Heparan Sulfate Proteoglycans Retain Noggin at the Cell Surface: A Potential Mechanism for Shaping BMP Gradients

Bone morphogenetic proteins (BMPs) are expressed broadly and regulate a diverse array of developmental events in vivo. Essential to many of these functions is the establishment of activity gradients of BMP, which provide positional information that influences cell fates. Secreted polypeptides, such as Noggin, bind BMPs and inhibit their function by preventing interaction with receptors on the cell surface. These BMP antagonists are assumed to be diffusible, and therefore potentially important in the establishment of BMP activity gradients in vivo. Nothing is known, however, about the potential interactions between Noggin and components of the cell surface or extracellular matrix that might limit its diffusion. We have found that Noggin binds strongly to heparin in vitro, and to heparan sulfate proteoglycans on the surface of culture cells. Noggin is detected only on the surface of cells that express heparan sulfate, can be specifically displaced from cells by heparin, and can be directly crosslinked to a cell surface proteoglycan in culture. Heparan sulfate bound Noggin remains functional and can bind BMP4 at the plasma membrane. A Noggin mutant with a deletion in a putative heparin binding domain has reduced binding to heparin and does not bind to the cell surface, but has preserved BMP binding and antagonist functions. Our results imply that interactions between Noggin and heparan sulfate proteoglycans in vivo regulate diffusion, and therefore the formation of gradients of BMP activity. 2 by gest on O cber 0, 2017 hp://w w w .jb.org/ D ow nladed from Introduction The gene noggin encodes a member of one of at least four distinct gene families encoding secreted polypeptides that bind to members of the TGF-ß superfamily, such as BMP4, and inhibit the function of these signaling proteins by preventing their interaction with receptors on the cell surface (1,2). Other antagonists with related functions include Chordin, Follistatin, and members of the DAN family (1,3-6). Although structurally distinct, members of these gene families have in some cases similar ligand specificity, overlapping patterns of expression, and in the case of Chordin and Noggin these proteins apparently are capable of at least partial compensation for each other, for example during forebrain development in the mouse (7). In addition to these multiple secreted BMP antagonists, there are other secreted proteins whose primary function is to overcome this antagonism (8,9). Thus, there is a highly complex system to regulate the bioavailability and consequently the activities of BMPs in the extracellular space. Members of the BMP gene family are broadly expressed and their functions have been implicated in a wide range of developmental processes (10,11). In most situations, BMPs appear to function as morphogens (12-14). Morphogens function over long ranges through the formation of gradients of activity. Cells sense the strength of the morphogen signal across this gradient and respond by induction of dose dependent patterns of gene activation, which then serve to specify cell fate. Despite evidence for gradients of BMP activity in many tissues, there is no evidence that bmp4 transcripts have graded expression patterns in many of these same tissues (15-17). This has led to the suggestion that post-translational mechanisms are responsible for the establishment of BMP activity gradients in vivo (12,14,18). Specifically, it has been hypothesized that the diffusion of a secreted BMP antagonist from organizer regions 3 by gest on O cber 0, 2017 hp://w w w .jb.org/ D ow nladed from might establish gradients of BMP antagonism, thereby resulting in inverse activity gradients of BMP. Therefore, BMP antagonists themselves may behave as morphogens. Blitz et al. have recently demonstrated that when Chordin is expressed at high doses by microinjection of mRNA it can directly function to inhibit BMP over large distances in the Xenopus embryo and can apparently establish a BMP activity gradient over several cell diameters (19). It remains untested, however, whether endogenous Chordin similarly acts over long distances during normal development (19). Testing of this hypothesis is restricted to indirect interpretations of the range of BMP antagonist function since adequate reagents do not exist to localize the precise physical range of any BMP antagonists in vivo in relation to their site of production. Accurately predicting this physical range is dependent on understanding the potential interactions between BMP antagonists and components of the cell surface and the extracellular matrix in vivo. Little is known about these potential properties, and typically it has been assumed that these proteins are in fact freely diffusible following secretion. Heparan sulfate proteoglycans are found abundantly on the surface of all adherent cells and within the extracellular matrix where they bind and regulate the functions of a wide range of ligands (20). More importantly, members of the glypican family of heparan sulfate proteoglycans have been found to specifically modify cellular responsiveness to BMPs in vivo (21,22). We therefore hypothesized that heparan sulfate proteoglycans in vivo might regulate BMP function through interactions with BMP antagonists. In this study we have evaluated the ability of heparan sulfate proteoglycans to bind to one well characterized BMP antagonist, Noggin, and to influence its cellular localization in cultured cells. We report that Noggin binds strongly to heparin-Sepharose in vitro and to heparan sulfate 4 by gest on O cber 0, 2017 hp://w w w .jb.org/ D ow nladed from proteoglycans on the surface of cultured cells, thereby localizing Noggin to the cell surface. This bound Noggin remains functional and indeed can bind BMP4 to the cell surface. Genetically engineered mutant Noggin proteins bearing deletions in a putative heparin-binding domain significantly reduce Noggin’s ability to bind heparin (23) and eliminate binding to the cell surface. Furthermore, these mutations have been shown to have little effect on the function of Noggin as a BMP antagonist (23) and are consistent with the crystal structure of the Noggin/BMP-7 complex which shows that the heparin binding site lies in a separate domain from that which binds BMP (Jay Groppe, Jason Greenwald, Aris N. Economides, Markus Affolter, and Senyon Choe, in preparation). These results suggest that it is likely that interactions between Noggin and heparan sulfate proteoglycans in vivo play a significant role in the physical range of Noggin actions. ‘ Experimental Procedures AntibodiesRP57-16, RP57-21, and 57-06 were a gift of Regeneron. These rat monoclonal antibodies were generated using native human noggin protein as immunogen. Ascites fluid from SCID mice, affinity purified by protein G affinity chromatography, was used in Western blotting, immunoprecipitation, and immunofluorescence as indicated below. PlasmidsThe eukaryotic expression plasmids, pSRα.hNog and pSRα.hNog∆B2, encoding expression of full-length human Noggin (hNog) and a genetically engineered deletion mutant were a gift of Regeneron. For the generation of stably transfected cell lines, cells were co5 by gest on O cber 0, 2017 hp://w w w .jb.org/ D ow nladed from transfected with pSVZeo from Invitrogen. Cell Culture and TransfectionParental Chinese Hamster Ovary cells (CHOK1) as well as mutant lines derived from these same cells, but which are defective in either heparan sulfate (PGSD-677) or both heparan and chondroitin sulfate (PGSA-745) biosynthesis, were a gift of Jeff Esko (UC San Diego, CA). Cells were maintained in DMEM/F12 media (BioWhittaker) containing 10% fetal bovine serum (Hyclone). Liposome mediated transfection was performed using Geneporter (Gene Therapy Systems) according to the manufacturers recommendations. Cells were cotransfected with pSVZeo (Invitrogen, San Diego, CA), selected with 0.5 mg/ml Zeocin (Invitrogen, San Diego, CA) in DMEM/F12 media containing 10% fetal bovine serum, and individual clones harvested and subcultured. Western blotting with RP57-16 antibody identified positive clones expressing similar levels of Noggin. Metabolic Labeling, Pulse Chase and ImmunoprecipitationFor metabolic labeling of Noggin expressing cultures, cells were incubated in methionineand cysteine-free DMEM (Life Technologies, Inc.) for 40 min. Translabel-3535 (ICN) was then added to each well at 200 μCi/ml, and cells incubated at 37°C for 30 minutes. Subsequent to this pulse, cells were washed once with DMEM/F12 media containing 10% fetal bovine serum, and chased in the same media. For competition experiments heparin or chondroitin sulfate (Sigma, St. Louis, MO) were added during the chase period at 1 μg/ml. At the specified time intervals, media was recovered and the cell layers lysed using cold 1% Nonidet P-40, 0.5% deoxycholic acid, 0.1% SDS, 1 mM magnesium chloride, 0.5 mM 6 by gest on O cber 0, 2017 hp://w w w .jb.org/ D ow nladed from calcium chloride in phosphate buffered saline (PBS) containing protease inhibitors of 1 μg/ml pepstatin A, 0.25 mg/ml N-ethylmaleimide, and 0.5 mM phenylmethylsulfonyl fluoride. Media was brought to similar conditions by addition of concentrated buffer. Cell debris was removed from both cell layer extracts and media samples by centrifugation at 14,000 rpm for 1 min. Supernatants were then precleared by the addition of rat IgG (Sigma, St. Louis, MO) at 1 μg/ml, incubated for 30 min at 4°C, followed by addition of Protein-G Sepharose and incubation overnight at 4°C. Samples were then centrifuged and the supernatants used for immunoprecipitation. RP57-16 (Regeneron, Tarrytown, N.Y.), was added to each sample at 1 μg/ml and incubated for 1 h at 4°C. Protein-G Sepharose was subsequently added and incubated for an additional 1 h. Beads were washed twice with the original lysis buffer without protease inhibitors and then twice with PBS containing magnesium and calcium chloride. Immunoprecipitated products were recovered from the beads by boiling in SDS-PAGE sample buffer and analyzed by electrophoresis on 10-20% gradient SDS-PAGE gels (BioRad, Richmond, CA), detecting immunoprecipitated products by autoradiography or PhosphorImager. Western Blot AnalysisConditioned media and extracts from cells expressing human noggin, as well as fractions separated by affinity chromatography of Noggin on heparin-Sepharose, were electrophoresed on 10-20% gradient SDS-PAGE gels (BioRad), and transferred to Zetaprobe (BioRad) by electoblotting. Filters were blocked for 1 h at room temperature using 5% non-fat dry milk (NFDM) in 20 mM Tris-HCl, pH 7.4, 150 mM sodium chloride (TBS), plus 0.1% Tween 20 (TBST). Blocked filters were probed with RP57-16 at 20 ng/ml in 2.5% NFDM in 7 by gest on O cber 0, 2017 hp://w w w .jb.org/ D ow nladed from TBST for 1 h at room temperature. Following three washes with TBST, incubation with an antirat HRP-conjugated secondary antibody in TBST for 1 h, and three remaining washes with TBST, protein bands were detected by ECL. Inorganic Sulfate Labeling, Crosslinking, and ImmunoprecipitationConfluent monolayers of control CHOK1 cells, and CHOK1 cells transfected with Noggin, were labeled overnight with 100 μCi/ml 35S-Na2SO4 (ICN) in S-MEM media containing dialyzed FCS (Gibco). Cells were washed three times with PBS, DTSSP (Pierce) was added to a final concentration of 5 mM, and incubated for 30 min at 4°C (24). Monolayers were washed three times with TBS to quench the crosslinking reaction, followed by three additional washes with PBS, and then subsequently extracted and immunoprecipitated as described above. After the final PBS wash, the beads were resuspended in 100 μl of 20 mM Tris pH 7.5 containing 5 mM CaCl2 and then incubated for 180 min at 37°C plus or minus two additions of Heparitinase I (Sigma) of 2 mU/ml spaced 90 min apart. Following two final washes with PBS, the immunoprecipitates were eluted with 2x SDS loading buffer (BioRad) containing ß-mercaptoethanol, boiled, centrifuged and analyzed on a 420% gradient SDS-PAGE gel (BioRad). Heparin binding affinity-A column containing 1 ml of heparin-Sepharose was prepared and attached to an AKTA FPLC unit (Amersham-Pharmacia) and run at a flow rate of 1 ml/min. The resin was equilibrated in 20 mM Tris-HCl, pH 7.4 and then 10 μg of purified recombinant Noggin in the same buffer was loaded and the column washed for 10 column volumes. Bound 8 by gest on O cber 0, 2017 hp://w w w .jb.org/ D ow nladed from protein was eluted in the same buffer using a linear gradient up to 2 M NaCl over 20 column volumes. For comparisons of hNog and hNog∆B2 binding to heparin, 160 microliters of heparin-Sepharose (Sigma) was packed into a mobicol column (Mobitec). Following equilibration with 20 mM Tris-HCL, pH 7.5, supernatants containing 250 ng of hNog or hNog∆B2 were applied to each column. Columns were the washed with 10 ml of equilibration buffer and subsequently eluted sequentially with two column volumes each of 20 mM Tris pH 7.5 with increasing amounts of salt. Eluted samples were separated by SDS-PAGE electrophoresis and Noggin detected by Western as described above. ImmunofluorescenceLive cells grown on glass coverslips were washed twice with warm serum-free medium, incubated with primary antibodies for Noggin (RP57-16 at 4.4 μg/ml) for 30 min at 37°C, then washed twice with warm medium, and incubated with secondary antibodies containing 0.01 mg/ml Hoechst for 30 min at 37°C. After washing twice with PBS, the cells were fixed with 4% paraformaldehyde in PBS for 10 min., washed twice with PBS followed by a water rinse, and then mounted in ProLong Antifade (Molecular Probes). Immunofluorescence microscopy was performed with an Olympus Fluoview 500 configured with Krypton and UV with the appropriate wavelength filters (568 and 351 nm) for CY3 and Hoechst excitation. Confocal images were assembled as montages using Adobe Photoshop. 125I-BMP-4 Binding, Affinity Cross-linking, and Analysis of ComplexesBMP4 (R and D Systems) was labeled with 125I using Iodogen tubes (Pierce). Confluent monolayers of cells 9 by gest on O cber 0, 2017 hp://w w w .jb.org/ D ow nladed from were washed 3 times with PBS at 4°C. 100 ng/ml of 125I-BMP4 was added and following incubation for 90 min at 4°C with gentle shaking, cells were washed 3 times with PBS. DTSSP (Pierce) at 5mM in PBS was added to the monolayer and incubated for 30 min at 4°C. The crosslinking reaction was halted by washing 3 times with TBS and then twice with PBS. Cell layers were extracted and immunoprecipitated as described above. Osteogenic Differentiation AssayC2C12 myoblasts were maintained in DMEM/F12 supplemented with 10% FCS. Cells were plated in 24 well plates to be 70% confluent after incubating at 37°C overnight. Dilutions of supernatants containing Noggin or ∆B2 Noggin were mixed with BMP4 and allowed to incubate at room temperature for 30 min, at which point they were added to the C2C12 cells and incubation extended overnight at 37°C. Following cell lysis, alkaline phosphatase activity was measured using a colorimetric assay (Sigma).