S100A8, MMP13, and BMP3b: Likely Gene Targets of the Anti-Resorptive Effects of Erα Signaling on Cancellous and Cortical Bone and the Periosteal Response to Mechanical Loading, Respectively

Presentation Number: OR05-6
Date of Presentation: April 3rd, 2017

Stavros C Manolagas*1, Maria S Almeida2, Srividhya Iyer2, Ha-Neui Kim2, Li Han2, Mark S Johnson3, Haibo Zhao4 and Charles A O'Brien2
1University of Arkansas for Medical Sciences/Central Arkansas Veterans Healthcare System, Little Rock, AR, 2University of Arkansas for Med Science/Central Arkansas Veterans Healthcare System, Little Rock, AR, 3UMKC School of Dentistry, Kansas City, MO, 4University of Arkansas for Medical Sciences, Little Rock, AR

Abstract

Characterization of the skeletal phenotype of mice with conditional deletion of the estrogen receptor (ER)α has revealed that the anti-resorptive effects of estrogens on cancellous and endosteal bone result from ERα signaling on distinct cell types: myeloid lineage cells (targeted by LysM-Cre) and mesenchymal cells (targeted by Prx1 or Osx1-Cre), respectively. ERα signaling in Osx1+ cells is also required for optimal periosteal bone accrual – an event that critically depends on BMP2 signaling. In addition, mice with global ERα deletion fail to respond to mechanical loading, indicating that ERα signaling is indispensable for the anabolic response of bone to mechanical forces. We report here that 5 month old female mice with conditional deletion of the ERα in Osx1 cells failed to exhibit the expected increase in bone formation and periosteal bone accrual in response to mechanical loading of the ulna (~2500 micro-strains for 3 weeks); whereas mice with ERα deletion in mature osteoblasts/osteocytes (targeted by Dmp1) had normal loading response. To elucidate the ERα gene targets responsible for these effects, we isolated ERα deleted cells from our conditional mouse models and performed microarray analysis. The expression of Fas ligand, a purported target gene of estrogens in osteoclasts, was not altered in ERα deleted macrophages or mature osteoclasts; and in contrast to published findings by others, FasLgld/gld mice which lack functional FasL lost cortical and cancellous bone following OVX indistinguishably from FasL-intact controls. The highest up-regulated mRNAs in macrophages from ERαf/f;LysM-Cre mice, as compared to ERαf/f controls, encode the heterodimeric calcium binding proteins S100A8 and S100A9, which stimulate osteoclast formation via the Toll-like receptor 4 and activation of NF-kB. Silencing S100A8 greatly attenuated osteoclastogenesis and increased osteoclast apoptosis in vitro. The highest among up-regulated genes in ERα deleted Osx1+ cells from GFP+ ERαf/f Osx1-Cre mice was the matrix metalloproteinase 13, which promotes osteoclast fusion independent of its enzymatic activity. BMP3b a.k.a. GDF10 – an inhibitor of BMP2 signaling – was also up-regulated (14-fold) in the ERα deleted Osx1+ cells. Collectively, these results suggest that S100A8, MMP13, and GDF10/BMP3b are likely targets of the anti-resorptive effects of ERα signaling on cancellous and cortical bone and the periosteal response to mechanical loading, respectively. Furthermore, these findings support the notion that searching for ERα gene targets in ERα-deficient cells isolated from mice with cell-targeted deletions represents an advantageous and fruitful approach that establishes a priori that: a) these genes are expressed in cell targets that have been functionally validated in vivo and b) are contextually relevant to bone mass regulation in the whole animal.

 

Nothing to Disclose: SCM, MSA, SI, HNK, LH, MSJ, HZ, CAO