Disruption of the FGFR1-FGF23-Phosphate Axis and Targeted Therapy in a Murine Model of Osteoglophonic Dysplasia

Abstract

Osteoglophonic Dysplasia (OGD) is an autosomal dominant skeletal dysplasia characterized by impaired bone growth resulting in short stature, severe craniofacial abnormalities, and in some patients FGF23-mediated hypophosphatemia. It is caused by gain-of-function variants in FGFR1, particularly in or near the transmembrane domain of the receptor. We used CRISPR in mice to knock-in the FGFR1 p.N330I variant, chosen based on its association with FGF23 excess. Skeletal phenotyping of this Fgfr1 ^+/N330I model demonstrated markedly reduced body weight and naso-anal length, shortened long bones, and craniosynostosis, all hallmarks of the human disease. Mutant mice exhibited profound microarchitectural changes in cortical bone and severe disorganization of the growth plate and articular cartilage, driven by decreased cell proliferation and increased apoptosis in skeletal tissues. In addition to osteochondrodysplasia, we noted dramatic increases in plasma FGF23 and hypophosphatemia, driven by upregulated Fgf23 expression and protein levels in bone, with consequent undermineralization. An in vivo ossicle assay allowed longitudinal evaluation of mineral metabolism. We modulated the signaling pathway by repurposing an inhibitor of the overactive receptor, infigratinib, resulting in partial restoration of naso-anal length in treated mutant mice. This first model of OGD offers insights into the disease pathogenesis and open avenues for targeted therapeutic strategies.

Figure 1.. Fgfr1^+/N330I mice show severe growth impairment.

(A) Design of the knock-in Fgfr1^+/N330I mice. Created with BioRender.com. (B) Representative images of whole-body radiography of 5-week-old mice (littermate control on the left; mutant mouse on the right). (C) 5-week-old Fgfr1^+/N330I mice are markedly smaller than their WT littermate controls as determined by the reduction of total weight and whole-body length. (D) 5-week-old Fgfr1^+/N330I mice have reduced naso-anal and tail length as compared to WT controls. (E) Representative microCT scan reconstructed at the same scale, showing severely shortened femur of 3-week-old mutant mouse (right) with lack of typical landmarks (femoral head and neck, greater trochanter, lesser trochanter, third trochanter) as compared to a femur from a littermate control (left). Mutant mice show rhizomelia, or shortening of the proximal appendicular long bones. (F) The lengths of both femur and tibia are dramatically reduced in the mutant mice. Horizontal and vertical lines represent the mean ± SD of 3 mice/group (**** p < 0.0001).

Figure 2.. Fgfr1^+/N330I mice show profound alterations of femoral microarchitecture and bone mineral density.

(A) Fgfr1^+/N330I mice show a significant reduction of bone volume fraction (BV/TV), cortical bone area fraction (Ct.Ar/Tt.Ar), cortex volume, and cortical thickness (Ct.Th) as compared to WT littermate controls (17%, 23%, 54% and 41%, respectively). (B) Fgfr1^+/N330I mice show an impaired cortical structure of femurs characterized by an increased cortical porosity (4% vs. 0.6% in WT controls). (C) Fgfr1^+/N330I mice display a significant reduction of femur bone mineral density (BMD). Horizontal and vertical lines represent the mean ± SD of 3 WT or 4 mutant mice. (D) Whole-body bone mineral density (BMD) and bone mineral content (BMC) are significantly reduced in mutant mice. Horizontal and vertical lines represent the mean ± SD of 3 or 4 mice/group (** p < 0.01, *** p < 0.001; **** p < 0.0001).

Figure 3.. Fgfr1^+/N330I mice show craniofacial involvement.

(A) Fgfr1^+/N330I mice show profound differences in skull morphology as shown by a reduction in total skull and snout length (35% and 72%) and reduction of the eye socket length and width (33% and 54%). Horizontal and vertical lines represent the mean ± SD of 3 WT or 4 mutant mice. (B) Representative images of unstained dissected calvaria from WT mice (top) and mutant (bottom) mice (scale bar = 1 mm) and (C) representative images from whole-mount skeletal stains show well-defined sutures in WT mice (top) whereas the mutant mice (bottom) show complete fusion of coronal and partial fusion of frontal sutures; magnified view of sutures in WT and mutant mice are shown on the right panel. fs- frontal suture, cs- coronal suture, ss- sagittal suture. (D) Representative 3D rendering from microCT scans of the skull (left), sagittal view (middle), and zoomed-in sagittal view (right), showing an open coronal suture in the wild-type (top panel) and a fused suture in the mutant (bottom panel). Resolution 10 µm, scale bar = 5 mm. (E) Fgfr1^+/N330I mice have reduced skull BMD (20% reduction), (F) a smaller skull as shown by a marked reduction of its total volume, and (G) reduced volume of the bone with normal density (NBD, 78% reduction), but not of lower density bone (LDB), as compared to WT controls. Horizontal and vertical lines represent the mean ± SD of 3 WT or 4 mutant mice. (H) 3D reconstructions of WT and Fgfr1^+/N330I skulls highlighting differences in bone density (NDB in white vs. LDB in orange). (ns = not significant, ** = p < 0.01, **** = p < 0.0001).

Figure 4.. Fgfr1^+/N330I mice display profound cartilage disorganization in femurs.

(A) Qualitative histological analysis of Hematoxylin & Eosin (H&E) staining in 5-week- old Fgfr1^+/N330I mice showing severe abnormalities of both growth plates (black arrows; scale bar = 500 µm) and (B) epiphyseal centers of ossification (scale bar = 100 µm). (C) Decreased Safranin O/Fast green (SafO/FG) staining for proteoglycans in mutant mice revealed altered composition of the cartilage extracellular matrix (red, black arrows). Scale bar = 500 µm (D) Representative images of cell proliferation assay in proximal tibial articular cartilage (AC) and growth plate cartilage (GPC) of mice representing nuclear staining with Hoechst 33342 (blue signal) and with EdU (green signal) revealed decreased proliferation in articular cartilage (AC) and secondary ossification center (SOC) in Fgfr1^+/N330I mice when compared to WT mice (n = 2, 3-week-old). Scale bar = 100 µm (AC, SOC); scale bar = 200 µm (GPC). (E) Dot plots represent the percentage of EdU+ (left) and percentage of EdU+ cells normalized to tissue area in GPC (right), revealing significantly lower proliferation in Fgfr1^+/N330I mice when compared to WT mice. Data points represent mean ± SD of three non-consecutive sections from two WT and two mutant mice. (F) Increased apoptosis demonstrated in femurs from mutant mice through quantification of TUNEL+ cells and relative H-score. Data points represent mean ± SD of 3 mice/group. (* p < 0.05, ** p < 0.01).

Figure 5.. In vitro and in vivo evaluation of the effect of FGFR signaling pathway modulation on the phenotype of OGD.

(A) Bone marrow stromal cells (BMSCs) from mutant mice were cultured in triplicate and a dose-response curve was obtained; the IC50, or the concentration of a substance (infigratinib) needed to inhibit a biological process (the phosphorylation of ERK) by 50%, was calculated at 1.26 nM. (B) BMSCs were isolated from Fgfr1^N330I/+ mice and cultured to ~80% confluence; cells were then incubated with 125 ng/mL of infigratinib for 1 hour at 37°C. A 76% decrease in downstream phosphorylation compared to vehicle control was noted. Horizontal and vertical lines represent the mean ± SD of three biological replicates. * p < 0.05. (C) Graphical representation of the experimental design. The preclinical study involved genotyping on day of birth followed by subcutaneous injection of infigratinib 2 mg/kg/d of SQ infigratinib x 15 days to assess its effect in vivo. (D) Growth curve during infigratinib treatment trial. The naso-anal length of treated mutant animals (green inverted triangles) was intermediate between that of untreated mutant mice (empty circles) and WT mice (black circles and blue squares), with a significant difference in length between the treated mutant and the untreated (vehicle-treated) mutant mice (**** p < 0.0001 using a mixed model for repeated measures with time x treatment interaction effect). n = 6–17 mice/group.

Figure 6.. Effect of the Fgfr1 N330I variant on FGF23 synthesis.

(A) Fgfr1^+/N330I mice show markedly elevated intact FGF23 levels (9211 ± 2625 pg/mL; n = 5), when compared to WT littermates (522.1 ± 50.2 pg/mL; n = 14; p<0.0001) and consequently reduced circulating phosphate levels (10.1 vs. 13.6 mg/dL, p = 0.0115; n = 19 WT and 6 mutant mice). Horizontal and vertical lines represent the mean ± SD. (B) Transcriptomics in bone (n = 5) revealed enrichment of the FGFR binding pathway (FDR adjusted p-value = 0.036), with Fgf23 having the highest rank metric score for the genes driving enrichment within this pathway. (C) Representative images of mouse femur sections processed for RNAscope in situ hybridization. Fgf23 mRNA probe shows red staining while nuclear staining is blue. The top panels demonstrate femur at 20× magnification (scale bar = 200 µm), while the middle panels represent zoomed-in view of cortical bone (scale bar = 50 µm) demonstrating increased Fgf23 staining in mutant femurs. Bottom panels represent zoomed in view of the marrow region (scale bar = 50 µm) showing Fgf23 staining in peri-sinusoidal cells. Dot plot represents the density of positively stained nuclei with Fgf23 probe with respect to the area of cortical bone. The mean number of cells staining for Fgf23 in the mutant bone was 161, significantly higher in comparison to that of the WT mice (22, p = 0.0027). Horizontal and vertical lines indicate the mean ± SD from 2 WT and 2 mutant mice. (D) FGF23 protein levels were also increased in femurs of mutant mice as shown by immunohistochemical staining. Original magnification = 40×; scale bar = 100 µm. (E) Quantification of FGF23 abundance in the femurs of WT and mutant mice, highlighting not only a higher percentage of FGF23-positive cells but also stronger intensity indicating increased per-cell production (H-score) in the mutant mice as compared to WT controls (3 mice/group). Horizontal and vertical lines represent the mean ± SD of 3 mice/group.

Figure 7.. Fgfr1^+/N330I mice show impaired mineral metabolism.

(A) Representative images of undecalcified mouse femur sections stained with Goldner trichrome stain showing the impaired development and organization of the articular cartilage, secondary ossification center and growth plate (top panel) and revealing the increased osteoid (mineralized bone in green, and osteoid seam in red marked with yellow arrows; bottom panel) in Fgfr1^+/N330I mice. (B) Dot plots representing static histomorphometry parameters that illustrate a trend towards increased osteoid surface fraction (40% vs. 16%, p = 0.054), and a significant increase in osteoid volume fraction (26% vs. 6%, p = 0.012) and osteoid width (5.1 mm vs. 2.8 mm, p = 0.046) in femurs from Fgfr1^+/N330I mice. (C) Representative images of von Kossa-stained sections from the distal end of undecalcified femurs from WT and mutant mice showing mineralized tissue as black (left panels, scale bar = 500 µm). Brightfield images from anatomically equivalent regions of distal femurs were used to quantify the area of mineralization using ImageJ. Zoomed-in view of distal growth plate (right panels, scale bar = 100 µm). (D) Dot plot representing the percentage (%) of mineralized area. Data are presented as mean ± SD; p = 0.019. Horizontal and vertical lines represent the mean ± SD of 5 WT and 2 mutant mice. Original magnification 20×. OS- osteoid surface, BS- bone surface, OV- osteoid volume, BV- bone volume, O.Wi- osteoid width. (ns p ≥ 0.05, * p < 0.05, ** p < 0.001, **** p < 0.0001).

Figure 8.. An in vivo ectopic ossicle model reveals osteodysplasia and allows longitudinal assessment of the disrupted FGF23-phosphate axis.

(A) Schematic representation of the in vivo transplantation assay. Gelatin sponges were loaded with 2 million BMSCs (WT vs. Fgfr1^+/N330I) depleted of hematopoietic cells and transplanted into subcutaneous pockets in immunocompromised mice for 80 days. Created with BioRender.com. (B) Representative 3D microCT reconstruction of in vivo transplants. Scale bar = 500 µm. (C) Low and high magnification H&E staining images of representative transplants loaded with WT BMSCs (top panel) and Fgfr1^+/N330I BMSCs (bottom panels). The low magnification image for the Fgfr1^+/N330I BMSC transplant was obtained by merging two consecutive section images from the scanned slide. Histology confirmed excess bone but lack of cortex in the Fgfr1^+/N330I BMSC transplants. (D) Plasma intact FGF23 concentrations remained stable in mice receiving WT BMSC transplants, while those receiving Fgfr1^+/N330I BMSC transplants showed elevation of intact FGF23 levels starting at 40 days post-transplantation (top). Similarly, plasma phosphate concentrations were decreased in mice receiving Fgfr1^+/N330I BMSC transplants starting at 40 days post-transplantation (bottom). ** p < 0.01. (E) Decreased bone volume fraction as measured via microCT (BV/TV 4.64 vs. 7.99%; p = 0.03) and decreased whole-body BMD as measured via DXA (0.0640 vs. 0.785 g/cm²; * p = 0.04) in mice receiving Fgfr1^+/N330I BMSC transplants compared to those receiving WT BMSC transplants. (* p < 0.05).