Controversy of physiological vs. pharmacological effects of BMP signaling: Constitutive activation of BMP type IA receptor-dependent signaling in osteoblast lineage enhances bone formation and resorption, not affecting net bone mass

Abstract

Bone morphogenetic proteins (BMPs) were first described over 50 years ago as potent inducers of ectopic bone formation when administrated subcutaneously. Preclinical studies have extensively examined the osteoinductive properties of BMPs in vitro and new bone formation in vivo. BMPs (BMP-2, BMP-7) have been used in orthopedics over 15 years. While osteogenic function of BMPs has been widely accepted, our previous studies demonstrated that loss-of-function of BMP receptor type IA (BMPR1A), a potent receptor for BMP-2, increased net bone mass by significantly inhibiting bone resorption in mice, indicating a positive role of BMP signaling in bone resorption. The physiological role of BMPs (i.e. osteogenic vs. osteoclastogenic) is still largely unknown. The purpose of this study was to investigate the physiological role of BMP signaling in endogenous long bones during adult stages. For this purpose, we conditionally and constitutively activated the Smad-dependent canonical BMP signaling thorough BMPR1A in osteoblast lineage cells using the mutant mice (Col1CreER™:caBmpr1a). Because trabecular bones were largely increased in the loss-of-function mouse study for BMPR1A, we hypothesized that the augmented BMP signaling would affect endogenous trabecular bones. In the mutant bones, the Smad phosphorylation was enhanced within physiological level three-fold while the resulting gross morphology, bodyweights, bone mass/shape/length, serum calcium/phosphorus levels, collagen cross-link patterns, and healing capability were all unchanged. Interestingly, we found; 1) increased expressions of both bone formation and resorption markers in femoral bones, 2) increased osteoblast and osteoclast numbers together with dynamic bone formation parameters by trabecular bone histomorphometry, 3) modest bone architectural phenotype with reduced bone quality (i.e. reduced trabecular bone connectivity, larger diametric size but reduced cortical bone thickness, and reduced bone mechanical strength), and 4) increased expression of SOST, a downstream target of the Smad-dependent BMPR1A signaling, in the mutant bones. This study is clinically insightful because gain-of-function of BMP signaling within a physiological window does not increase bone mass while it alters molecular and cellular aspects of osteoblast and osteoclast functions as predicted. These findings help explain the high-doses of BMPs (i.e. pharmacological level) in clinical settings required to substantially induce a bone formation, concurrent with potential unexpected side effects (i.e. bone resorption, inflammation) presumably due to a broader population of cell-types exposed to the high-dose BMPs rather than osteoblastic lineage cells.

Keywords: BMPR1A; Bone morphogenetic protein; Osteoblast lineage cells; Pharmacologic; Physiologic.

Figure 1.. Generation of the Bmpr1a^ca mice

(A) Schematic representation of the transgenic constructs in the Bmpr1a^ca mice. The mice conditionally express a constitutively active form of human Bmpr1a (caBmpr1a) following Cre-dependent DNA recombination. (B) Cre-dependent DNA recombination detected by β-galactosidase (β-gal) staining. ROSA26 Cre reporter mice crossed with Col1CreER™ mice following tamoxifen administration showed Cre activity in osteoblast lineage cells at 10 weeks of age. Cre activity was similarly detected both in trabecular and cortical bones. There was no gender difference in Cre activity. Bars; 20 μm (C) Gross image of the Bmpr1a^ca male mice at 34 weeks of age. No overt change was observed in male and female of Bmpr1a^ca mice compared to controls (control: n = 18, Bmpr1a^ca: n = 15). (D) Body weight of the Bmpr1a^ca mice at 34 weeks of age. There was no change in body weight between Bmpr1a^ca mice and control mice both in male and female (control: n = 18, Bmpr1a^ca: n = 15). (E) Expression levels of Id1, a target gene of BMP signaling, as well as those of Sost, one of downstream targets of BMPR1A, were measured in the Bmpr1a^ca male mice at 12 weeks of age. The expression levels of Id1 and Sost were both significantly higher in the Bmpr1a^ca bones by 5-fold and 4-fold, respectively (control: n = 4, Bmpr1a^ca: n = 4). Values are expressed as mean + SD. *p < 0.05 (F) Immunohistochemical detection of phosphorylated form of Smad1/5/9 in the Bmpr1a^ca femora at 34 weeks of age (i.e. male). Osteoblast lineage cells from comparable area were magnified. The ratio of positive cells to total cells in bone area was significantly higher by 3-times in the Bmpr1a^ca bones (Control: 7.2%, Bmpr1a^ca: 20.5%). Bars; 50 μm, **p < 0.01

Figure 2.. Gene expression for formation and resorption markers in the Bmpr1a^ca mice

(A) Bone formation markers (Cbfa1, Osterix, Akp2, Ibsp, and Col1a1) were analyzed in the Bmpr1a^ca male mice at 12 weeks of age (control: n = 4, Bmpr1a^ca: n = 4). (B) Bone resorption markers (Mmp9, Ctsk, Trap, Rankl, and Opg) and the Rankl/Opg ratio were analyzed in the Bmpr1a^ca male mice at 12 weeks of age. All values are expressed as mean + SD. *p < 0.05

Figure 3.. Skeletal morphology in the Bmpr1a^ca mice

(A) Radiological assessment of the skeleton (i.e. femora, spine) in the Bmpr1a^ca male mice at 34 weeks (control: n = 9, Bmpr1a^ca: n = 6). (B) Bone mineral density of femora and lumbar assessed by DEXA in the Bmpr1a^ca male mice at 34 weeks (control: n = 9, Bmpr1a^ca: n = 6). (C) Serum levels of calcium and phosphorus in the Bmpr1a^ca male mice at 12 weeks (control: n = 8, Bmpr1a^ca: n = 8).

Figure 4.. Morphological assessment of femora and lumbar in the Bmpr1a^ca mice

(A) H&E staining of femora and lumbar (L3) in the Bmpr1a^ca male mice at 34 weeks compared to controls (control: n = 9, Bmpr1a^ca: n = 6). Bars; 1 mm (B) Micro-CT based 2D and 3D images of femora (i.e. cortical and trabecular bones) and lumbar (i.e. trabecular bone) in the Bmpr1a^ca male mice at 34 weeks as presented in Table 1. Bars; 1 mm

Figure 5.. Bone histomorphometry in the Bmpr1a^ca mice

(A) Bone volume (BV/TV), trabecular thickness (TbTh), trabecular space (TbSp), and trabecular number (TbN) were analyzed in the Bmpr1a^ca lumbar at 34 weeks compared to controls both in male (control: n = 7, Bmpr1a^ca: n = 5) and female (control: n = 6, Bmpr1a^ca: n = 5). (B) Osteoblast surface (ObS/BS), osteoclast surface (OcS/BS), osteoclast number (NOc/BS) were analyzed in the Bmpr1a^ca lumbar at 34 weeks compared to controls both in male (control: n = 7, Bmpr1a^ca: n = 5) and female (control: n = 6, Bmpr1a^ca: n = 5). A modest to large effect size was determined using Cohen's d; (a) p < 0.05, d = 1.51, (b) p > 0.05, d = 0.82, (c) p > 0.05, d = 0.78, (d) p > 0.05, d = 0.78. All values are expressed as mean + SD. *p < 0.05 (C) Mineralizing surface (MS/BS), mineral apposition rate (MAR), and bone formation rate (BFR/BS) were analyzed in the Bmpr1a^ca femora at 34 weeks (i.e. male), A modest to large effect size was determined using Cohen's d; (e) p > 0.05, d = 0.80, (f) p > 0.05, d = 1.33, (g) p > 0.05, d = 1.23. All values are expressed as mean + SD.

Figure 6.. Mechanical test in the Bmpr1a^ca male mice

(A-F) Ultimate displacement, stiffness, failure strength, failure load, Young’s modules, and cortical bending moment of inertia were measured in the femora of Bmpr1a^ca male mice at 34 weeks (control: n = 9, Bmpr1a^ca: n = 7). A large effect size was determined using Cohen's d; (a) p = 0.09, d = 0.89, (b) p = 0.08, d = 0.92, (c) p = 0.05, d = 1.0. All values are expressed as mean + SD. *p < 0.05

Figure 7.. Rib fracture experiment in the Bmpr1a^ca male mice

(A) Fractures were introduced at two adjacent rib bones per mice at 10 weeks (i.e. male). Radiological assessment of rib fracture healing at 16 weeks of age, post-surgery 6 weeks. Type I: both two fractured sites were united, Type II: one of two fractured sites were united, Type III: both two fractured sites were non-union. (B) Union rate was compared between control rib bones (n = 12 mice) and Bmpr1a^ca rib bones (n = 9 mice). (C) A frequency of fracture type I, II, and III was compared between control and Bmpr1a^ca rib bones. (D) H&E staining of fractured site at 16 weeks of age, post-surgery 6 weeks. Yellow arrows indicate a gap of two ends in the non-union fractured site. Bars; 0.5 mm