With increased usage, it was soon evident that the decrease in bone fracture incidence induced by bisphosphonates was disproportional to their antiresorptive properties and effects on bone mass, suggesting an additional effect on bone strength unrelated to the drugs’ actions on osteoclasts.

During this time of increased use, in the late 1990s, another bone cell was coming to the center stage: the osteocyte (3, 4). Osteocytes — the most abundant bone cells — were hypothesized to detect damaged bone and orchestrate its removal through the sophisticated osteocytic network, expanding the entire mineralized bone matrix and reaching to the bone surfaces. However, how osteocytes buried within the mineral could coordinate bone repair was not understood. With the generation of new osteocytic cell lines and the development of unique molecular means to target osteocytes in animal models, an avalanche of research demonstrated that untimely death of osteocytes could account for disruption of this network, leading to decreased bone quality and increased bone fragility. It was also shown that accumulations of apoptotic osteocytes mark areas of bone that need to be replaced, signal to osteoclast precursors, and initiate “targeted” remodeling, i.e., bone resorption in particular areas of the skeleton that need replacement (5).

In 1999, work from my lab published in the JCI showed that osteocytes (and osteoblasts) were target cells of bisphosphonates and that the drugs prevented the increased prevalence of apoptosis of these cells induced by excess of glucocorticoids (6) (Figure 1).

Distinct biological actions of bisphosphonates on bone cells explain the protective effects of the drugs on the skeleton. Bisphosphonates inhibit the mevalonate pathway in osteoclasts, inducing their apoptosis, leading to inhibition of bone resorption. Additionally, the drugs open Cx43 hemichannels, promoting survival of osteocytes and osteoblasts, maintaining bone strength and bone formation.

This paper had an unanticipated impact in the field because it dismantled a few then widely accepted notions. Furthermore, the research that followed, by our laboratory and the research community in general, changed forever the perception of bisphosphonates as monodimensional drugs. These findings continue to reverberate today. Personally, as a junior faculty at that time, I learned to trust my instincts, follow the data, and interpret research findings with candor and without fear. The work was also a demonstration of team science and provided important lessons on how to collaborate effectively, lessons that I have embraced in my scientific career.

Our work was simple and at the same time remarkable. It provided irrefutable evidence that osteoclasts were not the only bone target cells of bisphosphonates and that osteocytes contribute to bone strength by mechanisms beyond the control of bone mass. Furthermore, it was clear that, besides the recognized direct effect of the drugs on osteoclasts, bisphosphonates interfere with remodeling indirectly by preserving osteocyte viability and thus regulating targeted remodeling.

Another startling outcome of our research was the recognition that the molecular mechanism of the antiapoptotic effect on osteocytes/osteoblasts was unrelated to interference with the mevalonate pathway and that the antiapoptotic effect on osteocytes and osteoblasts was exerted at much lower concentrations than those needed for the effect of bisphosphonates on osteoclasts (6). This discovery opened the possibility that the integrity of the osteocyte network could be maintained without affecting osteoclasts directly, thereby avoiding an excessive decrease in remodeling. Indeed, through biological screens, we discovered osteocyte/osteoblast-selective bisphosphonate analogs that preserve the osteocyte network, bone formation, and bone strength without decreasing bone resorption (7).