Evaluation of the therapeutic potential of a CNP analog in a Fgfr3 mouse model recapitulating achondroplasia

Abstract

Achondroplasia (ACH), the most common form of dwarfism, is an inherited autosomal-dominant chondrodysplasia caused by a gain-of-function mutation in fibroblast-growth-factor-receptor 3 (FGFR3). C-type natriuretic peptide (CNP) antagonizes FGFR3 downstream signaling by inhibiting the pathway of mitogen-activated protein kinase (MAPK). Here, we report the pharmacological activity of a 39 amino acid CNP analog (BMN 111) with an extended plasma half-life due to its resistance to neutral-endopeptidase (NEP) digestion. In ACH human growth-plate chondrocytes, we demonstrated a decrease in the phosphorylation of extracellular-signal-regulated kinases 1 and 2, confirming that this CNP analog inhibits fibroblast-growth-factor-mediated MAPK activation. Concomitantly, we analyzed the phenotype of Fgfr3(Y367C/+) mice and showed the presence of ACH-related clinical features in this mouse model. We found that in Fgfr3(Y367C/+) mice, treatment with this CNP analog led to a significant recovery of bone growth. We observed an increase in the axial and appendicular skeleton lengths, and improvements in dwarfism-related clinical features included flattening of the skull, reduced crossbite, straightening of the tibias and femurs, and correction of the growth-plate defect. Thus, our results provide the proof of concept that BMN 111, a NEP-resistant CNP analog, might benefit individuals with ACH and hypochondroplasia.

Figure 1

ACH-Related Clinical Features in Fgfr3^Y367C/+ Mice (A) Fgfr3^Y367C/+ mice exhibit a severe dwarfism phenotype. Mice presented are littermates and are 17 days old. (B) Computed-tomography (CT) scans of Fgfr3^Y367C/+ (n = 5) and Fgfr3^+/+ mice (n = 5) (top) and ACH and control human skulls (15 years of age) (bottom) show a domed skull (arrow) with an anterior crossbite (arrow) in Fgfr3 mutants. CT scans were performed with a VivaCT40 microscanner (SCANCO Medical, Laboratoire de Biologie Intégrative du Tissu Osseux, Institut National de la Santé et de la Recherche Médicale Unité U1059) according to the following protocols: 0.020 mm voxel resolution, 55 kV, sigma 1.5, support 2, threshold 148, and 150 uA. (C) MRI analysis of Fgfr3^Y367C/+ (n = 4) and Fgfr3^+/+ mice (n = 4) (3 weeks of age) (top) and ACH and control humans (16 years of age) (bottom) shows medullary and upper-spinal-cord compression (arrow) in a pathological case. After three averages, a repetition time of 1,200 ms (echo time 80 ms), and two dummy scans, the resolution in mice was 0.0078 cm/px × 0.0078 cm/px × 625 micron/px for a matrix size of 256 × 256 × 32. The total scan time was 30 min. In humans, MRI was performed with a 1.5 tesla (Signa General Electric Medical Systems, Milwaukee, WI, USA) scanner with the following sequences: FSE T2 (TR/TE: 5,000/108, 3 mm slices, 0.3 mm gap). MRI was performed without intravenous contrast enhancement. Osirix software was used for postprocessing. (D) Hematoxylin and eosin (H&E) staining of the growth plate of 2-week-old Fgfr3^Y367C/+ mice (n = 20) shows a decrease in the height of the hypertrophic (HZ) zone. The area of the Fgfr3^Y367C/+ hypertrophic cells is smaller than that of Fgfr3^+/+ cells (arrows). H&E analysis of the growth plate from a TD (c.1118A>G [p.Tyr373Cys] mutation) fetus (16 weeks of age), from an ACH (c.1138G>A [p.Gly380Arg] mutation) fetus, and from a control fetus (25 weeks of age) shows a gradient of extremely severe (TD) to moderate (ACH) (arrows) defects in the growth-plate architecture. The size of the hypertrophic cells from the ACH growth plate is smaller than that of the control cells. Comparative analyses of human and mouse growth plates shows that the cartilage of Fgfr3^Y367C/+ mice (n = 30) resembles that of ACH mice (n = 15) more than that of TD mice (n = 25). The scale bars represent 50 μm.

Figure 2

BMN 111 Reduced ERK1/2 Activation in ACH Growth-Plate Chondrocytes (A) Control and pathologic human chondrocytes (ACH) were isolated as described previously and treated with BMN 111 (10⁻⁵ M) prior to coincubation with fibroblast growth factor (FGF) (100 ng/ml FGF18). BMN 111 pretreatment partially prevented FGF-mediated increase in ERK1/2 phosphorylation (n = 6). The cell lysates were subjected to SDS polyacrylamide-gel electrophoresis and were hybridized overnight at 4°C with phosphorylated-ERK1/2 antibody (1/1,000) (Cell Signaling Technology, Danvers, MA, USA) or ERK1/2 antibody (1/1,000) (Cell Signaling Technology). A secondary antibody coupled to peroxidase was used at a dilution of 1:10,000 (GE Healthcare Life Sciences, Pittsburgh, PA, USA). Bound proteins were detected by enhanced chemiluminescence (GE Healthcare Life Sciences). (B) Immunoblot showing STAT3-P and STAT3 (dilution 1:1,000, Cell Signaling Technology) in human ACH growth-plate chondrocytes (n = 3). The phosphorylation of STAT3 did not appear decreased upon BMN 111 coincubation.

Figure 3

CNP Analog Improved the Size and Growth-Plate Defect in an Ex Vivo Fgfr3^Y367C/+ Mouse Model (A) FGFR3 and NPR-B localization in the proliferative (PZ) and prehypertrophic (PHZ) zones of the cartilage. Serial histology sections of the growth plate from Fgfr3^+/+ and Fgfr3^Y367C/+ mice (n = 6) were stained by immunochemistry with an FGFR3 antibody (1:250 dilution; Sigma-Aldrich, St. Louis, MO, USA) and an NPR-B antibody (1:100 dilution; Abcam [ab14357-ICC/IF/WB], Cambridge, UK) and were visualized with the Dako Envision kit with an anti-rabbit antibody (secondary antibody) (Dako North America, Carpinteria, CA, USA). Images were captured with an Olympus PD70-IX2-UCB microscope (Olympus, Tokyo, Japan). (B) A partial normalization of the growth defect, as well as enlargement of the epiphysis, was observed in the BMN-111-coincubated femurs (n = 25) isolated from Fgfr3^Y367C/+ mouse embryos (E16.5) after 6 days of culture in Dulbecco's modified Eagle's medium with antibiotics and 0.2% bovine serum albumin (Sigma-Aldrich) supplemented with BMN 111 (BMN 111 patent US2010-0297021, BioMarin Pharmaceutical, Novato, CA, USA). (C) Quantitative gain in femur length after coincubation with BMN 111 (10⁻⁶ M to 10⁻¹⁰ M) or vehicle. A partial concentration-dependent restoration of the growth-plate defect was observed in Fgfr3^Y367C/+ femurs coincubated with BMN 111 (BMN 111 10⁻⁶ M, n = 13; 10⁻⁷ M, n = 10; 10⁻⁸ M, n = 8; 10⁻⁹ M, n = 6; and 10⁻¹⁰ M, n = 6). ^∗p < 0.05, ^∗∗p < 0.001, Student's t test. Error bars indicate the SD. (D) H&E staining of longitudinal sections of Fgfr3^Y367C/+ (n = 11) and Fgfr3^+/+ (n = 16) femurs coincubated with BMN 111 (10⁻⁶ M) or vehicle. An increase in the height of the proliferative and hypertrophic zones of the growth plate, along with the normalization of the shape of the chondrocytes, was observed in BMN-111-treated Fgfr3^Y367C/+ femurs. The enlarged regions are boxed. The scale bar in the “Proximal femur” row represents 200 μm, the scale bar in the “Growth plate” row represents 100 μm, and the scale bar in the “Hypertrophic zone” row represents 20 μm. (E) Type-X-collagen (1:30 dilution, Quartett, Berlin, Germany) immunohistochemical staining of the hypertrophic zone of Fgfr3^Y367C/+ femurs coincubated with BMN 111 (10⁻⁶ M) or vehicle. A partial restoration of the size of the hypertrophic zone was observed in BMN-111-treated Fgfr3^Y367C/+ femurs (n = 11) compared to vehicle-treated Fgfr3^Y367C/+ femurs coincubated with vehicle (n = 11). The scale bar represents 100 μm.

Figure 4

CNP Analog Partially Corrected the Dwarfism of Fgfr3^Y367C/+ Mice (A) Phenotypic changes were observed in 800 μg/kg BMN-111-treated Fgfr3^Y367C/+ mice (n = 9) compared to vehicle-treated Fgfr3^Y367C/+ mice (n = 9) and included flattening of the skull and an increased length of the long bones and tail. Fgfr3^Y367C/+ mice received once-daily SC administrations of BMN 111 or vehicle (0.03 mol/l acetic acid buffer solution, pH 4.0, containing 1% [w/v] benzyl alcohol and 10% [w/v] sucrose) for 10 days. All mice were dosed approximately 2 hr prior to the dark cycle. (B) Histological safranin-O (top) or H&E staining of growth plates from Fgfr3^Y367C/+ mice treated or not with 240 μg/kg or 800 μg/kg BMN 111 during 10 days. Dose-dependent modifications of the epiphysis were observed in distal-femur sections of Fgfr3^Y367C/+ mice. The formation of the secondary ossification center was delayed in vehicle-treated Fgfr3^Y367C/+ mice compared to vehicle-treated Fgfr3^+/+mice (arrow). CNP analog treatment (800 μg/kg) did not appear to accelerate the ossification process. The scale bars in the “Cartilage” and “Distal epiphysis” rows represent 200 μm, the scale bars in the “Growth plate” row represent 50 μm, and the scale bars in the “Hypertrophic zone” row represent 20 μm. (C) Pictures of Fgfr3^+/+ (n = 5) and Fgfr3^Y367C/+ (n = 5) mice treated during 20 days with 800 μg/kg BMN 111. Phenotypic changes comprise evident lengthening and straightening of the limbs, as well as longer forelimbs and hind limbs, including paws and digits, in BMN-111-treated Fgfr3^Y367C/+mice (n = 5). (D) X-rays of the skeleton show flattening of the skull (arrow) and improvement in the snout (arrow) in BMN-111-treated Fgfr3^Y367C/+ mice. X-rays of the hind limbs show straightening of both the tibia and femur in BMN-111-treated Fgfr3^Y367C/+ mice (arrows) and a marked increased in the size of the digits (arrows). (E) Histological analysis of the distal femur of Fgfr3^Y367C/+ mice with or without BMN 111 treatment after 20 days. Rescue of the height and architecture of the different zones of the growth plate, along with replicating and hypertrophic chondrocytes organized in columns, was observed in BMN-111-treated Fgfr3^Y367C/+mice. The shape and size of the proliferative zone, visualized by type-II-collagen (Col II) in-situ-hybridization labeling (type-II-collagen riboprobe²¹) and KI67 immunolabeling (KI67 antibody, 1:3,000 dilution, Abcam), were modified after 800 μg/kg BMN 111 treatment in Fgfr3^Y367C/+ mice (arrows). The scale bar in the “Distal epiphysis” row represents 200 μm, and the scale bars in the “Growth plate,” “Col II,” and “KI67” rows represent 50 μm.