Fig 1.
Gorab conditional inactivation in mesenchymal progenitors and pre-osteoblasts, but not at later stages leads to severe cortical bone thinning.
(A) Gorab expression in chondrocytes, perichondrium (red arrowhead), and periosteum (black arrowhead) as indicated by LacZ staining of E18.5 GorabNull tibia. Scale bar = 50μm. (B) Sagittal view and (C) transverse view of reconstructed μCT image of tibia from twelve week old GorabPrx1, GorabRunx2, GorabDmp1 and control animals. (D) Quantitative μCT analysis given as fold changes for trabecular bone volume (BV/TV), length and cortical thickness (Cort.Th), of tibia of twelve week old GorabPrx1 (N = 15), GorabRunx2 (N = 5), GorabDmp1 (N = 8) and littermate control animals, N = 14, N = 5 and N = 9 for each individual conditional mouse line, respectively. Values were normalized to the control littermates of each mutant line to facilitate comparison.
Fig 2.
Cortical porosity and osteoblast dysfunction in the GorabPrx1 model recapitulate the gerodermia osteodysplastica bone phenotype.
(A) Von Kossa / hematoxylin stained cortical bone of twelve week old GorabPrx1 mouse. Inset showing magnified view of large pores in the mutant diaphysis. Scale bar = 200 μm. (B) Sections of metaphyseal cortical bone of the tibia of four week old GorabPrx1 mice stained with Masson Goldner trichrome. Scale bar = 100μm. (C) μCT reconstructed image of a spontaneously fractured humerus from a four week old GorabPrx1 mouse. Arrowhead = fracture site. Scale bar = 1mm. (D) Histomorphometric quantitation number of osteoblast per bone perimeter (N.Ob/B.Pm), (E) number of osteocytes per cortical bone perimeter (N.Ot/cort B.Pm). (F) Mineral apposition rate (MAR) at the endosteum of tibia midshaft in four week old GorabPrx1 (N = 4). Inset showing calcein double labeling, C = control, M = mutant. (G) osteoid volume (OV/TV) in secondary spongiosa of the proximal tibia in GorabPrx1 mutants vs. controls at four weeks of age (N = 4–6). (H) Altered expression of osteoblast lineage marker genes in femoral cortical bone of four week old GorabPrx1 mutants (N = 6–8). (I) Immunohistochemical detection of Sp7/osterix, Spp1/osteopontin and Dmp1 expression in cortical bone. Note higher number of osterix + cells in GorabPrx1 mutants. Scale bar = 50μm. (J) Number of osteoclasts per bone perimeter (N.Oc/B.Pm). (K) Opg to Rankl expression ratio in four week old GorabPrx1 mutants (N = 6). P = periosteum, C = cortical bone, E = endosteum.
Fig 3.
Disrupted collagen network in skin and bone tissue deficient for Gorab.
(A) Electron microscopy of skin from E18.5 GorabNull mouse (magnification = 11700x) and (B) bone from tibia of eight week old GorabPrx1 mouse. Asterisk marks collagen fibers, Ot = osteocyte. (C) Collagen fiber orientation in cortical bone of four week old GorabPrx1 mutants and in a GO patient bone biopsy stained with picrosirius red or toluidine blue, respectively, and imaged under polarized light. Note fragmentation of birefringence signals. For each experiment at least N = 3 mice were analyzed, representative images are shown.
Fig 4.
Loss of Gorab resulted in underglycanation of proteoglycans.
(A) Quantitation of dermatan sulfate and chondroitin sulfate in skin, cartilage, and lung samples from E18.5 GorabNull mice (N = 3–4). (B) Total amount of GAGs in the cortical bone of femora from four week old GorabPrx1 mutants and littermate controls (N = 3–4). (C) Percentage of dermatan sulfate in the total amount of glycosaminoglycans (GAGs). (D) Immunoblotting for decorin in skin samples from E18.5 GorabNull mice. Loss of the 100kDa band, corresponding to the fully glycanated decorin, and higher intensity the core protein band in mutant lysate indicate a glycanation defect. (E) Immunoblotting for decorin in cortical bone lysates from four week old GorabPrx1 mice and littermate controls also showing higher intensity of lower bands in mutant. (F) Immunoblot of decorin in cortical bone lysates from wildtype (WT) mice at different ages: newborn (P0), 5 weeks (5W), 26 weeks (26W) and 2 years (2Y). Note reduction in glycanation with increasing age. (G) Immunofluorescence detection of decorin and (H) biglycan in tibia of four week old GorabPrx1 mice. Sections were not pretreated with chondroitinase. Higher staining intensities therefore indicate lower glycanation of the core proteins in the periosteum. P = periosteum, C = cortical bone. Scale bar = 50μm. Experiments (D) to (H) were repeated at least three times with independent biological samples, representative results are shown.
Fig 5.
Golgi retention and reduced glycanation of decorin in Gorab-deficient fibroblasts.
(A) Western blot analysis of decorin in control and GorabNull MEF cell lysates. The size ranges for the fully glycanated form of decorin (DCN) and of its core protein are indicated. (B) Left, western blot analysis of decorin in lysates of the extracellular matrix produced by cultured control and GO fibroblasts. Right, levels of glycanated decorin were quantified against total decorin levels (N = 3). (C) Analysis of intra-Golgi levels of decorin in co-cultured control and GO human skin fibroblasts. Immunofluorescence labeling was performed with anti-decorin, anti-GORAB and anti-GM130 antibodies. GORAB staining was used to distinguish control (yellow arrows) and GO (red arrows) cells. Decorin fluorescence intensity in both cell types was normalized against that of the Golgi marker GM130 (N = 3, >500 cells analyzed per cell line). Scale bar = 10 μm. (D) Analysis of intra-Golgi levels of dermatan sulfate (DS)-modified proteins in co-cultured control (yellow arrows) and GO (red arrows) human skin fibroblasts. Cells were labeled with anti-DS (GD3A12), anti-GORAB and anti-TGN46 antibodies. The intensity of the GD3A12 fluorescence signals were measured relative to that of the Golgi marker TGN46 (N = 3, >500 cells analyzed per cell line). Scale bar = 10 μm.
Fig 6.
Elevated TGF-β signaling in Gorab-deficient skin, bone, and fibroblasts.
(A) Quantitation of active and total TGF-β in skin lysates from E18.5 GorabNull mice (N = 3). (B) Enhanced expression of TGF-β regulated genes in the diaphysis of four week old GorabPrx1 mutants (N = 7–8). (C) Immunofluorescence staining for p-SMAD2 in GorabPrx1 mutants and controls at four weeks of age. Representative picture of N = 4 per group. Note stronger signals in periosteum in mutants. Scale bar = 50μm. (D) Western blot of p-SMAD2 in confluent fibroblasts from GO patients and healthy controls (N = 3) and quantitative evaluation. (E) Quantitative PCR to measure expression of TGF-β responsive genes in GO patient-derived fibroblasts (N = 4). P = periosteum, C = cortical bone. E = endosteum.