6'-Methoxy Raloxifene-analog enhances mouse bone properties with reduced estrogen receptor binding

Abstract

Raloxifene (RAL) is an FDA-approved drug used to treat osteoporosis in postmenopausal women. RAL suppresses bone loss primarily through its role as a selective estrogen receptor modulator (SERM). This hormonal estrogen therapy promotes unintended side effects, such as hot flashes and increased thrombosis risk, and prevents the drug from being used in some patient populations at-risk for fracture, including children with bone disorders. It has recently been demonstrated that RAL can have significant positive effects on overall bone mechanical properties by binding to collagen and increasing bone tissue hydration in a cell-independent manner. A Raloxifene-Analog (RAL-A) was synthesized by replacing the 6-hydroxyl substituent with 6-methoxy in effort to reduce the compound's binding affinity for estrogen receptors (ER) while maintaining its collagen-binding ability. It was hypothesized that RAL-A would improve the mechanical integrity of bone in a manner similar to RAL, but with reduced estrogen receptor binding. Molecular assessment showed that while RAL-A did reduce ER binding, downstream ER signaling was not completely abolished. In-vitro, RAL-A performed similarly to RAL and had an identical concentration threshold on osteocyte cell proliferation, differentiation, and function. To assess treatment effect in-vivo, wildtype (WT) and heterozygous (OIM+/-) female mice from the Osteogenesis Imperfecta (OI) murine model were treated with either RAL or RAL-A from 8 weeks to 16 weeks of age. There was an untreated control group for each genotype as well. Bone microarchitecture was assessed using microCT, and mechanical behavior was assessed using 3-point bending. Results indicate that both compounds produced analogous gains in tibial trabecular and cortical microarchitecture. While WT mechanical properties were not drastically altered with either treatment, OIM+/- mechanical properties were significantly enhanced, most notably, in post-yield properties including bone toughness. This proof-of-concept study shows promising results and warrants the exploration of additional analog iterations to further reduce ER binding and improve fracture resistance.

Keywords: Bone mechanics; Bone quality; Osteogenesis Imperfecta; SERM.

Fig. 1

Structures of RAL (1) and RAL-A (2). RAL possesses a 6-hydroxy substituent, while RAL-A possesses a 6-methoxy substituent.

Fig. 2

Cells were treated with RAL (A) or RAL-A (B) at concentrations of 0 nM, 1 nM, 10 nM, 100 nM, 1 μM, and 10 μM. Absorbance was normalized to the 24 h 0 nM value for each compound. Cell proliferation had no qualitative impact until treatment reached 10 μM. (C) Alkaline Phosphatase and Alizarin Red staining for cells treated with RAL or RAL-A for concentrations at 0 nM, 1 nM, 10 nM, 100 nM, 1 μM, 10 μM. Qualitatively, cells were able to mineralize a matrix until treatment reached 10 μM.

Fig. 3

(A) The ERα binding assay indicates the compounds ability to bind to ERα and displace a fluorochrome tracer, measured by fluorescence polarization. Polarization was measured for 17β estradiol, RAL, and RAL-A for concentrations ranging from 10⁻¹⁰ to 10⁻⁶ M. The IC₅₀ value (50% tracer displaced) for 17β estradiol was 19.52 nM, RAL was 9.28 nM, and RAL-A was 183.2 nM. (B) Expression of C3 (multiplied ×1000) in MLO-Y4 osteocytic cells after being treated with vehicle (DMSO), 17β estradiol (10⁻⁸ M), RAL (10⁻⁷, 10⁻⁸, and 10⁻⁹ M), or RAL-A (10⁻⁷, 10⁻⁸, and 10⁻⁹ M). A significant change from vehicle is indicated by ‘*’ at p < 0.05. There was a significant increase noted for RAL at 10 nM and RAL-A at 100 nM compared to vehicle.

Fig. 4

Treatment effects on trabecular microarchitecture for (A) bone volume fraction (BV/TV), (B) bone mineral density (BMD), (C) trabecular thickness, (D) trabecular number, (E) trabecular separation, and (F) tissue mineral density (TMD). Significant change from control at p < 0.05 is indicated by ‘*’ within each genotype. Schematic representation of the average cortical profile for each treatment group compared to respective control for (G) Wildtype and (H) Heterozygous OIM+/− samples.

Fig. 5

Treatment effects on mechanical properties for (A) yield force, (B) ultimate force, (C) total displacement, (D) total work, (E) yield stress, (F) ultimate stress, (G) total strain, and (H) toughness. Significant change from control at p < 0.05 is indicated by ‘*’ within each genotype.