Author response: Control of craniofacial development by the collagen receptor, discoidin domain receptor 2
Year: 2023
Type: peer-review
Abstract: Article Figures and data Abstract Editor's evaluation eLife digest Introduction Results Discussion Materials and methods Data availability References Decision letter Author response Article and author information Metrics Abstract Development of the craniofacial skeleton requires interactions between progenitor cells and the collagen-rich extracellular matrix (ECM). The mediators of these interactions are not well-defined. Mutations in the discoidin domain receptor 2 gene (DDR2), which encodes a non-integrin collagen receptor, are associated with human craniofacial abnormalities, such as midface hypoplasia and open fontanels. However, the exact role of this gene in craniofacial morphogenesis is not known. As will be shown, Ddr2-deficient mice exhibit defects in craniofacial bones including impaired calvarial growth and frontal suture formation, cranial base hypoplasia due to aberrant chondrogenesis and delayed ossification at growth plate synchondroses. These defects were associated with abnormal collagen fibril organization, chondrocyte proliferation and polarization. As established by localization and lineage-tracing studies, Ddr2 is expressed in progenitor cell-enriched craniofacial regions including sutures and synchondrosis resting zone cartilage, overlapping with GLI1 + cells, and contributing to chondrogenic and osteogenic lineages during skull growth. Tissue-specific knockouts further established the requirement for Ddr2 in GLI +skeletal progenitors and chondrocytes. These studies establish a cellular basis for regulation of craniofacial morphogenesis by this understudied collagen receptor and suggest that DDR2 is necessary for proper collagen organization, chondrocyte proliferation, and orientation. Editor's evaluation This fundamental work substantially advances our understanding of the development of the craniofacial skeleton requires interactions between progenitor cells and the collagen-rich extracellular matrix (ECM). The evidence supporting the conclusions is compelling, with rigorous biochemical assays, state-of-the-art localization and lineage tracing studies, and phenotype and genetic analysis of Ddr2-deficient mice. The work will be of broad interest to cell biologists and developmental biologists. https://doi.org/10.7554/eLife.77257.sa0 Decision letter Reviews on Sciety eLife's review process eLife digest We each have unique facial features that are key to our identities. These features are inherited, but the mechanisms are poorly understood. People with the genetic disease spondylo-meta-epiphyseal dysplasia, or SMED, have characteristic facial and skull abnormalities including a flattened face and shortened skull. SMED is associated with mutations that inactivate the gene encoding a protein called discoidin domain receptor 2 (DDR2), which is a receptor for collagen. Collagen is the major structural protein in the human body, supporting the structure of cells and tissues. It also controls cell behaviors including growth, migration and differentiation, and it helps form tissues such as cartilage or bone. At least some of the effects of collagen on cells depend on its interaction with DDR2. Since the facial and skull abnormalities in mice with mutations that stop DDR2 from working correctly resemble those of SMED patients, these mice can be used to understand the cellular basis for this disease, as well as the role of DDR2 in the embryonic development of the face and skull. Therefore, Mohamed et al. set out to understand how loss of DDR2 causes the characteristic facial and skull defects associated with SMED. Mohamed et al. used mice that had been genetically modified so that DDR2 could be inactivated in skeletal progenitor cells, cartilage cells and bone cells (osteoblasts). Examining these mice, they found that the shortened skulls and flat face characteristic of mice lacking DDR2 are due to bones at the skull base failing to elongate correctly due to defects in the growth centers that depend on cartilage. Mohamed et al. also discovered that the cells that normally produce DDR2 are the progenitors of cartilage and bone-forming cells, which partly explains why lacking this protein leads to issues in growth of these tissues. In addition to shedding light on the causes of SMED, Mohamed et al.’s results also provide general insights into the mechanisms controlling the formation of facial and skull bones that depend on interactions between cells and collagen. This information may help explain how other abnormalities in the face and skull emerge, and provide a basis for how the shape of the skull has changed during human evolution. In the future, it may be possible to manipulate the activity of DDR2 to correct skull defects. Introduction Craniofacial abnormalities affecting bone formation in the skull and face are common birth defects in humans (Fish, 2016; Trainor and Richtsmeier, 2015). The craniofacial skeleton forms through regulated bone growth in the calvarium synchronized with growth of the developing brain and cranial base. Cranial sutures, fibrous tissues adjoining calvarial bones, are the principal sites for growth of the cranial vault. Sutures contain stem cells associated with several genetic markers including glioma-associated oncogene 1 (Gli1), an intermediate in hedgehog signaling, Axin-related protein 2 (Axin2), a component of the Wnt pathway, and paired related homeobox 1 (Prrx1) (Lana-Elola et al., 2007; Maruyama et al., 2016; Wilk et al., 2017; Zhao et al., 2015). The current paradigm suggests that sutural stem cells allow continued calvarial growth through osteogenic differentiation and bone formation at sutural edges of the calvarial bones. This process is coordinated with growth of the cranial base mediated by endochondral ossification centers called synchondroses. Histologically, the synchondrosis is a mirror-image growth plate with resting chondrocytes in the central zone and proliferative and hypertrophic chondrocytes on both sides. During endochondral ossification, the resting chondrocytes in the growth plate synchondrosis undergo sequential proliferation, hypertrophic differentiation, and apoptosis before being replaced by bone. The cranial base has been implicated in many human syndromes such as craniosynostosis (premature closure of cranial sutures), Down syndrome, Turner syndrome, cleidocranial dysplasia, cleft palate, and osteogenesis imperfecta (Cheung et al., 2011; McGrath et al., 2012; Nie, 2005; Paliga et al., 2014; Tahiri et al., 2014). The extracellular matrix (ECM) is critical for maintenance of stem cell niches, differentiation, organ development, and remodeling (Lu et al., 2011). During craniofacial development, skeletal stem cells interact with the collagen-rich ECM through cell-surface receptors. These interactions regulate cell proliferation, migration, differentiation, and remodeling to control and maintain shapes of individual cranial bones and the relative proportions of skeletal elements (Erlebacher et al., 1995). Disruption of these interactions can cause craniofacial abnormalities (Lu et al., 2011; Erlebacher et al., 1995; Velleman, 2000; Kuivaniemi et al., 1997). Discoidin domain receptors (DDR1 and DDR2) are a unique class of receptor tyrosine kinases (RTKs) that, together with β1 integrins, mediate interactions between cells and the ECM by specifically binding triple-helical collagens (Vogel et al., 1997; Shrivastava et al., 1997; Leitinger and Hohenester, 2007). DDRs contain a unique discoidin (DS) domain on their extracellular surface that is homologous to discoidin I-like domain of the slime mold Dictyostelium discoideum (Leitinger and Hohenester, 2007). DS domains are required for collagen recognition and specificity of DDRs such that DDR2 preferentially binds to fibrillar collagens I-III, but not nonfibrillar basement membrane type IV collagen that is exclusively recognized by DDR1 (Vogel et al., 1997; Shrivastava et al., 1997; Leitinger and Kwan, 2006). DDR2 plays an important role during human craniofacial morphogenesis, possibly being involved in skull changes occurring during the evolution of modern humans from archaic hominins (Kuhlwilm and Boeckx, 2019). Moreover, DDR2 loss of function mutations cause the autosomal recessive growth disorder, spondylo-meta-epiphyseal dysplasia, with short limbs and abnormal calcifications (SMED, SL-AC). This disorder is associated with a number of craniofacial abnormalities including a prominent forehead, open fontanelles, hypertelorism, a short nose with a depressed nasal bridge, long philtrum and micrognathia (Alkindi et al., 2014; Ali et al., 2010; Bargal et al., 2009; Mansouri et al., 2016; Ürel-Demir et al., 2018; Smithson et al., 2009; Rozovsky et al., 2011). Ddr2-deficient mice recapitulate craniofacial phenotypes seen in SMED SL-AC patients including short snout, protruding eyes, abnormal tooth development and impaired calvarial ossification (Ge et al., 2016; Kano et al., 2008; Labrador et al., 2001; Mohamed et al., 2020). However, the basis for these abnormalities is not understood. We recently described cell autonomous functions of Ddr2 in growth of the appendicular skeleton (Mohamed et al., 2022). Preferential localization of DDR2 in GLI1-positive skeletal progenitor populations associated with resting/growth zone chondrocytes of the growth plate and bone marrow was demonstrated while conditional inactivation of Ddr2 in Gli1-expressing cells and chondrocytes phenocopied the dwarf bone phenotype of globally Ddr2-deficient mice. In the present study, we focus on DDR2 functions in craniofacial development. We show that DDR2 is associated with putative suture stem cells that can serve as precursors for osteoblasts/osteocytes in the cranial vault and with synchondrosis-associated resting/proliferating chondrocytes that form hypertrophic chondrocytes, osteoblasts, and osteocytes in the cranial base. Furthermore, as shown by global and conditional inactivation approaches, DDR2 controls suture formation, frontal bone thickness and anterior/posterior growth of the skull. Notably, these DDR2 functions involve activity in different cell populations and are correlated with changes in collagen organization and chondrocyte polarity. Results Anterior-posterior skull growth and frontal suture/bone formation impairment in Ddr2-deficient mice To begin understanding the function of Ddr2 in craniofacial development, we first defined which parts of skull are altered in global Ddr2 deficiency using Ddr2slie/slie mice. These animals contain a spontaneous 150 kb deletion in the Ddr2 locus to generate an effective null (Kano et al., 2008). We performed linear measurements on micro-CT scans of 3-month-old skulls (n=10) using previously described landmarks (Vora et al., 2015). Our analysis along the anterior-posterior (AP) axis revealed a significant reduction (12%) in skull length (SL) mainly due to shortening of the nasal bone (NB), cranial vault (CV), and anterior and posterior cranial base (ACB and PCB) in Ddr2-deficient mice compared with WT littermates (Figure 1a–c). In addition, we detected an increase in anterior skull width without any change in skull height (Figure 1d–e). We further selected orthogonal planes in MicroView to measure the thickness of calvarial bones (Figure 1f). Interestingly, frontal bone thickness was reduced by 55% with Ddr2 deficiency; however, no significant differences were observed in the thickness of parietal or occipital bones (Figure 1f–g). Further analysis using Alcian blue and Alizarin red whole mount and H&E staining of 2-week-old skulls showed defective formation of frontal sutures in Ddr2-deficient mice (defect in 3/3 mice examined). In contrast, coronal and lambdoid sutures had a normal morphology (Figure 1h–i). Ddr2-deficient calvaria also had a reduced calvarial bone marrow cavity that was most prominent in the anterior skull bones (Figure 1f and i). No major differences were observed in transverse cranial sutures, such as coronal and lambdoid. Consistent with results from 2-week-old mice, reduced mineralization in the posterior frontal suture region was also seen in most skulls from 3-month-old Ddr2slie/slie mice, although some phenotypic variability was observed at this age with 6/10 skulls affected (Figure 1b). In summary, skull deformities in Ddr2-deficient mice are associated with defects in AP skull growth, impaired frontal suture mineralization, and frontal bone thickness. Figure 1 with 1 supplement see all Download asset Open asset Ddr2 deficiency results in impaired anterior-posterior skull growth with abnormal frontal bone and suture formation. WT and Ddr2slielsile mice were compared at 3 months (a–g) and 2 weeks (h,i). (a-c) Short snout, and reduced skull length in Ddr2slielsile mice. (a) Side (upper) and top (lower) head views of 3-month-old Ddr2slie/slie mice and WT littermates. Scale bar: 1 cm. (b) 3D rendering of μCT scans of 3-month-old skulls. Scale bar: 1 mm. (c) Linear measurements along anteroposterior axis of skulls, where SL: skull length; NB: nasal bone; CV: calvaria vault; ACB: anterior cranial base; PCB: posterior cranial base. (d) Quantification of anterior (ant.) and posterior (post.) skull width showed a selective increase only in the anterior skull of Ddr2slie/slie vs WT mice. (e) No changes were observed in skull height at any of the regions measured (anterior cranial height, ACH; middle cranial height, MCH; posterior cranial height, PCH). (f,g) μCT scans of calvarial bones and quantification showing a significant reduction of frontal bone (blue) in 3-month-old Ddr2slie/slie mice in the absence of changes in parietal (orange) or occipital (green) calvarial bones. Note frontal suture defect in Ddr2slie/slie mice (red asterisk). Data are presented as mean ± SD. (n=10). *p<0.01, **p<0.01, ***p<0.001, ****p<0.0001, ns, not significant, two-tailed unpaired t test. (h) Alcian blue and Alizarin red staining of 2-week-old mouse skull whole mounts shows delayed frontal suture formation (yellow arrow) and abnormal suture morphology (white dotted lines) in Ddr2slie/slie mice compared with WT. Boxed region is shown in higher magnification; right. Scale bar: 100 μm. (i) Hematoxylin and eosin (H&E) staining shows open frontal sutures in Ddr2slie/slie, but transverse sutures, such as coronal and lambdoid were not affected (highlighted by yellow lines). Ddr2slie/slie calvariae also had a smaller bone marrow cavity (Bm, white arrows) compared with WT. Frontal suture, coronal section; transverse sutures, sagittal section. Scale bar: 100 μm. Figure 1—source data 1 Skull linear measurements in Wildtype versus Ddr2 knockout mice. https://cdn.elifesciences.org/articles/77257/elife-77257-fig1-data1-v3.xlsx Download elife-77257-fig1-data1-v3.xlsx Reduced anterior-posterior skull growth in Ddr2 deficiency is explained by abnormal chondrogenesis and delayed synchondrosis ossification Skull elongation is mainly driven by endochondral ossification at the cranial base (Hallett et al., 2022). The impaired AP skull growth observed in Ddr2slie/slie mice prompt us to examine the cranial base synchondroses and associated bones. We focused on intersphenoid (ISS) and spheno-occipital synchondroses (SOS) that form the anterior and posterior borders of basisphenoid bone (BS), the most affected cranial base bone in Ddr2slie/slie mice (Figure 2c). The BS bone forms the mid-posterior cranial base providing a foundation to support the brain, along with presphenoid (PS) and basio-occipital (BO) bones (Figure 2a). Our analysis showed that the ISS and SOS were abnormally wide in Ddr2slie/slie mice compared with WT littermates (Figure 2a–d). At 2 months, the width of the ISS and SOS was increased by 187% and 23%; respectively, while a significant increase in the height was only observed in the SOS of knockout mice (12%) (Figure 2b). Accordingly, growth of basisphenoid bone was significantly inhibited in the knockout mice (reduced by 14%) (Figure 2c). The basio-occipital (BO) bone was also significantly affected but to lesser extent. No major difference was seen in the presphenoid (PS) bone (Figure 2c). H&E staining of 2-week-old skulls revealed that the Ddr2-deficient synchondroses, particularly ISS, showed severe defects in chondrocyte organization associated with loss of the columnar arrangement of proliferative chondrocytes (Figure 2d). Figure 2 with 1 supplement see all Download asset Open asset Cranial base hypoplasia due to chondrocyte disorganization and reduced chondrocyte proliferation in Ddr2-knockout synchondroses. (a) H&E staining (upper) and μCT scans (lower) of WT and Ddr2slie/slie skulls showing wide cranial base synchondroses. Scale bar: 500 μm. Boxed region (red) is shown in higher magnification. Br: Brain; t: Tongue. ISS: Intersphenoid synchondrosis; SOS: Spheno-occipital synchondrosis; PS: Presphenoid bone; BS: Basisphenoid bone; BO: Basis-occipital bone. In μCT scan of skulls (lower), arrowheads point to cranial base synchondroses; yellow and red lines highlight the width and height of synchondroses; respectively (Quantification is shown in b). Cyan lines highlight shortening of basisphenoid bone between ISS and SOS in Ddr2slie/slie vs WT mice (Quantification of cranial base bone lengths is shown in c); Data are presented as mean ± SD. (n=10). *p<0.01, ****p<0.0001, ns, not significant, two-tailed unpaired t test. (d), H&E staining of ISS and SOS sections showing loss of columnar organization of proliferative chondrocytes (red arrowheads) in Ddr2slie/slie mice at 2 weeks of age. Red arrows point to resting chondrocyte zone. Boxed region is shown in higher magnification. Scale bar: 100 μm. (e) EdU staining (green) of ISS and SOS sections showing reduction in EdU + cells (white arrowheads) in Ddr2slie/slie mice compared with WT littermates. Boxed region is shown in higher magnification. Scale bar: 100 μm. Br: Brain, Ms: Muscle; Bm: Bone marrow. White dotted lines define the ventral surface of brain. (f) Percentage of EdU + cells in ISS and SOS of WT and Ddr2slie/slie mice. (g) TUNEL staining (brown, black arrowheads) shows no changes in apoptotic levels between mice Ddr2slie/slie and WT. Cell nuclei were stained with methyl green (green). Scale bar: 50 μm. (h) Quantification of TUNEL-positive cells in cranial base synchondroses. Data in are presented as mean ± SD (panel f, n=5-7; h, n=5). ***p<0.001, ****p<0.0001, ns, not significant, two-tailed unpaired t test. Figure 2—source data 1 Cranial base linear measurements in Wildtype versus Ddr2 knockout mice. https://cdn.elifesciences.org/articles/77257/elife-77257-fig2-data1-v3.xlsx Download elife-77257-fig2-data1-v3.xlsx Figure 2—source data 2 Proliferation (EdU labeling) and apoptosis (TUNEL assay) in Wildtype versus Ddr2 knockout mice. https://cdn.elifesciences.org/articles/77257/elife-77257-fig2-data2-v3.xlsx Download elife-77257-fig2-data2-v3.xlsx To begin understanding the cellular basis for the observed changes in the size of cranial base bones and synchondrosis structure, we assessed cell proliferation and apoptosis. EdU staining (shown in green) indicated labeling mainly in the proliferative chondrocyte zone of ISS and SOS (Figure 2e, white arrowheads); however, EdU +chondrocytes were significantly reduced in Ddr2-deficient synchondroses: 55% in ISS and 26% in SOS, compared with WT controls (Figure 2f). We also performed TUNEL assay to measure apoptosis, but did not detect differences between Ddr2-deficient and control mice in either ISS or SOS regions (Figure 2g–h). Consistent with this result, staining for cleaved caspase 3 was also not affected by Ddr2 deficiency (Figure 2—figure supplement 1). Ddr2 deficiency is associated with abnormal cranial base ECM organization While a reduction in chondrocyte proliferation may explain the smaller BS and BO bones seen in Ddr2-deficient mice, proliferation changes do not readily explain the observed synchondrosis widening. To gain further insight into the pathogenesis underlying this phenomenon, we examined the distribution of ECM proteins associated with resting and proliferative chondrocytes (type II collagen), hypertrophic chondrocytes (type X collagen) and osteoblasts (type I collagen and integrin-binding sialoprotein (IBSP)) (Robey, 2002). In wild-type mice, immunofluorescent staining of type II collagen was homogenously distributed in the cartilage matrix around chondrocytes. However, Col II was unevenly distributed in Ddr2-deficient synchondroses where more intense territorial matrix staining was observed around small clusters of chondrocytes with diminished staining in the interterritorial matrix (Figure 3a, arrows). These data suggest that DDR2 may regulate Col II fibril distribution and/or orientation. To explore this further, second harmonic generation (SHG) imaging was used to determine collagen fibril orientation in the central resting zone of ISS from WT and Ddr2slie/slie mice (Figure 3—figure supplement 1). In control synchondroses, collagen fibrils were evenly distributed and generally oriented along the A-P axis of the ISS. In contrast, fibrils in Ddr2-deficient synchondroses were more concentrated adjacent to cells and had a more randomized orientation as reflected by a dramatic decrease in anisotropy relative to WT controls. Figure 3 with 2 supplements see all Download asset Open asset ECM defects in Ddr2-deficient synchondroses. (a-d) Immunofluorescent staining of ISS and SOS sections from 2-week-old WT and Ddr2slie/slie synchondroses. (a) Representative images of COL2 immunostaining (red) in ISS and SOS showing homogenous distribution around chondrocytes in WT synchondroses, while Ddr2slie/slie mice showed uneven, ring-like immunostaining around chondrocytes (white arrowheads). Boxed region is shown in higher magnification Scale bar: μm. (b) Immunofluorescent images of immunostaining in the hypertrophic zone of synchondroses showing no major changes in staining distribution between WT and Ddr2slie/slie mice. Scale bar: 50 μm. (c) images showing (red) staining in and bones of the cranial base in WT and Ddr2slie/slie mice. Scale bar: μm. (d) Immunofluorescent images showing (red) in and bones of cranial base is in Ddr2slie/slie synchondrosis compared with WT littermates. Figure 3—figure supplement 2 for of Scale bar: μm. Cell nuclei were stained with (blue) in Bm: Bone Br: Brain; bone; bone; Ms: X collagen immunostaining showed a in the hypertrophic zone although no differences in staining were observed between WT and Ddr2slie/slie mice (Figure using I and showed staining in and bones (Figure Figure 3—figure supplement 2 for quantification of staining no differences in staining were observed WT and Ddr2slie/slie mice were However, staining was reduced in hypertrophic chondrocytes and in Ddr2-deficient SOS compared with defective bone formation and mineralization, although these differences were not seen in the ISS. In summary, Ddr2 deficiency is associated with abnormal chondrogenesis in the cranial chondrocytes, reduced chondrocyte proliferation, abnormal COL2 ECM randomized fibril orientation and delayed endochondral of Ddr2 in the craniofacial skeleton To begin the cellular functions of DDR2 to the observed craniofacial phenotype of Ddr2-deficient mice, we first examined the and distribution of cells using a mouse The of Ddr2 was examined using a of whole mount and sections from mice. During development, activity was first detected at seen at The whole mount staining at this revealed broad Ddr2 in the developing midface including the and nasal process and and and around (Figure was seen at and (Figure sections of the embryonic activity was detected in cartilage of the cranial nasal and cartilage in the developing as previously developing tooth (Mohamed et al., 2020). However, no activity was detected in the developing brain (Figure supplement Figure with 1 supplement see all Download asset Open asset Ddr2 in craniofacial (a) staining (green) of skulls showing of Ddr2 in cranial and cranial Scale bar: 50 μm. (b) staining of sections of calvaria from mice showing in suture and of bones. Scale bar: 100 and 50 right. (c) staining of of ISS and SOS from mice Ddr2 in resting and proliferative chondrocyte but or in hypertrophic chondrocytes. Boxed regions are shown in higher right. Scale bar: 50 μm. mounts of skulls from mice showed broad Ddr2 the calvarial surface with in cranial This distribution more to sutures in 3-month-old mice (Figure sections of skulls revealed Ddr2 in the suture of bones, and on the ventral surface of calvaria (Figure and were also seen in calvaria as well as in cells the bone marrow (Figure supplement 1b). However, activity was not seen in the cells calvarial bones. This that Ddr2 is during of Ddr2 was also detected in synchondroses, in the resting and proliferative chondrocyte zone but or in hypertrophic chondrocytes and in the associated bone marrow and (Figure Ddr2 was in regions in skeletal progenitor cells including cranial sutures, and and resting chondrocytes. This suggests that Ddr2 has functions in skeletal progenitor cells which to development of the craniofacial cells with GLI1 in cranial sutures and synchondroses, contributing to osteogenic and chondrogenic lineages Cranial sutures contain stem cells that to craniofacial bone formation et al., 2015). Since Ddr2 is also expressed in sutures, we DDR2 is in skeletal progenitor cells can form the major cranial bone cell To this we used a lineage-tracing by mice with mice. in mice was with for birth and skulls were and at and 2 months of age (Figure mounts of calvaria showed labeling that for at least 2 months in all cranial coronal and occipital sutures (Figure supplement At sections revealed labeling in a cells in the cranial suture and in developing and of calvarial bones (Figure At this the calvarial bone is and of a marrow weeks labeling was in cranial sutures, and a labeling intense in cranial suture where cells in the cells in the bone marrow of calvarial bones, and in osteocytes (Figure Figure supplement Therefore, the osteoblasts, and bone cells for calvarial bone formation are from Figure with 2 supplements see all Download asset Open asset progenitors of the skeletal lineage during craniofacial development. (a) used for of and of protein (red) (b) on sections of calvaria at the and showing labeling (white arrowheads) in suture and contributing to osteocytes and bone marrow of calvarial bones Scale bar: 100 μm. cranial base synchondrosis (ISS) showing labeling at in resting and proliferative chondrocyte but not in the hypertrophic zone dotted with at shows an increase in + cells in all synchondrosis regions and associated bone marrow At labeling is in the middle zone and shows a of labeling in proliferative and
Source: eLife
Language: English
Subfield: Surgery
Field: Medicine
Domain: Health Sciences
Open Access status: gold
Estimated APC paid: $2,500
OpenAlex ID: W4322629338