while circulating levels of antioxidants are reduced in MM patients (Sharma et al., 2009). Heme oxygenase 1 (HMOX1) is an intracellular inducible antioxidant that acts in response to oxidative stress and mediates anti-inflammatory and wound healing processes. The aim of the study was to identify dysregulated microenvironmental factors in myelomatous bone using mesenchymal stem cell (MSC) cytotherapy as an approach previously shown by us to induce bone formation and inhibit osteolysis and MM progression (Li et al., 2011; 2012). Intrabone injection of fetal MSCs into myelomatous bones of SCID-hu mice engrafted with a patient's MM cells resulted in significant upregulation(≥2 folds) of 120 probe sets and downregulation of 44 probe sets 24 hours after MSC cytotherapy, assessed by global gene expression profile (GEP, 9 mice/group). HMOX1, one of top overexpressed genes by MSC cytotherapy (5.2 folds, p<0.009), was further found to be underexpressed in myelomatous versus non-myelomatous bones from SCID-hu mice using GEP (p<0.04), and immunohistochemistry. HMOX1 was also induced in myelomatous bone following cytotherapy with healthy donors MSCs (n=3, p<0.04) and was consistently upregulated by MSCs cytotherapy in bones engrafted with myeloma cells from 4 patients (p<0.0001) using qRT-PCR. GEP of clinical biopsies revealed lower HMOX1 expression in newly diagnosed MM patients (n=352) than healthy donors (n=20, p<0.001) and normal levels in biopsies from patients at variable time during remission (n=98, p<0.0001 vs. active MM). In our Total Therapy 2 (TT2) and 3 (TT3) patients, using optimal HMOX1 signal cut-point based on statistical significance value, lower HMOX1 expression in biopsies was associated with poor survival in each of the trials; the 5-years estimated overall survival was 62% vs. 85% in TT2 (p<0.0001) and 57% vs. 81% in TT3 (p<0.002) patients with low and high HMOX1 expression, respectively. Survival analysis based on HMOX1 expression in purified plasma cells revealed similar significant trend in TT2 (p<0.003) but not TT3 (p<0.056) patients. Induction of HMOX1 in osteoclast precursors through noncontact co-culture with MSC or with the HMOX1 inducer, hemin, resulted in downregulation of RANK and NFATC1 and inhibition of osteoclast formation demonstrated by reduced expression of osteoclast markers cathepsin K, vitronectin receptor and acid phosphates 5, and by fewer numbers of TRAP-expressing multinucleated osteoclasts (p<0.0001). In the human hFOB preosteoblastic cell line, HMOX1 expression was increased during osteoblast differentiation and treatment with hemin or MSC conditioned media further stimulated mineral deposition in these cells. These data highlight the involvement of HMOX1 in controlling osteoclastogenesis and osteoblastogenesis. We conclude that HMOX1 is suppressed in experimental and clinical myelomatous bone and that higher expression is associated with superior outcome. Approaches to induce HMOX1 expression may help control bone disease and MM progression.
No relevant conflicts of interest to declare.
Asterisk with author names denotes non-ASH members.