Parental Origin of Gsα Inactivation Differentially Affects Bone Remodeling in a Mouse Model of Albright Hereditary Osteodystrophy

Abstract

Albright hereditary osteodystrophy (AHO) is caused by heterozygous inactivation of GNAS, a complex locus that encodes the alpha-stimulatory subunit of heterotrimeric G proteins (Gsα) in addition to NESP55 and XLαs due to alternative first exons. AHO skeletal manifestations include brachydactyly, brachymetacarpia, compromised adult stature, and subcutaneous ossifications. AHO patients with maternally-inherited GNAS mutations develop pseudohypoparathyroidism type 1A (PHP1A) with resistance to multiple hormones that mediate their actions through G protein-coupled receptors (GPCRs) requiring Gsα (eg, parathyroid hormone [PTH], thyroid-stimulating hormone [TSH], growth hormone-releasing hormone [GHRH], calcitonin) and severe obesity. Paternally-inherited GNAS mutations cause pseudopseudohypoparathyroidism (PPHP), in which patients have AHO skeletal features but do not develop hormonal resistance or marked obesity. These differences between PHP1A and PPHP are caused by tissue-specific reduction of paternal Gsα expression. Previous reports in mice have shown loss of Gsα causes osteopenia due to impaired osteoblast number and function and suggest that AHO patients could display evidence of reduced bone mineral density (BMD). However, we previously demonstrated PHP1A patients display normal-increased BMD measurements without any correlation to body mass index or serum PTH. Due to these observed differences between PHP1A and PPHP, we utilized our laboratory's AHO mouse model to address whether Gsα heterozygous inactivation differentially affects bone remodeling based on the parental inheritance of the mutation. We identified fundamental distinctions in bone remodeling between mice with paternally-inherited (GnasE1+/-p) versus maternally-inherited (GnasE1+/-m) mutations, and these findings were observed predominantly in female mice. Specifically, GnasE1+/-p mice exhibited reduced bone parameters due to impaired bone formation and enhanced bone resorption. GnasE1+/-m mice, however, displayed enhanced bone parameters due to both increased osteoblast activity and normal bone resorption. These in vivo distinctions in bone remodeling between GnasE1+/-p and GnasE1+/-m mice could potentially be related to changes in the bone microenvironment driven by calcitonin-resistance within GnasE1+/-m osteoclasts. Further studies are warranted to assess how Gsα influences osteoblast-osteoclast coupling. © 2021 The Authors. JBMR Plus published by Wiley Periodicals LLC on behalf of American Society for Bone and Mineral Research.

Keywords: BONE; DISEASES AND DISORDERS OF/RELATED TO BONE – OTHER; GENETIC ANIMAL MODELS; Gnas; MOLECULAR PATHWAYS – REMODELING; OSTEOBLASTS; OSTEOCLASTS.

Conflict of interest statement

The authors declare that they have no disclosures or conflicts of interest.

© 2021 The Authors. JBMR Plus published by Wiley Periodicals LLC on behalf of American Society for Bone and Mineral Research.

Figures

Fig 1

Cortical bone parameters are dependent upon inheritance pattern of Gsα mutation. (A) Gsα mRNA expression is reduced within flushed tibia samples of Gnas E1+/−m and Gnas E1+/−p mice compared to WT. (B) Representative cross‐sectional images of cortical bone of distal femur across both WT and Gnas E1+/− animals. (C) Both male and female Gnas E1+/−p mice had a significantly reduced femur length compared to WT. Male Gnas E1+/−m mice also had a significantly reduced femur length compared to WT but no significant variations were observed among female Gnas E1+/−m and WT mice. (D) Phenotypic differences observed within cortical bone of both Gnas E1+/−p and Gnas E1+/−m animals are not attributed solely to decreased length, based on the lack of significant differences seen within the ratio of cortical area to total bone area; (E–H) Statistically significant decreases in cortical bone parameters were observed in Gnas E1+/−p compared to Gnas E1+/−m in both male and female animals. Gnas E1+/−p animals demonstrated significant decreases in cortical bone parameters compared to Gnas E1+/−m mice, and displayed normal to reduced parameters when compared to WT. Conversely, Gnas E1+/−m demonstrated normal to increased cortical bone parameters compared to WT. Sample size per genotype per experiment is listed on each bar graph. All statistical tests completed using ANOVA with post‐hoc Tukey test for multiple comparisons, and p values are displayed for each comparison.

Fig 2

Femur and vertebral trabecular bone parameters are mildly influenced by Gsα heterozygous inactivation. Representative μCT images of trabecular bone from the (A) distal femur and (B) L5 vertebrae of WT and Gnas E1+/− animals. (C–F) Gnas E1+/−p mice, when compared to WT, displayed no significant femur trabecular bone phenotype. However, female Gnas E1+/−m mice when compared to WT displayed a significantly elevated BV/TV, trabecular number, and a reduction in trabecular spacing (no variations observed between male mice). Gnas E1+/−m mice, when compared to Gnas E1+/−p, displayed a significant increase in trabecular thickness. (G–J) No significant differences were observed between WT and Gnas E1+/− mice with respect to vertebral trabecular bone parameters. Sample size per genotype per experiment is listed on each bar graph. All statistical tests completed using ANOVA with post‐hoc Tukey test for multiple comparisons, and p values are displayed for each comparison.

Fig 3

Enhanced cortical and trabecular bone parameters in Gnas E1+/−m mice are directly associated with elevated osteoblast activity despite a reduction in osteoblast number. (A) Representative images of undecalcified femur cortical bone sections containing calcein (green) and alizarin complexone (red) labels for dynamic histomorphometry. (B) Significant reductions in total AP activity on the bone surface (AP_BS) were observed within female Gnas E1+/−p and Gnas E1+/−m femur sections compared to WT. No significant differences were observed in male specimens across all genotypes. (C) Gnas E1+/−m mice displayed a reduction in the total number of actively mineralizing osteoblasts (AP_L_BS) when compared to WT and Gnas E1+/−p mice. (D) Dynamic histomorphometry on the femur trabecular surface revealed Gnas E1+/−m mice displayed a significant reduction in MS/BS within the femur when compared to both WT and Gnas E1+/−p mice (E) Female Gnas E1+/−p mice had a significantly reduced mineral apposition rate compared to both Gnas E1+/−m and WT. MAR among Gnas E1+/−m was comparable to WT. We observed no significant variations within male mice. (F) Gnas E1+/−m mice displayed a significantly increased MAR on the femoral endosteal surface compare to WT mice, as well as when compared to Gnas E1+/−p mice for males. (G) Serum P1NP levels were significantly elevated within Gnas E1+/−m mice compared to Gnas E1+/−p and WT samples. No significant variations were observed between Gnas E1+/−p and WT samples. (H) Bglap1, Dmp1, and Ibsp mRNA expression in flushed tibia specimens were significantly reduced in Gnas E1+/−p male and female mice compared to WT. Gnas E1+/−m samples displayed no significant reduction in Bglap and Ibsp expression when compared to WT but displayed a reduction in Dmp1 expression. Gnas E1+/−m samples, when compared to Gnas E1+/−p, displayed a significantly elevated Bglap1 and Dmp1 expression, but displayed no significant variations in Ibsp expression. Sample size per genotype per experiment is listed on each bar graph. All statistical tests completed using ANOVA with post‐hoc Tukey test for multiple comparisons, and p values are displayed for each comparison. AP_BS = Alkaline phosphatase positive cells on the bone lining surface (indicative of total number of osteoblasts); AP_L_BS = Alkaline phosphatase positive cells on the bone lining surface superimposed over a mineral label (indicative of actively mineralizing osteoblasts).

Fig 4

Inheritance pattern of Gsα mutation does not differentially influence osteoprogenitor differentiation in vitro. (A) Representative entire‐well images of: alkaline phosphatase positive (AP+) colony formation units within primary BMSCs at day 0 from male and female mice; Alizarin Red staining to detect calcium deposition on culture days 7 and 14; and Von Kossa staining to detect phosphate on culture day 14. (B) Both Gnas E1+/−m and Gnas E1+/−p demonstrated no significant differences within the osteoprogenitor populations compared to WT BMSCs as measured by AP colony forming unit assays. (C) Gnas E1+/− and WT BMSCs following 7 and 14 days of osteogenic differentiation displayed no significant variations in Alizarin Red staining absorbance. (D‐G) Both Gnas E1+/−m and Gnas E1+/−p BMSCs displayed significant reductions in Gsα expression at each time point when compared to WT. Prior to osteogenic differentiation, no significant differences were observed in (E) Alpl, (F) Sp7, and (G) Bglap1 mRNA expression among Gnas E1+/− and WT BMSCs. However, Gnas E1+/−m mice displayed elevated mRNA expression of (E) Alpl and (F) Sp7 after 7 days of osteogenic differentiation and (E) Alpl and (G)Bglap1 after 14 days of osteogenic differentiation when compared to both WT and Gnas E1+/−p. Sample size per genotype per experiment is listed on each bar graph. For AP+ colony assay, statistical tests were completed using a one‐way ANOVA with post‐hoc Tukey test for multiple comparisons. For alizarin red staining and RT‐PCR analyses at multiple time points, statistical tests were completed using a two‐way ANOVA with post‐hoc Tukey test for multiple comparisons. Values of p are displayed for each comparison.

Fig 5

Gnas E1+/−m osteoclasts display elevated Sphk1 expression in vitro. (A) Representative images of chromogenic TRAP staining (top) and phalloidin‐DAPI fluorescence microscopy staining of primary BMM cultures (bottom) from male and female mice in vitro. (B) Gnas E1+/−m and Gnas E1+/−p demonstrated increased rates of osteoclastogenesis and osteoclast surface area compared to WT as indicated by increased osteoclast number following 5 days of differentiation. Osteoclasts were defined as TRAP(+) cells with ≥3 nuclei. No significant differences were observed in osteoclast number between Gnas +/−p and Gnas +/−m cultures, respectfully; (C) Gsα mRNA expression was significantly reduced in both Gnas E1+/−m and Gnas E1+/−p cultures when compared to WT. No significant variations were observed in mRNA expression of Nfatc1 or Ctsk among Gnas E1+/− cultures and WT at day 5 of osteoclast differentiation. However, both Gnas E1+/−p and Gnas E1+/−m cultures demonstrated a significant increase in Rank expression compared to WT. (D) Gnas E1+/−m cultures display significantly elevated Sphk1 mRNA expression when compared to WT cultures. Sample size per genotype per experiment is listed on each bar graph. All statistical tests completed using ANOVA with post‐hoc Tukey test for multiple comparisons, and p values are displayed for each comparison.

Fig 6

Gsα heterozygous inactivation differentially affects bone resorption in a parental inheritance‐specific pattern in vivo. (A) Representative images of undecalcified distal femur sections demonstrating TRAP enzymatic activity on the bone surface (TRAP_BS) following incubation with Elf97 fluorescent substrate. (B) Representative images of distal femur trabeculae stained with TRAP for visualization of osteoclasts on the bone surface; counterstained with aniline blue. (C,D) Female Gnas E1+/−p mice demonstrated a significant increase in (C) total TRAP activity on the bone surface (TRAP_BS) and (D) active bone remodeling sites (TRAP_L_BS) within the femur when compared to WT and Gnas E1+/−m mice. No significant variations were observed between female Gnas E1+/−m and WT mice. No significant phenotype was observed within male mice. (E) Chromogenic TRAP staining revealed female Gnas E1+/−p mice demonstrated a significant increase in the ratio of osteoclast surface/bone surface when compared to both WT and Gnas E1+/−m mice. No significant variations were observed between female Gnas E1+/−m and WT mice. No significant phenotype was observed within male mice. (F) Female Gnas E1+/−p mice displayed a significant increase in the number of osteoclasts per bone surface when compared to WT mice. No significant variations were observed between female Gnas E1+/−m and WT mice or between Gnas E1+/−m and Gnas E1+/−p mice. No significant phenotype was observed within male mice. (G) Female Gnas E1+/−p mice displayed significantly elevated fasting serum CTX‐1 measurements when compared to WT. (H) RT‐PCR analysis of flushed tibia diaphysis from male and female mice identified both Gnas E1+/−p and Gnas E1+/−m mice had reduced mRNA expression of Tnfsf11 (Rankl) and Tnfsf11b (Opg) when compared to WT. Gnas E1+/−m mice displayed significantly increased Rankl and Opg mRNA expression when compared to Gnas E1+/−p. However, Gnas E1+/−p mice displayed a significantly elevated Rankl:Opg ratio when compared to WT but were not statistically significant when compared to Gnas E1+/−m mice. Sample size per genotype per experiment is listed on each bar graph. All statistical tests completed using ANOVA with post‐hoc Tukey test for multiple comparisons, and p values are displayed for each comparison.

Fig 7

Gnas E1+/−m osteoclasts display impaired calcitonin receptor bioactivity in vitro. (A) Gnas E1+/−m and Gnas E1+/−p osteoclasts displayed significant elevations in Calcr mRNA expression when compared to WT; however, Calcr expression within Gnas E1+/−m cultures was significantly greater than Gnas E1+/−p cultures. Treatment of Gnas E1+/− and WT osteoclasts with 1 × 10−7M sCT or 1 × 10−5M forskolin resulted in significant reductions in Calcr expression. (B,C) RT‐PCR analysis of Crem and Ramp3 mRNA expression of Gnas E1+/−m, Gnas E1+/−p, and WT cultures following exposure to sCT or vehicle controls for 6 hours. sCT‐treated Gnas E1+/−m cultures displayed a significant reduction in Crem and Ramp3 mRNA expression compared to Gnas E1+/−p and WT sCT‐treated cultures. No significant variations in Crem or Ramp3 expression were observed between Gnas E1+/−p and WT sCT treated cultures. (D) Gnas E1+/−m and Gnas E1+/−p BMSCs following 7 and 14 days of osteogenic differentiation demonstrate a significant reduction in Calcr mRNA expression when compare to WT. (E,F) RT‐PCR analysis of c‐Fos and Ramp3 mRNA expression of Gnas E1+/−m, Gnas E1+/−p, and WT BMSCs following exposure to sCT, PTH, forskolin, or vehicle controls for 6 hours. BMSCs overall displayed no significant response to sCT treatment. However, Gnas E1+/−m and Gnas E1+/−p BMSCs displayed a reduction in c‐Fos and Ramp3 when compared to WT BMSCs following PTH treatment. No significant variations in c‐Fos or Ramp3 expression were observed between PTH treated Gnas E1+/−m and Gnas E1+/−p BMSCs. Sample size per genotype per experiment is listed on each bar graph. All statistical tests completed using two‐way ANOVA with post‐hoc Tukey test for multiple comparisons, and p values are displayed for each comparison. sCT = salmon calcitonin.