Gene Modified Human Mesenchymal Stem Cells Expressing Osteoprotegerin Inhibit Osteoclast Activation and Increase Trabecular Bone Area in a Xenogeneic Murine Model of Myeloma.

Bone disease in multiple myeloma (MM) results from increased osteoclast (OCL) numbers and activity, which is associated with an increase in RANK Ligand and reduction in osteoprotegerin (OPG). Systemic administration of recombinant OPG reduces MM bone disease, but the short half life of OPG limits its usefulness. Gene modified mesenchymal stem cells (MSCs) offer a potential means of delivering stable expression of OPG in vivo to reduce OCL activation and bone destruction.

Bone marrow derived human MSCs were transduced with a self-inactivating bicistronic lentiviral vector containing human OPG and GFP (MSC^OPG). Control vector was identical except the OPG was cloned in reverse orientation (MSC^GPO). Efficient transduction was demonstrated by high GFP expression (96% MSC^OPG, 92% (MSC^GPO). Stable transgene expression of human OPG (hOPG) occurred for beyond 20 passages in vitro, and hOPG was detected in vivo after tail vein administration of MSC^OPG (2ng/mL hOPG detected in mouse serum 1 week after tail vein administration of 3 x 106 MSC^OPG). Immunophenotype and differentiation potential of MSCs were maintained following transduction.

A xenogeneic model of MM was developed. 1 x 10⁷ KMS-12-BM cells injected tail vein into b2 m NOD/SCID mice leads to tumour infiltration in the bone marrow at 6 weeks, with varied tumour take between the bones examined. Using histomorphometric analysis trabecular bone area (TBA) was assessed as the proportion of trabecular bone in 0.5625 mm² of marrow space 0.2 mm from growth plate. OCL were recorded as the proportion lining the endocortical surface (%OcPm). Reduction of trabecular bone in the tibia is related to the amount of tumour (KMS-12-BM tibia with >70% tumour mean TBA 0.7+/− 0.2 vs. KMS-12-BM tibia with <70% tumour mean 5.1+/− 0.8, p<0.01, which is similar to non diseased animals). All subsequent analysis were carried out on tibia with >70% tumour. There was no change in trabecular bone in the lumbar vertebrae. OCL were increased in the tibia and lumbar vertebrae of tumour bearing mice (PBS group mean %OcPm 0.9+/− 0.3 and 1.1+/− 0.4 vs. KMS-12-BM group mean 7.2+/− 3.2 and 7.5 +/− 2.2 in tibia and lumbar vertebrae respectively, p=0.01 in both groups).

We hypothesised that MSCs expressing OPG will prevent the increase in OCL and subsequent loss of trabecular bone. Infusion of unmanipulated MSC or MSC^GPO had no effect on %OcPm or TBA in diseased animals. 1 x10⁶ MSC^OPG or MSC^GPO were injected by tail vein 2, 3 and 4 weeks after KMS-12-BM injection. Another group received KMS-12-BM alone. All mice were culled at 6 weeks. Trabecular bone was increased in the tibia of tumour bearing mice treated with MSC^OPG (mean TBA 1.4 +/− 0.5) compared to control animals receiving MSC^GPO or tumour alone (mean TBA 0.6 +/− 0.2), p=0.03, with a trend showing a reduction of OCL in the tibia of the MSC^OPG group (mean %OcPm 2.6+/− 1.0) vs. control group (mean %OcPm 4.2+/− 1.5, NS). Importantly in the lumbar vertebrae, OCL were reduced in the MSC^OPG group (mean %OcPm 1.9 +/− 0.4) compared to control animals (mean %OcPm 3.5+/− 0.5), p<0.01.

Conclusion: MSCs gene modified with OPG are able to increase TBA in the tibia and reverse OCL activation in a xenogeneic model of MM. Gene modified MSCs may have future potential in treating MM induced bone disease.