CSF1 is the principal colony stimulating activity released by osteoblasts in response to PTH treatment. Its
receptor, c-fms, is more highly expressed on mature osteoclasts than any other cell in bone. We found that
deleting c-fms in osteoclasts attenuates the anabolic response to PTH. This indicates that part of PTH's anabolic
actions could be via a paracrine loop in which PTH stimulates expression of CSF1 in osteoblasts, which then
acts on osteoclasts to induce anabolic clastokines that augment bone formation. There are two major isoforms
of CSF1: soluble (sCSF1) and membrane-associated (mCSF1). We found that the anabolic response to PTH
is augmented in animals only expressing the sCSF1 isoform due to a greater increase in osteoblast
number in these mice compared to PTH-treated controls. In striking contrast, animals only expressing
mCSF1 had no increase in bone mass in response to an anabolic PTH regimen. Importantly, sCSF1 and
mCSF1 differ in the kinetics and extent to which they activate the CSF1 receptor, c-fms, suggesting that their
divergent in vivo actions may be due in part, to intrinsic differences in cell-signaling. Based on these data, we
hypothesize that mCSF1 inhibits PTH anabolism by opposing the actions of sCSF1 on osteoclasts. When
unopposed, sCSF1 contributes to PTH anabolism by inducing production of anabolic clastokines.
Specifically, single daily doses of PTH induce transient increases in sCSF1 in osteoblasts that cause
bursts of anabolic clastokine production. Consistent with this, we found that sCSF1 stimulates production of
the anabolic clastokine, sphingosine-1-phosphate (S-1-P) in osteoclasts. To test these hypotheses, we will, in
SA1, determine if adding intermittent dosing of sCSF1 to an anabolic PTH treatment regimen augments
the skeletal response to PTH in wild type animals, while adding mCSF1 to that regimen attenuates the
response. We will then try to restore the anabolic response to PTH in the sCSF1-/- mice by treating with PTH
plus sCSF1. These experiments will provide pharmacologic evidence that the very different response to PTH in
sCSF1-/- and mCSF1-/- mice is directly due to differing actions of the two CSF1 isoforms in bone. In SA 2, we
will treat wild type mice that were ovariectomized 5 months earlier, with PTH plus sCSF1 to determine if
it restores bone mass to pre-OVX levels as a model of therapy for established post-menopausal
osteoporosis. In SA 3, we will determine if sCSF1 and mCSF1 induce production of different clastokine
profiles in mature osteoclasts. We will use osteoblast/osteoclast cocultures as a model of in vivo paracrine
signaling in bone and profile the transcriptome of osteoclasts exposed to PTH-treated osteoblasts expressing
only sCSF1 or only mCSF1 to identify differences in the types of clastokines produced. We will also determine
differences in the transcriptomes of osteoblasts in the cocultures stimulated by these different clastokine profiles.
Finally, we will examine binding kinetics of sCSF1 and mCSF1 to c-fms, and differences in downstream signaling,
to gain molecular insight into the divergent in vivo and in vitro actions of these two isoforms.
