The collagen receptor, discoidin domain receptor 2, functions in Gli1-positive skeletal progenitors and chondrocytes to control bone development

Abstract

Discoidin Domain Receptor 2 (DDR2) is a collagen-activated receptor kinase that, together with integrins, is required for cells to respond to the extracellular matrix. Ddr2 loss-of-function mutations in humans and mice cause severe defects in skeletal growth and development. However, the cellular functions of Ddr2 in bone are not understood. Expression and lineage analysis showed selective expression of Ddr2 at early stages of bone formation in the resting zone and proliferating chondrocytes and periosteum. Consistent with these findings, Ddr2⁺ cells could differentiate into hypertrophic chondrocytes, osteoblasts, and osteocytes and showed a high degree of colocalization with the skeletal progenitor marker, Gli1. A conditional deletion approach showed a requirement for Ddr2 in Gli1-positive skeletal progenitors and chondrocytes but not mature osteoblasts. Furthermore, Ddr2 knockout in limb bud chondroprogenitors or purified marrow-derived skeletal progenitors inhibited chondrogenic or osteogenic differentiation, respectively. This work establishes a cell-autonomous function for Ddr2 in skeletal progenitors and cartilage and emphasizes the critical role of this collagen receptor in bone development.

Fig. 1

Ddr2 is highly expressed in long bone growth plates, periosteum and bone marrow. a Left, whole-mount X-gal staining of Ddr2^Lacz/+ mouse embryos from E9.5-E16.5, showing of Ddr2 expression starting at E11.5; right, sagittal section (right of midline) of newborn showing forelimb (fl), ribs, and hindlimb (hl) staining. Scale bar: 2 mm. b X-gal staining of cryostat section of spine (intervertebral discs) from newborn mice. Scale bar: 100 μm. c Side views of Ddr2^+/+ (left) and Ddr2^Lacz/+ (right) embryos at E13.5. Midline sagittal section (right) of Ddr2^Lacz/+ embryo is indicated in dotted lines. Arrowhead points to cranial base cartilage template. M, Meckles cartilage; V, cartilaginous primordia of the vertebral bodies; MP, maxillary process. Scale bar: 2 mm. d Side views of rib cages from Ddr2^Lacz/+ mice at E18.5 and birth. Scale bar: 2 mm. e Whole-mount X-gal staining of Ddr2^Lacz/+ embryo forelimbs (upper panel) and hindlimbs (lower panel) at E11.5, E13.5, and E18.5. Scale bar: 500 μm. f, g X-gal staining of cryostat sections of long bone from newborn and 3 month-old mice. R Resting zone; P Proliferative zone; H Hypertrophic zone; gp Growth plate. Arrowheads indicate areas of positive X-gal staining. Note: Newborn growth plate in (f) is a composite of three images. Scale bar: 200 μm (left) and 50 μm (right)

Fig. 2

Skeletal contribution of Ddr2-expressing cells. a Protocol. b–d, Lineage tracing in Ddr2^mer-iCre-mer; R26R^tdTomato mice at the postnatal day 5 (P5), P14 and 2 months (n = 2 mice). Red: Ddr2-positive cells; gray: cell nuclei; gp Growth plate; bm Bone marrow; SOC Secondary ossification center; ms Muscle; R Resting zone; P Proliferative zone; H Hypertrophic zone. b Fluorescent tdTomato (red) distribution in the proximal tibia at P5 showing initial labeling in perichondrium, periosteum, primary spongiosa, and sporadic cells in the growth plate. Scale bar: 200 μm. The boxed region is shown in higher magnification (right). Arrowheads indicate expression in perichondrium. Scale bar: 100 μm. c Fluorescent tdTomato distribution at P14 showing label in the resting and proliferative zones of growth plate and in primary spongiosa, but not in the hypertrophic zone. Scale bar: 200 μm. Boxed region is shown in higher magnification (right) Scale bar: 100 μm. Dotted Line denotes chondro-osseous junction. d Fluorescent tdTomato distribution at 2 months showing persistence of tdTomato labeling and contribution to (1) growth plate columns, (2) bone cells in primary spongiosa, and (3) periosteum and osteocytes in cortical bone (arrowheads). Scale bar: 200 μm. Images at a higher magnification for different areas of tibia are shown on the right side. Scale bar: 20 μm. e High power image of 2-month growth plate and trabecular regions showing co-staining of tdTomato⁺ cells with an anti-OSX antibody (green) in osteoblasts on the trabecular surface. Scale bar: 100 μm

Fig. 3

Ddr2-derived cells are distinct from osteoclasts; overlapping distribution of Ddr2 and GLI1. a Fluorescence-based TRAP staining on frozen sections from Ddr2^mer-iCre-mer; R26R^tdTomato mice at P14 showing no overlap between TRAP⁺ osteoclasts and Ddr2 tdTomato-labeled cells or their progeny. Scale bar: 50 μm. Arrowheads indicate multinucleated osteoclasts (Cyan). Dotted Line denotes trabecular bone surface. The boxed region is shown in higher magnification (top). b Gli1-Cre^ERT; R26R^tdTomato mice were treated with tamoxifen as in Fig. 2a and examined at P14. Section of the proximal tibia showing Gli1-Cre^ERT fluorescent tdTomato fluorescence with intense labeling in the resting zone of the growth plate and primary spongiosa, in cell populations having partial overlap with Ddr2^mer-iCre-mer Ddr2-positive cells in Fig. 2. Scale bar: 100 μm. c IF colocalization of Ddr2 (red) and Gli1 (green). Proximal tibias were isolated from P5 C57BL6 mice. Extensive Gli1 and Ddr2 colocalization were seen in the resting zone (R) and proliferative (P) chondrocytes with low staining in hypertrophic cells (H). More limited staining was seen in select cells in primary spongiosa (PS) and metaphysis (M). Right panels, Quantitation of Ddr2⁺, Gli1⁺, and Gli1⁺ Ddr2⁺ cells in growth plate and PS/MS. The percentage of Gli1⁺ cells also containing Ddr2 is shown in brackets. Values are means ± SD with n = 3 mice

Fig. 4

Requirement for Ddr2 in Gli1-positive chondro-osteoprogenitors. a Experimental protocol for embryonic study (b). b Safranin O and von Kossa staining of tibia from Ddr2^fl/fl (n = 3) and Gli1-Cre^ERT;Ddr2^fl/fl (n = 4) newborns. von Kossa staining (lower panel) shows no evident difference in mineralization (black staining). Scale bar: 200 μm. Right; quantification of growth plate length. c Safranin O staining of proximal tibia growth plates from WT and Ddr2^Slie/Slie newborns. The dotted line indicates the growth plate lengths of mutant in relation to WT. Scale bar: 200 μm. Right; quantification of growth plate length (n = 3). d Safranin O staining and quantification (right) showing reduced growth plate lengths in mutant mice at 2 weeks of age (n = 3), experimental protocol at the bottom. Scale bar: 100 μm. e Experimental protocol for adult study. f Gross phenotype of 3 month-old Ddr2^fl/fl and Gli1-Cre^ERT;Ddr2^fl/fl littermates. Gli1-Cre^ERT;Ddr2^fl/fl mice have skeletal dwarfism (n = 20, 10 mice/sex). Scale bar, 2 cm. g Gene expression analysis of Ddr2 from whole 3 month-old tibiae to confirm knockout efficiency (n = 6 mice). h, i Body length (cm) and weight (g) measurements (n = 20, 10 mice/sex). j 3D micro-computed tomography (microCT) models of the tibial metaphysis and cortical mid-shaft in 3 month-old Ddr2^fl/fl and Gli1-Cre^ERT;Ddr2^fl/fl mice. Scale bar 100 µm. k–o Quantification of trabecular bone volume fraction (Trab. BV/TV), trabecular number (Tb.N), trabecular thickness (Tb.Th), trabecular spacing (Tb.Sp), and bone mineral density (Trab. BMD). (n = 20, 10 mice/sex). p, q Quantification of cortical bone volume fraction (Cort. BV/TV), and bone mineral density (Cort. BMD). Statistics: Data are expressed as mean ± SD. Unpaired t-test, *P < 0.05, **P < 0.01; ***P < 0.001; ****P < 0.000 1; ns, not significant

Fig. 5

Requirement for Ddr2 in Col2a1-positive chondroprogenitor cells. a Gross phenotype of 3 month-old Ddr2^fl/fl and Col2a1-Cre;Ddr2^fl/fl littermates. Col2a1-Cre; Ddr2^fl/fl mice have skeletal dwarfism (n = 20, 10 mice/sex). Scale bar: 2 cm. b Gene expression analysis of Ddr2 from whole 3 month-old tibiae to confirm knockout efficiency (n = 6 mice). c, d Measurements of body length (cm) and weight (g) (n = 20, 10 mice/sex). e 3D micro-computed tomography (microCT) models of the tibial metaphysis and cortical mid-shaft in 3 month-old Ddr2^fl/fl and Col2a1-Cre; Ddr2^fl/fl mice. Scale bar: 100 µm. f–j Quantitative data of trabecular bone volume fraction (Trab. BV/TV), trabecular number (Tb.N), trabecular thickness (Tb.Th), trabecular spacing (Tb.Sp) and bone mineral density (Trab. BMD), (n = 20, 10 mice/sex). k, l Quantification of cortical bone volume fraction (Cort. BV/TV), and bone mineral density (Cort. BMD), (n = 20, 10 mice/sex). m Gene expression analysis of osteogenic and chondrogenic differentiation markers. Statistics: Unpaired t-test *P < 0.05, **P < 0.01; ***P < 0.001; ****P < 0.000 1; ns Not significant. Data were presented as mean ± SD

Fig. 6

Ddr2 loss in chondroprogenitor cells results in deficient chondrocyte proliferation, organization, and abnormal type II collagen. a Hematoxylin and eosin (H&E), alcian blue and von Kossa staining of developing tibiae at E15.5 showing delays endochondral ossification in developing skeleton. Red arrowheads point to the delay in the formation of primary ossification center (n = 3). Scale bar: 200 μm and 100 μm in lower panel. b Safranin O staining of tibial growth plate sections of 2-week-old Col2a1-Cre; Ddr2^fl/fl shows chondrocyte disorganization. Boxed region is shown in higher magnification (bottom). Scale bar: 100 μm (top) and 20 μm (bottom). c Immunofluorescent staining of tibial growth plate sections of 2-week-old Col2a1-Cre; Ddr2^fl/fl and control littermates with an antibody against type II collagen (red) showing uneven distribution with no immunostaining in between chondrocytes. DAPI (blue) stains cell nuclei. Scale bar: 20 μm. d–g Hematoxylin and eosin (H&E) staining and quantification showing reduced growth plate lengths in Col2a1-Cre;Ddr2^fl/fl mice at 2 weeks. Scale bar: 200 μm. Unpaired t-test ***P < 0.001; ns Not significant. h EdU staining (green) of tibial growth plate sections show reduced chondrocyte proliferation in Col2a1-Cre; Ddr2^fl/fl mice. Scale bar: 100 μm. i Quantification of EdU-positive cells in proliferative zone (n = 4). j Fluorescent TUNEL staining (green) for detecting chondrocyte apoptosis at chondro-osseous junction (dotted lines). DAPI (blue) stains cell nuclei. Scale bar: 100 μm. k Quantification of chondrocyte apoptosis (n = 3), unpaired t-test, ns, not significant

Fig. 7

No obvious bone defects with osteoblast specific Ddr2 knockout. a Gross phenotype of 3 month-old Ddr2^fl/fl and Ocn-Cre;Ddr2^fl/fl littermates. Ocn-Cre;Ddr2^fl/fl mice have no skeletal phenotype (n = 20, 10 mice/sex). Scale bar, 2 cm. b Gene expression analysis of Ddr2 from whole 3 month-old tibiae to confirm knockout efficiency (n = 8 mice). c, d Measurements of body length (cm) and weight (g) (n = 20, 10 mice/sex), unpaired t-test. ns Not significant. e–j Quantitative data of trabecular bone volume fraction (Trab. BV/TV), trabecular number (Tb.N), trabecular thickness (Tb.Th), trabecular spacing (Tb.Sp) and bone mineral density (Trab. BMD), (n = 20, 10 mice/sex). h 3D micro-computed tomography (microCT) models of the tibial metaphysis and cortical mid-shaft in 3 month-old Ddr2^fl/fl and Ocn-Cre; Ddr2^fl/fl mice. Scale bar 100 µm. k, l Quantification of cortical bone volume fraction (Cort. BV/TV), and bone mineral density (Cort. BMD). (n = 20, 10 mice/sex), unpaired t-test, ns Not significant. Data were presented as mean ± SD

Fig. 8

Ddr2 deficiency inhibits chondrogenic differentiation in micromass cultures. a Chondroprogenitor cells isolated from limb buds of E12.5 Ddr2^fl/fl mice were treated with AdLacz or AdCre and differentiated for 8 days followed by alcian blue staining (top two panels) or were subsequently stained with eosin (bottom panel). Scale bar: 500 μm. b Total DNA. b, c–f Quantitative real-time RT-PCR (qRT-PCR) gene expression analyses of micromass cultures. b, c Efficient knockdown of Ddr2. c Gene expression analysis of chondrocyte differentiation markers. d–g Gene expression analysis shows downregulation of chondrocyte and hedgehog signaling in Ddr2-deficient chondrocytes. Unpaired t-test *P < 0.05, **P < 0.01; ***P < 0.001; ****P < 0.000 1; ns Not significant

Fig. 9

Requirement for Ddr2 in osteoblast differentiation of purified marrow-derived skeletal progenitors. a Flow cytometry analysis and isolation of CD140α/CD51⁺ cells from Ddr2^fl/fl mice. Boxed region (right) shows percentage of BM stem/progenitor cells. b Fold enrichment of Ddr2 and stem cell marker mRNAs in CD140α/CD51⁺ cells. c–e Colony forming unit-fibroblast (CFU-F) and CFU-osteoblasts (CFU-Ob) assays showing Ddr2 deficiency inhibits CFU-Ob. c Representative images of CFU-F and CFU-Ob. d CFU-F. e CFU-Ob. f–i Gene expression analysis of CD140α/CD51⁺ bone marrow cells treated with AdLacz or AdCre and grown in osteogenic medium. Closed bars, AdLacZ (control); open bars, AdCre. Unpaired t-test *P < 0.05, **P < 0.01; ***P < 0.001; ****P < 0.000 1; ns Not significant