6B4 Proteoglycan/Phosphacan, an Extracellular Variant of Receptor-like Protein-tyrosine Phosphatase (cid:122) /RPTP (cid:98) , Binds Pleiotrophin/Heparin-binding Growth-associated Molecule (HB-GAM)*

A major chondroitin sulfate proteoglycan in the brain, 6B4 proteoglycan/phosphacan, corresponds to the extracellular region of a receptor-like protein-tyrosine phos- phatase, PTP (cid:122) /RPTP (cid:98) . Here, we purified and character-ized 6B4 proteoglycan-binding proteins from rat brain. From the CHAPS (3-[(3-cholamidopropyl)dimethylam-monio]-1-propanesulfonic acid) extract of brain micro- somal fractions, 18-, 28-, and 40-kDa proteins were spe-cifically isolated using 6B4 proteoglycan-Sepharose. N-terminal amino acid sequencing identified the 18-kDa protein as pleiotrophin/heparin-binding growth-associ- ated molecule (HB-GAM). Scatchard analysis of 6B4 pro-teoglycan-pleiotrophin binding revealed low ( K d (cid:53) 3 n M ) and high ( K d (cid:53) 0.25 n M ) affinity binding sites. Chon- droitinase ABC digestion of the proteoglycan decreased the binding affinities to a single value ( K d (cid:53) 13 n M ) without changing the number of binding sites. This suggested the presence of two subpopulations Other Methods— IgG fraction from anti-6B4 proteoglycan antibody The protein concentration was determined a as the

The expression of the phosphate-buffered saline (PBS 1 )-soluble brain-specific chondroitin sulfate proteoglycan, 6B4 pro-teoglycan/phosphacan, with a 300-kDa core glycoprotein, is dynamically regulated in the developing rat brain (1,2). Cloning the cDNA for 6B4 proteoglycan revealed that it is an extracellular variant of a receptor-like protein-tyrosine phosphatase, PTP/RPTP␤ (3,4). We then demonstrated that PTP is also present in the form of chondroitin sulfate proteoglycan with a 380-kDa core glycoprotein in the rat brain (3). The 6B4 proteoglycan is composed of an N-terminal carbonic anhydrase (CAH)-like domain, a fibronectin type III domain, and a Cterminal serine, glycine-rich domain that is thought to be the chondroitin sulfate attachment region (3,4). In PTP, the serine, glycine-rich domain is followed by a membrane-spanning region and two tyrosine phosphatase domains (5,6). PTP and 6B4 proteoglycan are widely distributed in the developing nervous system and may play roles in neuronal cell migration, neurite extension, axonal outgrowth, and development of the cerebellar mossy fiber system (1,2,(7)(8)(9)(10). We demonstrated that 6B4 proteoglycan is a repulsive substrate for cell adhesion but promotes neurite extension and differentiation of cortical neurons (11). These results suggest that 6B4 proteoglycan as well as PTP participate in various aspects of brain development by regulating the tyrosine phosphorylation level of intracellular proteins. However, little is known about the signal transduction system coupled extracellularly or intracellularly with these proteoglycan molecules.
Several proteins bind 6B4 proteoglycan/phosphacan and PTP/RPTP␤ (8,9,12). Screening various extracellular matrix and cell adhesion molecules revealed that N-CAM, Ng-CAM, and tenascin bind to phosphacan (8). Contactin was recently identified as PTP/RPTP␤-binding protein by screening an expression cDNA library using the CAH domain as probe (12). These studies indicated that 6B4 proteoglycan and PTP bind various extracellular and cell surface molecules. At present, however, the specific ligands that regulate the phosphatase activity of PTP upon their binding are unknown. In this study, we investigate ligands that bind to 6B4 proteoglycan and PTP by means of affinity chromatography using a matrix coupled with intact or chondroitinase ABC-digested 6B4 proteoglycan. Several proteins bound to 6B4 proteoglycan-Sepharose dependently on or independently of chondroitin sulfate moiety. An 18-kDa protein, which bound to 6B4 proteoglycan-Sepharose independently of chondroitin sulfate, was identified as pleiotrophin.
Pleiotrophin (13), also known as heparin-binding growthassociated molecule (HB-GAM) (14), or heparin-binding neurite-promoting factor, HBNF (15), is a mitogenic and neurite-promoting factor isolated from the brain. Pleiotrophin shows 50% homology to midkine, and they are considered to constitute a new growth factor gene family (16). Syndecan-3/N-syndecan binds pleiotrophin (17), but the signal-transducing receptor for this factor is unknown. In this study, we examine the function of 6B4 proteoglycan and PTP as a possible receptor of pleiotrophin. Preparation of CHAPS Extract-Postnatal 16-day-old (P16) Sprague-Dawley rats were anesthetized and killed by decapitation, then the whole brains were dissected. 16 g of tissues were mixed with 5 volumes of a solution containing 0.32 M sucrose, 5 mM EDTA, 0.1 mM PMSF, 10 M pepstatin A, 10 M leupeptin, and 50 mM Tris-HCl, pH 7.5, and homogenized in a glass-Teflon Potter homogenizer with 10 strokes at 850 rpm. The homogenate was centrifuged at 1,000 ϫ g for 5 min at 2°C, and the precipitate was washed again under the same conditions. The combined supernatants were centrifuged at 105,000 ϫ g for 60 min at 2°C to precipitate the P2ϩP3 fraction. This fraction was resuspended in 40 ml of a solution containing 0.2 M NaCl, 1 mM EDTA, 10 M pepstatin A, 10 M leupeptin, and 10 mM Tris-HCl, pH 7.5. After adding 4.4 ml of 10% CHAPS, the solution was stirred overnight at 0°C. The sample was centrifuged at 105,000 ϫ g for 60 min at 2°C, and the supernatant was applied to affinity chromatography.

Materials-Human
Affinity Chromatography-We purified 6B4 proteoglycan from 20day-old Sprague-Dawley rat whole brains under dissociative conditions as described (2). This preparation was estimated to be more than 97% pure based on SDS-PAGE. Chondroitinase ABC digestion of 6B4 proteoglycan proceeded as follows. The purified proteoglycan was dialyzed against 200 ml of 50 mM Tris-HCl, pH 7.5, containing 1 mM EDTA and 30 mM sodium acetate, then incubated for 1 h at 37°C in the presence of 50 milliunits of protease-free chondroitinase ABC. 2.5 ml of 150 g/ml (as protein) intact or chondroitinase ABC-digested 6B4 proteoglycan was dialyzed three times against 0.5 M NaCl, 0.1 M NaHCO 3 at 4°C. The proteoglycan was coupled to 1 ml of CNBr-activated Sepharose 4B according to the manufacturer's protocol.
A mixture of 50 l of 1 M MgCl 2 and CaCl 2 and 10 ml of the CHAPS extract was added to 0.5 ml of 6B4 proteoglycan-Sepharose in a 15-ml centrifuge tube. After rotating overnight at 4°C, the treated gel was poured into a small column and washed with 5 ml of 0.2% CHAPS, 0.2 M NaCl, 10 mM Tris-HCl, pH 7.5, followed by 5 ml of 10 mM EDTA, 0.2% CHAPS, 0.2 M NaCl, 10 mM Tris-HCl, pH 7.5. The bound proteins were then eluted with 5 ml of 1 M NaCl, 0.2% CHAPS, 10 mM Tris-HCl, pH 7.5.
Amino Acid Sequence Analysis-The eluted 6B4 proteoglycan-binding proteins (ϳ3 g) were separated by 12.5% SDS-PAGE and transferred to a polyvinylidene difluoride membrane according to the method of Towbin et al. (19). The proteins were stained for 3 min with 0.5% Ponceau S in 1% acetic acid and then destained with 1% acetic acid. Protein bands were excised with a razor blade, and the amino acid sequences were determined using an Applied Biosystems sequencer 492.
The reaction mixture was incubated for 3 h on ice, and then 30 l of 1 M glycine, pH 7.5, was added. After a 2-h incubation at 4°C, free [ 125 I]Bolton-Hunter reagent was removed by passing through a Chroma Spin 30 column equilibrated with 0.05% Triton X-100, 0.5 mg/ml bovine serum albumin (BSA), 0.15 M NaCl, 10 mM sodium phosphate, pH 7.2. The specific radioactivity was 2.0 ϫ 10 6 cpm/g. Binding Assay-Wells of Nunc Maxisorp Immunoplates were coated with 2ϳ5 g/ml of proteins in 35 l of 5 mM Tris-HCl, pH 8.0, at 4°C overnight. The wells were washed three times with PBS and then blocked with 1% BSA/PBS for 1 h at room temperature. We diluted 125 I-6B4 proteoglycan in 0.5% BSA, 2 mM CaCl 2 , 2 mM MgCl 2 , 0.1% CHAPS, 0.15 M NaCl, 10 mM sodium phosphate, pH 7.2, and then added 125 I-6B4 proteoglycan (15,000 cpm/35 l for the usual assay and 2,000ϳ80,000 cpm/35 l for Scatchard analyses) to the coated wells and incubated the plates for 5 h at room temperature. The wells were washed three times with 2 mM CaCl 2 , 2 mM MgCl 2 , 0.15 M NaCl, 10 mM Tris-HCl, pH 7.2. The bound materials were released by adding 200 l of 0.1 M NaOH, 0.2% SDS to the wells. The plate was shaken for 15 min at room temperature, and then the eluted radioactivity was measured using a gamma counter.
We digested 125 I-6B4 proteoglycan with glycanases as follows. First, we diluted 125 I-6B4 proteoglycan with 0.5% BSA, 2 mM CaCl 2 , 2 mM MgCl 2 , 0.15 M NaCl, 10 mM sodium acetate, 10 mM Tris-HCl, pH 7.5, to a final concentration of 0.2 g/ml. Various concentrations of proteasefree chondroitinase ABC and heparitinases I and II were then added to the samples, and the solutions were incubated for 1 h at 30°C for use in binding assays.
Preparation of Dissociated Cortical Neurons-Cerebra were dissected from embryonic 16-(E16) or 17-day-old Sprague-Dawley rats, and the meninges were removed. The tissues were incubated in Ca 2ϩand Mg 2ϩ -free Hanks' balanced salt solution (CMF-HBSS) containing 0.1% trypsin for 15 min at 37°C. After three washes with CMF-HBSS, the tissues were triturated with Pasteur pipettes in CMF-HBSS containing 0.025% DNase I, 0.4 mg/ml soybean trypsin inhibitor, 3 mg/ml BSA, and 12 mM MgSO 4 . The cell suspension was centrifuged at 160 ϫ g for 5 min at 4°C, and the pelleted cells were washed once with CMF-HBSS. Cells were resuspended in culture medium consisting of a 1:1 mixture of Dulbecco's modified Eagle's medium and F12 medium (DF medium) containing 2% B-27 supplement (DF/B-27 medium) and were seeded as described below.
Cell Culture-Glass coverslips (13 mm in diameter) were incubated in 0.002% poly-L-lysine, washed five times with distilled water, and then air dried. Cells (40 l of a 1 ϫ 10 5 cells/ml suspension) were plated on coverslips in 24-well plates. After a 1-h incubation at 37°C under 5% CO 2 , 0.5 ml per well of DF/B-27 medium was added. The cells were cultured overnight and processed for immunocytochemistry as described below.
For the neurite extension assay, wells of Nunc Maxisorp Immunoplates were coated overnight with 2 g/ml pleiotrophin or 2 g/ml poly-L-lysine at 4°C. The wells were washed three times with distilled water and then blocked with 0.25% BSA for 2 h at 37°C. The solutions were removed, and then 40 l of DF/B-27 medium containing antibodies at various concentrations were added to the wells followed by 1.5 ϫ 10 4 cortical neurons (10 l in the medium) per well. Phase-contrast micrographs were taken after a 50-h incubation, and neurite length distribution was analyzed as described (11).
Immunocytochemistry-Cultures were washed once with PBS and then fixed with 4% paraformaldehyde, 0.1 M sodium phosphate buffer, pH 7.5, for 15 min. Fixed cells were rinsed three times with Trisbuffered saline, permeabilized with 0.2% Triton X-100/PBS for 30 min, then blocked with 2% BSA, 4% goat serum in PBS. Cells were then incubated for 2 h with the mixture of anti-phosphorylated neurofilament antibodies, SMI 312 (1/500), and anti-6B4 proteoglycan antiserum (1/500). After three washes with PBS, the cells were incubated for 30 min with fluorescein isothiocyanate-conjugated anti-mouse IgG (1/100) and biotinylated anti-rabbit Ig (1/200), followed by three washes with PBS, and then incubated once more for 30 min with Texas Red Avidin D (1/1000). The cells were washed three times with PBS, mounted, and observed using a Zeiss fluorescence microscope. All solutions were diluted with PBS containing 0.05% BSA. Incubations were at room temperature.
Immunoprecipitation-Cortical neurons (10 7 cells) in DF/B-27 medium were plated on 6-cm culture plates coated with poly-L-lysine as described above. After 3 days of culture under 5% CO 2 at 37°C, the cells were extracted for 2 h at 4°C with 1.5 ml of 0.2% Triton X-100, 0.1% sodium deoxycholate, 1 mM PMSF, 10 M pepstatin A, 10 M leupeptin, 1 mM EDTA, 50 mM Tris-HCl, pH 7.5 (buffer A). After centrifugation at 15,000 ϫ g for 15 min, 500-l aliquots of the supernatant were adsorbed for 1 h at 4°C with 25 l of Protein G Sepharose 4FF. After a brief centrifugation, 1 g of anti-6B4 proteoglycan antibody (3) or control rabbit IgG was added to the supernatants and incubated for 1 h at 4°C. Thereafter, 25 l of Protein G Sepharose 4FF was added to the solution, and the tubes were rotated for 1 h at 4°C. The gels were washed three times with 0.5 ml of buffer A and then mixed with the same volume of 5 mM EDTA, 1 mM PMSF, 0.1 mM pepstatin A, 60 mM sodium acetate, 0.2 M Tris-HCl, pH 7.5. Protease-free chondroitinase ABC (5 milliunits) was added to the samples, which were then incubated for 1 h at 37°C. The samples were mixed with 16 l of 4-fold concentrated SDS-PAGE sample buffer and boiled for 5 min. The samples were applied to 5% SDS-PAGE and processed for immunoblotting as described (2).
Other Methods-IgG fraction from anti-6B4 proteoglycan antibody (3) was prepared with HiTrap Protein G according to the supplier's protocol. The protein concentration was determined using a MicroBCA kit with BSA as the standard.

RESULTS
Identification of 6B4 Proteoglycan-binding Protein-To purify 6B4 proteoglycan-binding proteins, the CHAPS extracts of the P2ϩP3 fraction from P16 rat whole brain were mixed with 6B4 proteoglycan-Sepharose in the presence or absence of calcium and magnesium ions. The mixture was rotated overnight and then poured into a column. After washing the column with a buffer containing 0.2 M NaCl, the bound proteins were eluted with 10 mM EDTA and then with 1 M NaCl (Fig. 1). Several proteins that eluted with EDTA seemed to be aggregated in the presence of divalent cations because similar proteins were also eluted with EDTA from a control Sepharose CL-4B column (data not shown). Three major proteins (18-, 28-, and 40-kDa proteins) were eluted with 1 M NaCl from the 6B4 proteoglycan-Sepharose column but not from Sepharose CL-4B. Divalent cations (5 mM each) were not essential, but in their absence, these proteins did not bind efficiently to the column.
Next, chondroitinase ABC-digested 6B4 proteoglycan was coupled to a Sepharose gel for affinity chromatography and tested as described above. Under these conditions, 40-and 28-kDa proteins flowed through the column, and only the 18-kDa protein was eluted with 1 M NaCl, suggesting that the two former proteins are chondroitin sulfate-binding proteins (data not shown). Gas phase amino acid sequencing indicated that the N-terminals of the 28-and 40-kDa proteins were blocked, since the N-terminal amino acid was undetectable. Amino acid sequences obtained from the CNBr-digested samples could not be found in the sequence of proteins that have been reported to date (data not shown). On the other hand, the N-terminal amino acid sequence of the 18-kDa protein obtained was GKKEKPEKKVKKSDXGEXQXSVXVPT, which faithfully corresponded to that of pleiotrophin except for the four X positions (13).
Binding of 6B4 Proteoglycan to Pleiotrophin-The results of the affinity chromatography suggested that pleiotrophin specifically binds to 6B4 proteoglycan. We therefore measured the binding of 125 I-6B4 proteoglycan using ELISA plates coated with various proteins. Among these, pleiotrophin and tenascin significantly bound to 6B4 proteoglycan, whereas aFGF, fibronectin, and laminin did not ( Fig. 2A). Tenascin reportedly binds phosphacan (8). The binding of 125 I-6B4 proteoglycan to pleiotrophin was dependent on the concentration of the proteoglycan, and the nonspecific binding was negligibly low (Fig.  2B). These results indicate that 6B4 proteoglycan bound to pleiotrophin.
Effects of Glycosaminoglycans on the Binding of 6B4 Proteoglycan to Pleiotrophin-Pleiotrophin is a heparin-binding protein that binds to the heparan sulfate chains of syndecan (17,20). The 6B4 proteoglycan preparation did not contain any detectable amount of heparan sulfate proteoglycan. It was also confirmed that the preparation did not contain N-syndecan by immunoblotting analysis using anti-N-syndecan monoclonal antibody H5 (18) (data not shown). Next, we analyzed the effects of various glycosaminoglycans on 6B4 proteoglycanpleiotrophin binding. As shown in Fig. 3, heparin potently inhibited the binding of 6B4 proteoglycan to pleiotrophin (IC 50 ϭ 3.5 ng/ml), and heparan sulfate moderately suppressed the binding (IC 50 ϭ 150 ng/ml). Chondroitin sulfate C also had a moderate effect (IC 50 ϭ 400 ng/ml), but in contrast, chondroitin sulfate A scarcely influenced the 6B4 proteoglycan-pleiotrophin binding (IC 50 Ͼ 100 g/ml). However, chondroitinase ABC digests of chondroitin sulfate A and C (unsaturated disaccharides) had no effect on the binding.
Characterization of the Binding of 6B4 Proteoglycan to Pleiotrophin-The roles of chondroitin sulfate chains in the 6B4 proteoglycan-pleiotrophin interaction were tested by digesting 125 I-6B4 proteoglycan with various concentrations of chondroitinase ABC, and then we assayed the binding (Fig.  4A). Digestion with increasing amounts of enzyme reduced the binding to about 60% of the control. Additional heparitinase digestion did not affect the binding (Fig. 4A, open square), which was consistent with the finding that 6B4 proteoglycan does not contain heparan sulfate (2). Scatchard analysis of the binding of native 6B4 proteoglycan to pleiotrophin indicated that there are low (K d ϭ 3.0 nM) and high (K d ϭ 0.25 nM) affinity binding sites (Fig. 4, B and C). We then performed a binding assay using the chondroitinase ABC-digested 125 I-6B4 proteoglycan (Fig. 4B). Scatchard analysis indicated that chondroitinase ABC digestion of the 6B4 proteoglycan reduced the affinity of both sites (K d ϭ 13 nM) without changing the number of binding sites (Fig. 4C).
Presence of 6B4 Proteoglycan and PTP on Cortical Neurons-Dissociated cortical neurons from E16 rats were doubly stained with anti-6B4 proteoglycan antibody (anti-6B4 PG) and anti-phosphorylated neurofilament monoclonal antibody (anti-P-NF). As shown in Fig. 5, P-NF-positive neurons were stained with anti-6B4 PG. The cell bodies and a portion of neurites showed strong immunoreactivity to anti-6B4 PG. It is notable that the rims of the growth cones and the filopodial processes were 6B4 proteoglycan-positive. A polyclonal antibody against the C-terminal portion of 6B4 proteoglycan (anti-31-5) (3) gave similar results.
Immunoprecipitates with anti-6B4 PG were prepared from extracts of cultured cortical neurons and analyzed by immunoblotting with anti-6B4 PG (Fig. 6). Two core protein bands of chondroitin sulfate proteoglycan (300 and 380 kDa) were detected after chondroitinase ABC digestion. Judging from the molecular weight and enzyme activity, the former was 6B4 proteoglycan and the latter was PTP (3). There were two additional minor bands of 230 and 260 kDa, which were not sensitive to chondroitinase ABC digestion. These proteins are probably the precursor form of the proteoglycans in the cells, which are not yet glycosylated with chondroitin sulfate, or dvPTP, which is another splicing variant of PTP (6). Because this culture was estimated to be 98% pure neurons (11), these results suggest that cortical neurons synthesize both 6B4 proteoglycan and PTP.
Effect of Anti-6B4 Proteoglycan Antibody on Pleiotrophininduced Neurite Outgrowth-Pleiotrophin stimulates neurite outgrowth (13,14). As shown in Fig. 7, neurons aggregated without neurites on the uncoated ELISA plates after 20 h. By contrast, cortical neurons actively extended long neurites on the pleiotrophin-coated plates (Fig. 7B). To examine whether 6B4 proteoglycan-pleiotrophin interaction is involved in the neurite outgrowth, anti-6B4 PG was added to the culture medium. Anti-6B4 PG significantly suppressed the pleiotrophininduced neurite outgrowth of cortical neurons at a concentration of over 200 g/ml, in contrast to the control IgG (Fig. 7, C  and D). On the other hand, poly-L-lysine-induced neurite outgrowth was not influenced by anti-6B4 PG, indicating that the effect of the antibody is not a nonspecific cytotoxic activity (Fig.  7D). In fact, the cell viability determined by trypan blue exclusion was not influenced by adding the antibody to the medium (data not shown). However, it was revealed that anti-6B4 proteoglycan (200 g/ml) did not interfere with 6B4 proteoglycanpleiotrophin binding by the solid phase binding assay (data not shown). This result suggests that anti-6B4 proteoglycan does not directly mask the pleiotrophin binding site but affects the signal transfer mechanism of PTP and 6B4 proteoglycan. DISCUSSION This study provides evidence that 6B4 proteoglycan binds pleiotrophin. The 6B4 proteoglycan had high and low affinity binding sites for pleiotrophin. Chondroitinase ABC digestion of FIG. 2. Binding of 6B4 proteoglycan to pleiotrophin. A, wells of ELISA plates were coated with the listed proteins (2 g/ml), and 125 I-6B4 proteoglycan binding to them was measured. 125 I-6B4 proteoglycan bound to pleiotrophin (PTN) and tenascin (TN) but not to acidic fibroblast growth factor (aFGF), fibronectin (FN), laminin (LN), and BSA. B, wells of ELISA plates were coated with 5 g/ml pleiotrophin, and various concentrations of 125 I-6B4 proteoglycan were added to the wells in the presence or absence of 50 nM cold 6B4 proteoglycan. Specific (q) and nonspecific (E) binding was plotted against the 125 I-6B4 proteoglycan concentration. 6B4 proteoglycan reduced the affinity of the proteoglycan to pleiotrophin without changing the number of binding sites. These results suggested that the core protein of 6B4 proteoglycan together with chondroitin sulfate chains constitute the binding sites for pleiotrophin. Among the various glycosaminoglycans, heparin potently inhibited the binding of 6B4 proteoglycan to pleiotrophin, consistent with the fact that pleiotrophin has high affinity for heparin (13,15). N-Syndecan/ syndecan-3 and syndecan-1 were reported to bind pleiotrophin through heparan sulfate chains (17,20). Heparitinase digestion of 6B4 proteoglycan did not affect the binding activity to pleiotrophin, showing that heparan sulfate is not involved in the binding. This is consistent with the finding that 6B4 proteoglycan does not contain heparan sulfate (2).
Our Scatchard analysis indicated that there were two binding sites with different affinities to pleiotrophin. This may be explained by the presence of two populations of 6B4 proteoglycan in the brain bearing chondroitin sulfate chains with different structures. In fact, Rauch et al. (21) have reported that monoclonal antibodies distinguished two subpopulations of phosphacan with different chondroitin sulfate structures. We showed here that chondroitin sulfate C was a potent inhibitor of 6B4 proteoglycan-pleiotrophin binding as well as heparan sulfate. In contrast, chondroitin sulfate A was a poor inhibitor. From this point of view, it is also notable that the structure of chondroitin sulfate chains on 6B4 proteoglycan changes during development of the brain. In the early developmental stages, substantial amounts of chondroitin sulfate C unit are found, but later, the chondroitin sulfate chains of 6B4 proteoglycan are virtually composed of only chondroitin sulfate A units (2,21). This developmental change in the structure of chondroitin sulfate may change the binding affinity of the proteoglycan to pleiotrophin. The contribution of chondroitin sulfate to the protein function is similar in thrombomodulin (22). Digesting thrombomodulin with chondroitinase abolished the inhibitory effect of this molecule on thrombin-induced fibrinogen clotting, indicating that chondroitin sulfate constitutes part of the active site.
Raulo et al. (17) purified N-syndecan from the membrane fractions of the rat brain and from the cultured neurons as a FIG. 4. Contribution of chondroitin sulfate chains to the binding of 6B4 proteoglycan to pleiotrophin. A, 125 I-6B4 proteoglycan was exposed to various concentrations of chondroitinase ABC, and then the binding to pleiotrophin was measured by means of a solid phase assay (q). Chondroitinase ABC digestion reduced the binding of 6B4 proteoglycan to pleiotrophin. The remaining binding after chondroitinase ABC digestion is not likely to be due to the residual unsaturated disaccharide stubs, since unsaturated disaccharides did not influence the binding activity (see Fig. 3). Additional digestion of 125 I-6B4 proteoglycan with heparitinase I and heparitinase II (2 milliunits/ml) did not affect the binding (Ⅺ). Removal of chondroitin sulfates was monitored by SDS-PAGE and autoradiography of the digested samples and was also confirmed by dot blotting using an anti-chondroitin sulfate monoclonal antibody (CS-56). These analyses indicate that chondroitin sulfate was completely removed by 14 milliunits/ml chondroitinase ABC. B, wells of ELISA plates were coated with 2 g/ml pleiotrophin, and the binding of intact (q) or chondroitinase ABC-digested (E) 125 I-6B4 proteoglycan to pleiotrophin was measured by means of a solid phase assay. C, bindings of intact (q) and chondroitinase ABC-digested (E) 125 I-6B4 proteoglycan to pleiotrophin was analyzed by means of a Scatchard plot. Intact proteoglycan has high (K d ϭ 0.25 nM) and low (K d ϭ 3.0 nM) affinity binding sites, but chondroitinase ABC-digested proteoglycan has a single low affinity binding site (K d ϭ 13 nM).
FIG. 5. Cortical neurons were immunohistochemically stained with anti-6B4 proteoglycan antibody. Cortical neurons from E16 rats were cultured overnight on poly-L-lysine-coated coverslips. Cells were fixed and double-stained with anti-6B4 proteoglycan antibody (A) and anti-phosphorylated neurofilament antibody (B). Anti-6B4 proteoglycan antibody stained cell bodies, neurites, and the rims of growth cones (arrow heads). It is notable that some neurofilament-positive neurites were 6B4 proteoglycan-negative (arrows). Scale bar, 10 m.

FIG. 6. Presence of 6B4 proteoglycan and PTP on the cortical neurons.
Cortical neurons from E16 rats were cultured for 50 h and then extracted with a buffer containing Triton X-100 and sodium deoxycholate. The extracts were immunoprecipitated with an anti-6B4 proteoglycan antibody, and the precipitates were analyzed by immunoblotting with anti-6B4 proteoglycan antibody before (Ϫ) and after (ϩ) chondroitinase ABC digestion (A). Control IgG did not precipitate any proteins recognized by anti-6B4 proteoglycan antibody (B). major binding protein for pleiotrophin. The 6B4 proteoglycan is quite soluble, and most of it is recovered in the soluble fraction of the brain extract. Moreover, there is much less PTP than N-syndecan. This may be why Raulo et al. (17) did not detect 6B4 proteoglycan as pleiotrophin-binding protein. They indicated that adding anti-N-syndecan antibody or digesting the neurons with heparitinase suppresses the pleiotrophin-induced neurite outgrowth (17,23), suggesting that N-syndecan is involved in the signal transduction of pleiotrophin. The K d value of the binding of N-syndecan to pleiotrophin is 0.6 nM, which is in a similar range to those of 6B4 proteoglycan (0.25 and 3 nM). The content of pleiotrophin in the developing brain is extremely high, being about 10ϳ15 g/g of wet tissue (24). This is also true for 6B4 proteoglycan or N-syndecan in the brain. About 10ϳ18 g of 6B4 proteoglycan and 10 g of N-syndecan can be purified from 1 g of brain (2,17). Thus, 6B4 proteoglycan and N-syndecan have an apparently equivalent binding capacity for pleiotrophin, and 6B4 proteoglycan and PTP may compose another signal transduction pathway for pleiotrophin.
Immunoelectron microscopy indicated that pleiotrophin is located not only on the cell surface but also in the extracellular matrix, suggesting that pleiotrophin binds to some extracellular matrix components (25). Our studies indicated that 6B4 proteoglycan is localized at extracellular matrix as well as cell surface (10). Thus, 6B4 proteoglycan may serve as an extracellular reservoir of pleiotrophin. In addition to the soluble 6B4 proteoglycan, an alternatively spliced variant PTP is located on the neuronal cell surface as a receptor-like protein-tyrosine phosphatase. Although it is not confirmed whether this molecule binds pleiotrophin, its carbohydrate modification is quite similar to that of 6B4 proteoglycan, including the chondroitin sulfate (3). 2 Therefore, it is conceivable that PTP is a cell surface signal-transducing receptor for pleiotrophin.
The developmental expression of pleiotrophin is dynamically regulated during development of the brain (13,23,(25)(26)(27)(28)(29). In the rat cerebral cortex, the expression of pleiotrophin begins at around E12-E14, peaks at the perinatal period, and persists at a lower concentration into adulthood (23,25). In the early cerebral cortex, pleiotrophin is located along radial glial fibers, a scaffold for neuronal migration, and in the subplate and the marginal zone (23,25). Later, in the early postnatal period, pleiotrophin is more widely expressed in the cerebral cortex including the pathways of the developing axon (23,25). These expression profiles are quite similar to those of 6B4 proteoglycan and PTP (2,7), suggesting that these molecules concertedly function together in the development of the cortex, especially in neuronal migration and in axonal outgrowth.
Increasing evidence indicates that pleiotrophin and 6B4 proteoglycan promote neurite extension (2, 11-13, 23, 24). Pleiotrophin attached to the substratum induces neurite outgrowth from cortical neurons in vitro. 6B4 proteoglycan itself does not promote neurite extension but rather exerts repulsive effects on neurons. However, plates coated with 6B4 proteoglycan together with poly-L-lysine, fibronectin, or tenascin promoted neurite extension from cortical neurons (2,11). The CAH region of PTP/RPTP␤ coated on a substratum promotes neurite extension from tectal neurons (12). Here, we demonstrated that anti-6B4 proteoglycan antibody suppressed the pleiotrophin-induced neurite outgrowth, suggesting that 6B4 proteoglycan and PTP are the functional receptors for pleiotrophin.
The mechanism by which 6B4 proteoglycan regulates neurite outgrowth seems to be rather complex. Milev et al. (30) have reported that complex-type N-linked oligosaccharides mediate the binding of phosphacan to Ng-CAM, N-CAM, and tenascin. The CAH region of phosphacan binds contactin (12). Matrixbound 6B4 proteoglycan/phosphacan may be recognized by neuronal cell surface receptors such as contactin and N-CAM, which trigger the neurite outgrowth. On the other hand, 6B4 proteoglycan and PTP on the neuronal cell surface may bind to matrix-bound ligands such as pleiotrophin, which also leads to neurite outgrowth.
An immunohistochemical analysis of the cultured cortical neurons revealed immunoreactivity to anti-6B4 proteoglycan antibody on the rims of the growth cones and the filopodial processes, which are actin-rich structures that play important roles in axonal path finding and neurite outgrowth. Since this antibody recognizes both 6B4 proteoglycan and PTP (3), it is not clear which protein is responsible for this immunoreactivity. However, our recent analysis of cDNA transformants, stably expressing in the mouse fibroblast L cells, indicated that PTP was localized at cell cortical structures such as membrane ruffles and lamellipodia. 3 The cell cortex is rich in actin filaments and actin-binding proteins and plays important roles in cell locomotion and cell-cell and cell-substratum interaction 2  A-C, cortical neurons from E17 rat embryos were cultured on the Nunc Maxisorp plates. Plates used were uncoated (A) or coated with 2 g/ml pleiotrophin in the absence (B) or presence (C) of 300 g/ml anti-6B4 proteoglycan antibody. Phase-contrast micrographs were taken after 20 h in vitro. Scale bar, 50 m. D, cortical neurons from E17 rat embryos were cultured on Nunc Maxisorp plates coated with 2 g/ml pleiotrophin in the presence of 300 g/ml of control rabbit IgG (PTNϩIgG) or anti-6B4 proteoglycan antibody (PTNϩanti6B4). Neurons were also cultured on the plates coated with 2 g/ml poly-L-lysine in the presence of 300 g/ml of rabbit IgG (PLLϩIgG) or anti-6B4 proteoglycan antibody (PLLϩanti6B4). The average length of neurites extending from 100 neurons was measured after 50 h in vitro. Neurite length from neurons cultured on the plates coated (PTN) or uncoated (NONE) with 2 g/ml pleiotrophin was also measured. This experiment was performed three times, and the results of a typical experiment are shown. (31). We therefore postulate that PTP functions as a receptor that regulates the organization of actin filaments and that the ligand binding to this receptor leads to the reorganization of cytoskeleton in the growth cones and filopodial processes, resulting in neurite outgrowth and cell locomotion. Further studies are required to elucidate the involvement of pleiotrophin and PTP in the regulation of the neuronal cytoskeleton.