Deletion of Estrogen Receptor Beta in Osteoprogenitor Cells Increases Trabecular but Not Cortical Bone Mass in Female Mice

Abstract

Although the role of ERα in regulating bone metabolism has been extensively studied, ERβ has been largely dismissed as a relevant modulator of bone mass. Previous studies examining ERβ utilized a germline knockout mouse expressing transcript variants of ERβ and displaying systemic hormonal changes that confounded interpretation of the skeletal phenotype. Thus, we used a conditional ERβ mouse model to achieve deletion of ERβ specifically in early osteoprogenitor cells using the Prx1-Cre driver. We observed marked increases in the trabecular bone volume fraction (of 58% [p < 0.003] and 93% [p < 0.0003] in 6- and 12-week-old female ERβ(Prx1-CKO) mice, respectively) but no changes in cortical bone. Serum estradiol and IGF-I levels were unaltered in ERβ(Prx1-CKO) mice. Bone formation and resorption indices by histomorphometry and serum assays were unchanged in these mice, suggesting that alterations in bone turnover may have occurred early in development. However, the ratio of colony-forming unit-osteoblasts (CFU-OBs) to CFU-fibroblasts (CFU-Fs) was increased in bone marrow cultures from ERβ(Prx1-CKO) compared with control mice, indicating increased differentiation of osteoblast precursor cells into osteoblasts in ERβ(Prx1-CKO) mice. Detailed quantitative polymerase chain reaction analyses of 128 genes in 16 prespecified pathways revealed significant downregulation of 11 pathways in ERβ(Prx1-CKO) mice. Thus, deletion of ERβ specifically in osteoblast lineage cells, in the absence of all splice variants, increases trabecular bone mass and modulates multiple pathways related to bone metabolism. These findings suggest that pharmacological inhibition of ERβ in bone may provide a novel approach to treat osteoporosis.

Keywords: BONE; ESTROGEN RECEPTOR; MICROSTRUCTURE.

Fig. 1

ERβ deletion in osteoprogenitor cells. Genomic DNA was isolated from either ex vivo bone marrow stromal cells (BMSCs) of the long bones of 12-week-old control and ERβ^Prx1–CKO mice cultured under osteogenic conditions or estrogen-responsive tissues harvested at the time of bone marrow isolation. DNA rearrangement of ERβ exon 3 was determined using primers that span exon 3. L = DNA ladder; C = control.

Fig. 2

Trabecular parameters by μCT analyses in 12-week-old ERβ^Prx1–CKO mice. (A) Representative reconstructed images of distal femur trabecular bone in 12-week-old control or ERβ^Prx1–CKO mice. Changes in (B) bone volume to tissue volume (BV/TV), (C) trabecular number (Tb.N), (D) trabecular thickness (Tb.Th), and (E) trabecular separation (Tb.Sp) (n = 16 per group).

Fig. 3

Cortical parameters by μCT analyses in 12-week-old ERβ^Prx1–CKO mice. (A) Representative reconstructed image of femur cortical bone (diaphysis) in 12-week-old control or ERβ^Prx1–CKO mice. Changes in (B) cortical volumetric bone mineral density (Ct.vBMD) and (C) cortical thickness (Ct.Th) (n = 16 per group).

Fig. 4

Changes in CFU-Fs and CFU-OBs with ERβ deletion in osteoprogenitor cells. CFU-Fs and CFU-OBs were assessed in ex vivo bone marrow stromal cell cultures of cells isolated from long bones of 12-week-old control or ERβ^Prx1–CKO mice. Changes in (A) total colonies (CFU-F), (B) mineralized bone nodules (CFU-OB), and (C) ratio of mineralized bone nodules to total colonies formed (CFU-OB/CFU-F) (n = 4 wells per animal/16 animals per group).

Fig. 5

ERβ knockdown enhances ERα-specific transcriptional activity. MC3T3-E1 cells were cotransfected with either a negative control or ERβ-specific siRNA and an ERE-luciferase reporter construct. (A) QPCR analyses of ERβ expression in both the negative control and ERβ-specific siRNA-transfected cells to account for knockdown efficiency (n = 3). The asterisk represents p <0.01 compared with the negative control siRNA. (B) A parallel set of identically transfected cells were treated with either vehicle control or the ERα-specific agonist PPT (10 nM). Twenty-four hours later, the cells were harvested and luciferase and protein assays were performed. The data are expressed as luciferase activity/μg protein, graphed relative to the vehicle control condition for each siRNA and represent mean ± SD (n = 6). A single asterisk (*) represents p <0.01 compared with the vehicle control for each siRNA condition. A double asterisk (**) represents p <0.01 comparing the PPT-treated cells between the negative control and ERβ-specific siRNA conditions.