Efficient generation of osteoclasts from human induced pluripotent stem cells and functional investigations of lethal CLCN7-related osteopetrosis

Human induced pluripotent stem cells (hiPSCs) hold great potential for modelling human diseases and the development of innovative therapeutic approaches. Here, we report on a novel, simplified differentiation method for forming functional osteoclasts from hiPSCs. The three-step protocol starts with embryoid body formation, followed by hematopoietic specification, and finally osteoclast differentiation. We observed continuous production of monocyte-like cells over a period of up to nine weeks, generating sufficient material for several osteoclast differentiations. The analysis of stage-specific gene and surface marker expression proved mesodermal priming, the presence of monocyte-like cells, and of terminally differentiated multinucleated osteoclasts, able to form resorption pits and trenches on bone and dentine in vitro. In comparison to peripheral blood mononuclear cell (PBMC)-derived osteoclasts hiPSC-derived osteoclasts were larger and contained a higher number of nuclei. Detailed functional studies on the resorption behaviour of hiPSC-osteoclasts indicated a trend towards forming more trenches than pits and an increase in pseudo-resorption. We used hiPSCs from an ARO patient (BIHi002-A, ARO hiPSCs) with compound heterozygous missense mutations p.(G292E) and p.(R403Q) in CLCN7, coding for the Cl(-)/H(+) -exchanger ClC-7, for functional investigations. The patient's leading clinical feature was a brain malformation due to defective neuronal migration. Mutant ClC-7 displayed residual expression and retained lysosomal co-localization with OSTM1, but only ClC-7 harboring the mutation p.(R403Q) gave strongly reduced ion currents. An increased autophagic flux in spite of unchanged lysosomal pH was evident in undifferentiated ARO hiPSCs. ARO hiPSC-derived osteoclasts showed an increased size compared to hiPSCs of healthy donors. They were not able to resorb bone, indicating a loss-of-function effect of the mutations. In summary, we developed a highly reproducible, straightforward hiPSC-osteoclast differentiation protocol. We demonstrated that osteoclasts differentiated from ARO-hiPSCs can be used as a disease model for ARO and potentially also other osteoclast-related diseases.