CNTNAP4 Functions as a Specific Receptor of NELL-1 Affecting Osteoblastic Differentiation Chen-Shuang Li1,2, Xinli Zhang1, Yan-Heng Zhou2, Zhong Zheng1, Chia Soo3, Ting Kang1 1Division of Growth and Development and Section of Orthodontics, School of Dentistry, University of California, Los Angeles, Los Angeles, CA USA, 2Department of Orthodontics, School and Hospital of Stomatology, Peking University, Beijing, China, 3UCLA Division of Plastic Surgery and Department of Orthopaedic Surgery and the Orthopaedic Hospital Research Center, University of California, Los Angeles, Los Angeles, CA 90095, USA Introduction The NELL-1 gene was identified upregulation in craniosynostosis (CS) patients within prematurely fusing sutures (Ting et al., 1999). Transgenic Nell-1 overexpression mice exhibit gross bone overgrowth and a CS- like phenotype (Zhang et al., 2002). In contrast, Nell-1 deficient mice result in major skeletal anomalies with reduced bone formation (Desai et al., 2006; Zhang et al., 2012). Subsequently, NELL-1 was approved as a novel osteogenic protein inducing regeneration of bone in multiple animal models (Cowan et al., 2012; Kwak et al., 2013; Zhang et al., 2011a). To understand the mode of action of NELL-1’s osteogenic effects, several studies have been taken to indentify the potential cell membrane associated receptor(s) or interactive proteins of NELL-1. Recently, the intracellular molecule, apoptosis-related protein 3 (APR3) (Zou et al., 2011), was identified to physically bind to NELL-1 with only modest functional alteration of NELL-1’s ostegenic effects. Integrinβ1 was also found to bind with NELL-1 to mediate osteoblastic cell adhesion (Hasebe et al., 2012a; Shen et al., 2012). Generally, the Integrins are adhesion receptors for an enormous ligands population (Goessler et al., 2008), and not specific to NELL-1. Despite these accumulative data, the specific cell surface receptor(s) of NELL-1 responsible to its osteogenic effects hasn’t been identified yet. Materials and Methods 1. T7 human brain cDNA phage library construction and biopanning As NELL-1 is highly expressed in brain tissue (Nakamura et al., 2012; Ting et al., 1999), T7Select Human Brain cDNA Library was used to construct T7 phages library. For the maximal exposure of NELL-1 to its binding partners during biopanning, HIS-Select Nickel Magnetic Agarose Beads were selected to immobilize his- tagged NELL-1. An aliquot of the amplified phages was incubated with his-NELL-1 coated beads for 4 rounds of bipanning screen. The phages bound to his-NELL-1 coated beads were eluted, and 100 plaques were selected to amplify the phage DNA by PCR. The sequences larger than 500 bp were sequenced and analyzed (Fig. 1). 2. Dissociation constant (Kd) Fig 1. Diagram of phage biopanning by using His-NELL-1 coated magnetic beads. ELISA to confirm specific phage cDNA library was constructed from human brain mRNA and packaged with T7 phage. binding affinity The library was probed by his-NELL-1 coated magnetic beads. After 4 round of 10ug/ml of rhNELL-1 or short biopanning, most of the non-specific binding phages were washed off; the remaining phages were considered binding candidate and amplified the cDNA inserts by PCR using NELL-1 was used as bait to coat specific primers for T7 phage vector. PCR products over 0.5 kb were sequenced and the ELISA plate. After removed analyzed. unbounded bait, the plate was blocked with 3% non-fat milk. 100µl/well diluted phage lysate (1X104 – 1X1010 phages/ml) was added to the coated plate and incubated at 37 for 1 hour. Then, the unbound phages were washed out thoroughly, and T7 Tail-Fiber Monoclonal Antibody was added for 1 hour incubation. The HRP conjugated anti-mouse secondary ℃ antibody was added to the plate. For HRP Table 1. List of cell types used for Cntnap4 expression. detection, TMB substrate was used, and absorbance was read at 450nm. The T7 phages without insert were used as negative control, and each concentration of phage was triplicate. 3. Realtime-PCR of Cntnap4 gene expression of cell lines and primary cells 12 types of cell lines and primary cells (Table 1) which have significant response to NELL-1 were cultured under standard condition. When the cells reached subconfluence, RNA was isolated using standard protocol. And Cntnap4 gene expression level was detected by realtime-PCR. Table 2. List of rhNELL-1 binding candidate screened by 4. Pull-Down and Co-immunoprecipitation phage biopanning. Pull-Down PolyHis Protein : Protein Interaction Kit (Pierce) was used for Pull-Down assay. In brief, prey protein was isolated from MC3T3 or NMCC, and applied to the spin column which had pre-incubated with his-NELL-1 as bait protein. The bait-prey complex was eluted for SDS-PAGE. For CO-IP assay, MC3T3 or NMCC are trypsinized, suspended in PBS, and incubated with NELL-1. BS3 was added to the cells as crosslinker. Proteins were isolated by adding cold RIPA lysis buffer to the collected cells, and incubated with anti-NELL-1 antibody coated agarose beads. The his-NELL-1 – candidate receptor(s) complex was eluted for SDS- PAGE. 5. Small hairpin RNA transfection MC3T3 cells were transfected with Cntnap4 shRNA or nontarget control shRNA. The positive transfected colonies were selected by Puromycin, and validated by Cntnap4 mRNA expression level to establish Cntnap4 KD (knockdown) cell line. 6. Immunocytochemistry and confocal microscopy Serum-starved MC3T3 cells were treated with PBS or 500ng/ml rhNELL-1 for 30min, and fixed with ice-cold methanol. The fixed cells were blocked with 3% bovine serum albumin and incubated with anti-CNTNAP4, anti- NELL-1 antibodies in blocking buffer overnight at 4 . Following three washes with PBST, cells were incubated with FITC-conjugated biotinylated anti-rabbit IgG and℃ Texas Red-conjugated biotinylated anti-Goat IgG secondary antibodies, counterstained with DAPI, and then mounted. Fluorescence was observed using an Fig 2. Confirmed binding affinity between the CNTNAP4 Olympus BX51 microscope (200X) and Leica Confocal phages and rhNELL-1 by Kd ELISA. A. By increasing the microscope (1000X). number of phages using to incubate with rhNELL-1 pre-coated 7. Osteogenic differentiation assays ELISA plate, CNTNAP4 phage shows higher binding affinity For osteogenic differentiation, MC3T3 and stable than control phage. *: p<0.05 when compared to control phage. B. CNTNAP4 phage shows high binding affinity to full Cntnap4 KD cells were seeded on 24-well plates for ALP, length NELL-1 but not NELL-1 short isoform. *: p<0.05 when Alizarin Red staining, and Immunocytochemistry, on 6- compared to control phage. well plates for osteogenic genes expression test by realtime-PCR. Cells were treated with osteogenic differentiation medium (αMEM, 10% FBS, 50µg/ml ascorbic acid, 10mM β-glycerophosphate) with or without 500ng/ml rhNELL-1 or 100ng/ml BMP-2. ALP staining and Alizarin Red staining were performed as previously described (Shen et al., 2012). 8. MAPK activation assays by western blot Sub-confluent control and Cntnap4 KD MC3T3 were subjected to serum starvation for 18 hr, and then treated with PBS or rhNELL-1 (500ng/ml) for 10 min, or 30 min at 37 . Protein isolation and western blot were performed as previously described (Zhang et al., 2011b). ℃ Results 1. Identification and confirmation of CNTNAP4 as a cell surface protein binding to NELL-1 After sequencing, 22 of rhNELL-1 binding candidates were obtained (Table 2). The phages containing partial amino acid sequence of CNTNAP4 showed high binding ability to full-length rhNELL-1 and low binding ability to short rhNELL-1 isoform (Pang et al., 2014) by binding Kd ELISA (Fig 2). Further, both pull-down and Co-IP confirmed the physical interaction between NELL-1 and CNTNAP4 in MC3T3 and NMCC which had highest expression level of Cntnap4 among 12 types of cell lines and primary cells being screened (Fig 3). Phenotypically, high level of plasma membrane staining of CNTNAP4 was detected in control MC3T3 in contrast to barely detectable expression in stable Fig 3. Confirmed the binding between CNTNAP4 and NELL-1 proteins by pull-down and Co-immunoprecipitaion in MC3T3 Cntnap4 KD cells (Fig 4A). Significantly, there is and NMCC. A. among 8 types of tested cell lines, MC3T3 has the much higher NELL-1 staining intensity in control highest expression level of Cntnap4. B. among 4 types of tested MC3T3 than Cntnap4 KD cells after 30 min incubation primary cells, NMCC has the highest expression level of Cntnap4. with rhNELL-1, which indicates rhNELL-1 binding to NMCC: newborn mouse calvaria cells. mRC: mouse rib chondrocytes. hBMSC: human bone marrow stem cells. hARC: the cell surface specifically through CNTNAP4 (Fig human articular chondrocytes. C. Pull-down assay was performed 4A). The co-localization of CNTNAP4 and NELL-1 with MC3T3 and NMCC. More CNTNAP4 was detected when was further evaluated using confocal microscopy. coated beads with his-NELL-1. D. Co-IP assay was performed Immunofluorescent staining presented that majority of with MC3T3 and NMCC. More CNTNAP4 was detected when CNTNAP4 located on the plasma membrane, and the added his-NELL-1 to the cells. No bands were detected when not coated the agrose beads with anti-NELL-1 antibody. higher staining intensity of NELL-1 was found predominantly on the cell membrane after addition of exogenous rhNELL-1 although the intensity varies among stained cells (Fig 4B). This is indicative of formation of a binding complex of these two proteins on cell surface. 2. CNTNAP4 is indispensable for NELL-1’s osteogenic effect ALP staining at day 9 and Alizarin Red staining at day 14 revealed that rhNELL-1 can significantly induce osteogenesis and mineralization with control MC3T3, while these osteogenic effects of rhNELL-1 were completely abrogated by Cntnap4 KD.