Connective Tissue Growth Factor (CTGF/CCN2) is Required for Normal Intramembranous and Endochondral Ossification

1Lambi, A G; 2Pankratz, T L; 1Hendesi, H; 1Pixley R A; 1Barbe, M F; 2Richtsmeier, J T; 1Popoff, S N.

1Temple University School of Medicine, Philadelphia, PA, 2Pennsylvania State University, University Park, PA Senior author: spopoff@temple.edu

Introduction: Connective tissue growth factor (CTGF, CCN2) has emerged as an

important growth factor for skeletogenesis. Its importance in skeletogenesis is seen in global CTGF knockout (KO) mice, which demonstrate defects in growth plate chondrogenesis and skeletal defects including kinked ribs, tibiae, radii and ulnae, and craniofacial abnormalities. Ultimately, global ablation of CTGF results in neonatal lethality from respiratory failure. This study addressed the effects of global CTGF ablation on skeletogenesis using quantitative techniques, such as micro-CT. Furthermore, to discern any differences in the separate processes of ossification affected by CTGF ablation, specific bone sites and cells were studied. Methods: To address our goals, we used: (a) micro-CT analysis of long bones and skulls to assess phenotypic differences, (b) histological analyses to assess parameters of endochondral bone formation, and (c) ex vivo studies to assess bone cell function. CTGF heterozyogous mice (CTGF+/LacZ) were used as breeders to obtain the mice for this study. Animal care and use was monitored by the University Animal Care and Use Committee to assure compliance with Federal and NIH regulations. Newborn animals used for this study were sacrificed at birth (P0). Micro-CT scanning was accomplished using a Skyscan 1172, 11 MPix camera model, high-resolution cone-beam micro-CT scanner. For trabecular analysis, distal femora and proximal tibiae were scanned at a pixel size of 5.2µm and a distance of 250µm from the growth plate was taken. For skull analyses, entire skulls were scanned at a pixel size of 9.4µm. For landmark analysis, 3D coordinate locations of cranial landmarks was recorded using eTDIPS software (www.getahead.psu.edu). For histologic analysis, plastic-embedded sections of tibiae and femora were stained with von Kossa, Safranin-O, and Masson’s trichrome for mineralization, cartilage, and a general bone stain, respectively. Immunostaining of paraffin sections was also conducted for osteoblast (Runx2, Osterix, Osteocalcin) and osteoclast (Cathepsin K, actin, TRAP) markers. Histologic quantification was accomplished with BioquantOsteo program. Ex vivo studies of bone cell shape and function was carried out with proliferation, differentiation, and adhesion studies. Primary osteoblasts were isolated from newborn mouse calvariae and used for either cell proliferation assay, differentiation under osteogenic culture conditions, or to assess the ability of cells to adhere to various matricellular components (fibronectin, vitronectin). Results:

To quantify differences in bone resulting from the process of endochondral ossification, we first performed analysis on distal femoral and proximal tibial metaphyses. Our results demonstrated a decrease in trabecular bone volume in CTGF KO mice compared to WT littermates. In CTGF-null tibial diaphyses (the region of the structural kink), total TV, BV and bone perimeter (B.Pm.) were markedly increased, compared to WT tibiae.

Figure Legend. *p<0.05 and **p<0.01, compared to WT.

CTGF-null skulls demonstrated decreases in cranial bone mineralization (being absent in some sites), increases in suture and fontanel size, and a clear failure of midline convergence of the maxillary and palatine processes to form the palate. This was also accompanied by asymmetry of the vomer, a consistent feature found in CTGF KO skulls. Histologic analyses of WT and CTGF-null tibiae showed differences in various markers of endochondral bone formation, including decreased osteoblast cell numbers immediately below the growth plate, decreased bone matrix formation, and increased hypertrophic zone thickness. We also demonstrated a unique dichotomy of bone formation in the kinked tibiae, where ossification at the kink appears to form by a mechanism distinct from the region under the growth plate.

Ex vivo studies from isolated primary osteoblasts revealed differences in cell function comparing CTGF-null cells to WT. We demonstrated that CTGF-null osteoblasts have decreased ability to adhere to bone matricellular proteins, such as fibronectin. Furthermore, these cells also show decreased ability to proliferate, a feature consistent over various time periods and cell concentrations. Figure Legend. Boxes 1,2, and 3 in histologic sections correspond to the graphically depicted regions ‘Proximal Metaphysis’, ‘Proximal Diaphysis’, and ‘Mid Diaphysis’, respectively. Scale bars are equal to 500µm. Discussion: CTGF KO mice demonstrated quantitative defects in skeletogenesis, through both endochondral and intramembranous ossification processes, as well as at the cellular level. Micro-CT and histologic assays revealed abnormal endochondral ossification of long bones, such as decreased trabecular bone formation and osteoblast cell number in proximal tibial metaphyses, as well as increases in hypertrophic zone length in proximal tibial growth plates. In addition, a unique dichotomy of bone formation within the kinked tibiae alone was seen. Three dimensional landmark analysis of CTGF KO skulls revealed decreased ossification of skull flat bones, indicative of disruptions in intramembranous ossification. Furthermore, craniofacial aberrations in CTGF KO mice included a failure of palatine bones to converge toward the midline, and asymmetry of the vomer. Lastly, CTGF-null osteoblasts demonstrated decreased ability to adhere and proliferate in ex vivo culture. While the defects in osteoblast function could explain, in part, the gross decreases in bone formation in CTGF KO mice, findings such as the dichotomy of bone formation within the tibiae suggest a more complex disruption of normal skeletogenesis is occurring in these mice. Significance: This study is crucial in elucidating the role of CTGF in bone development, because for the first time it addressed the effects of CTGF ablation on bone development using quantitative techniques, such as micro-CT analyses. These analyses have generated novel information regarding the global skeletal effects of CTGF ablation (i.e. kinked long bones), information necessary to comprehensively define the role of CTGF in bone formation and provide physicians with better tools for preventing and treating patients with conditions of bone loss and/or gain. Acknowledgement: This work was supported by NIAH-NIAMS grant to SNP (AR047432) and PA Department of Health grant to AGL.

Poster No. 0535 • ORS 2012 Annual Meeting