Phosphorylation of the Cytoplasmic Tail of Syndecan-4 Regulates Activation of Protein Kinase Cα*

Syndecans are transmembrane proteoglycans capable of carrying both heparan and chondroitin sulfate chains. The cytoplasmic tail of syndecan-4 was recently reported to undergoin vivo phosphorylation on Ser183 in the membrane-proximal part of the tail (Horowitz, A., and Simons, M. (1998)J. Biol. Chem. 273, 10914–10918). However, the functional consequences of this event remain unknown. The cytoplasmic tail of syndecan-4 is known to undergo multimerization and to activate protein kinase Cα (PKCα), with both events depending on the presence of the commonly occurring phospholipid phosphatidylinositol 4,5-bisphosphate (PIP2). In the present investigation we found that phosphorylation of Ser183 produced a 10-fold reduction in the ability of syndecan-4 to activate PKCα, without affecting its ability to bind the PKC. Because Ser183 is adjacent to positively charged lysine groups that resemble PIP2-binding regions in several other proteins, phosphorylation of this serine may affect the binding affinity of the syndecan-4 cytoplasmic tail to PIP2. We found that the Ser183-phosphorylated cytoplasmic tail of syndecan-4 has indeed a significantly lower affinity to PIP2 compared with the nonphosphorylated tail. Furthermore, Ser183phosphorylation abolished PIP2-dependent oligomerization of syndecan-4 cytoplasmic tails. We conclude that Ser183 phosphorylation regulates syndecan-4-dependent activation of PKCα by reducing the affinity to PIP2 and inhibiting the oligomerization of syndecan-4 cytoplasmic tails. These results further support the role of syndecan-4 in signal transduction in endothelial cells.

Heparan sulfate (HS) 1 chains on the cell surface and in the extracellular matrix are present on a specific group of proteins referred to as proteoglycans, which is composed of several classes of core proteins that serve as acceptors for different glycosaminoglycan chains. Heparan sulfate proteoglycans are thought to play an important role in a number of biological processes including regulation of blood coagulation, cell adhesion, and signal transduction. These functions are largely thought to be mediated by HS chains capable of binding growth factors, cell adhesion receptors, and other biologically active molecules (1). Relatively little attention has been paid, however, to the function of the proteoglycan core proteins themselves. Of the three main classes of HS-carrying core proteins (syndecans, glypicans, and perlecan) only the syndecans possess a highly conserved transmembrane domain and a short cytoplasmic tail.
Accumulating evidence suggests that the cytoplasmic tails of several members of the syndecan family can participate in transduction of extracellular signals into the cell. Thus, the cytoplasmic tail of syndecan-3 was reported to bind Src family tyrosine kinases and to mediate neurite growth factor signaling (2). The ubiquitously expressed syndecan-4, which had been reported earlier to be incorporated into focal adhesions upon PKC stimulation (3), was found more recently to bind and regulate the activity of PKC (4 -6). Moreover, we have recently reported that Ser 183 located in the cytoplasmic tail of syndecan-4 becomes phosphorylated upon PKC stimulation, whereas cell treatment with bFGF reduces the phosphorylation of Ser 183 (7). The functional effects of this phosphorylation on the molecular properties of syndecan-4 and on its signaling activity have not yet been elucidated, however.
The regulation of PKC activity by syndecan-4 appears to be mediated by PIP 2 (4,6), which binds directly to the cytoplasmic tail of syndecan-4 (8) and facilitates its multimerization (4). This study addresses, therefore, the effect of phosphorylation of the cytoplasmic tail of syndecan-4 on its interaction with PIP 2 , as well as on its capacities to undergo multimerization and to activate PKC␣.

EXPERIMENTAL PROCEDURES
Materials-PIP 2 , phosphatidylserine (PS), and diolein were purchased from Sigma. Recombinant PKC␣ and PKC␦ were synthesized and prepared as described (9). PKC␤I optimal substrate peptide (FKLKRKGSFKKFA) was purchased from Tufts University Medical School (Boston, MA). A 28-amino acid-long syndecan-4 cytoplasmic tail peptide (S4c) (RMKKKDEGSYDLGKKPIYKKAPTNEFYA) was synthesized by Genemed Synthesis (South San Francisco, CA). A similar peptide with a phosphorylated Ser (S4c-P) was synthesized by the Biopolymers Laboratory, Harvard Medical School (Boston, MA). PIP 2 Binding Assay-PIP 2 (from Sigma, dissolved at 2 mg/ml in 20 parts CHCl 3 , 9 parts MeOH, 1 part H 2 O, 0.1 part 1 N HCl) was dried under N 2 and sonicated for 5 min in ice-cold H 2 O at a final concentration of 1 mg/ml. Syndecan-4 cytoplasmic tail peptides S4c or S4c-P (100 M) were incubated on ice for 30 min with the indicated concentrations of PIP 2 in 10 mM Tris-HCl (pH 7.5), 75 mM KCl, 0.5 mM dithiothreitol in aliquots of 100 l. The samples were layered on 30-kDa molecular mass cut-off cellulose filters (Ultrafree-MC, Millipore, Bedford, MA) and spun at 2000 ϫ g for 1 min following the method described in Ref. 10. The samples (40 l of each in Laemmli sample buffer, 2% SDS, 10% glycerol, 0.5% ␤-mercaptoethanol, 0.004% bromphenol blue, 50 mM Tris-HCl, pH 6.8) were resolved by SDS-PAGE on 16.5% Tris-Tricine gels (Bio-Rad). Gels were stained with Coomassie Brilliant Blue G-250 (Bio-Rad), and * This work was supported in part by National Institutes of Health Grants R01 HL53793 and P50 HL56993 (to M. S.), National Institutes of Health Training Grant HL07374 (to A. H.), and American Heart Association Scientist Development Grant 9730282N (to A. H.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Size Exclusion Chromatography-Syndecan-4 cytoplasmic tail peptides S4c or S4c-P (300 M) were incubated with PIP 2 (350 M, prepared as above) in 0.5 ml 50 mM HEPES (pH 7.3), 150 mM NaCl on ice for 30 min. Samples were applied at 4°C to a Sephadex G-50 (Amersham Pharmacia Biotech) 30 ϫ 1.6 cm column equilibrated with the incubation buffer, and the absorbency of the flow-through was measured at 280 nm.
Electrophoresis, Transfer, and Immunoblotting-Immunoprecipitated syndecan-4 cytoplasmic tail was resuspended in Laemmli sample buffer and resolved by SDS-PAGE on a 4 -20% Tris-glycine gel (Bio-Rad) and transferred for 2 h at 250 mA in 150 mM glycine, 20 mM Tris-HCl, and 20% methanol to a polyvinylidene fluoride (PVDF) membrane (Immobilon-P, Millipore). The membranes were immunoblotted as described (7) using polyclonal antibodies to PKC␣ or to PKC␦ (both at 2 g/ml, purchased from Santa Cruz Biotechnology, Santa Cruz, CA).
Syndecan-4 Cytoplasmic Tail Peptide-PKC Binding Assay-Cytoplasmic tail peptides S4c or S4c-P (10 M) were incubated on ice for 30 min either in the presence or absence of PIP 2 (20 M, prepared as above) with recombinant PKC␣ (4 M) in 0.5 ml of the same buffer used in the PIP 2 binding assay. The cytoplasmic tail peptide was immunoprecipitated, and the samples were resolved by SDS-PAGE, transferred, and immunoblotted as described above.
PKC in Vitro Assays-Samples (30 l) consisted of PKC␤I optimal substrate peptide (100 M) either with or without syndecan-4 cytoplasmic tail peptides S4c or S4c-P (both at 50 M) in 25 mM Tris-HCl (pH 7.4), 5 mM MgCl 2 , 1 mM dithiothreitol, 50 M ATP, and 5 Ci of [␥-32 P]ATP (NEN Life Science Products). In some assays the buffer was supplemented with either PIP 2 (50 M) or PS (4 g/ml), diolein (6.2 g/ml), and 0.2 mM CaCl 2 . In PKC␦ assays the buffer was supplemented with PS and diolein as above and with 0.5 mM EGTA. Upon addition of either PKC␣ (120 ng/ml) or PKC␦ (430 ng/ml), samples were incubated at 30°C for 10 min, and reactions were stopped by boiling in Laemmli sample buffer for 4 min. The samples were resolved on 16.5% Tris-Tricine gels (Bio-Rad), transferred to PVDF membranes, and detected as described (7).

Effect of Syndecan-4 Cytoplasmic Tail Phosphorylation on
Activation of PKC␣-Syndecan-4 cytoplasmic tail has been shown to activate a mixture of Ca 2ϩ -dependent PKCs and of recombinant PKC␣ in the presence of PIP 2 (4,5). To assess the effect of Ser 183 phosphorylation on syndecan-4-dependent PKC activation, we studied the ability of the S4c and S4c-P peptides to activate recombinant PKC␣ using the PKC␤I optimal substrate peptide (9) in an in vitro assay. When the assays were carried out with the standard cPKC cofactors PS, diacylglycerol, and calcium, the presence of neither the S4c nor the S4c-P peptides had any additional effect on the catalytic activity of PKC␣ (Fig. 1). That was also the case in PKC assays where no cofactors were added. However, in the presence of PIP 2 together with the S4c peptide, the catalytic activity of PKC␣ toward the PKC␤I peptide was approximately 10-fold larger than in assays with PIP 2 alone. On the other hand, when the S4c-P peptide was added instead of S4c, the phosphorylation level of the substrate was similar to that obtained with PIP 2 alone. Unlike PKC␣, the S4c peptide did not activate PKC␦ under the same conditions (data not shown). The activity of PKC␣ in the presence of the S4c peptide and PIP 2 was 72 Ϯ 10% (ϮS.D., n ϭ 3) of its activity in the presence of the S4c peptide, PS, diacylglycerol, and calcium.
Binding of PKC␣ to the Cytoplasmic Tail of Syndecan-4 -The ability of the unphosphorylated but not the phosphorylated cytoplasmic tail of syndecan-4 to activate PKC␣ in vitro may relate to a reduced PKC␣ affinity upon phosphorylation of the cytoplasmic tail. Previous studies have demonstrated the ability of the cytoplasmic tail of syndecan-4 to bind PKC (5), and the identity of the bound PKC isozyme in vivo was narrowed down to a group of four (␣, ␤I, ␤II, ␥, and ␦ (5)). Although PKC␣ was shown to bind to the cytoplasmic tail of syndecan-4 in vitro (5), the cytoplasmic tail could also bind to and be a substrate of PKC␦ (7). To determine the ability of syndecan-4 to bind PKC␣ or ␦ in vivo, RFPEC lysates were immunoprecipitated with an antiserum specific to the cytoplasmic tail, and the immunoprecipitants were probed with antibodies specific either to the ␣ or ␦ PKC isozymes. The presence of PKC␣ but not of PKC␦ was detected in the immunoprecipitants (Fig. 2, A and B).
To analyze the effect of syndecan-4 cytoplasmic tail phosphorylation on its ability to bind PKC␣, we carried out in vitro assays with recombinant PKC␣ and the S4c and S4c-P peptides. Incubation of PKC␣ with either peptide produced, however, similar degrees of binding both in the presence and absence of PIP 2 (Fig. 2C). It follows, therefore, that PKC␣ binding is not affected by the Ser 183 phosphorylation in the syndecan-4 cytoplasmic tail and thus cannot explain the effect of syndecan-4 phosphorylation on the activity of the enzyme.
Effect of Phosphorylation on PIP 2 Binding to the Cytoplasmic Tail of Syndecan-4 -Both the oligomerization and PKC␣ activation capacities of the cytoplasmic tail of syndecan-4 were found to depend on the presence of PIP 2 (4,8). It was of interest, therefore, to determine whether the phosphorylation of Ser 183 in the cytoplasmic tail of syndecan-4 affects the affinity of the tail to PIP 2 . To this end, we compared the in vitro binding between PIP 2 micelles and S4c or S4c-P peptides using a filtration assay. The filter retains the PIP 2 micelle-bound peptide, whereas the unbound peptide passes through it. The binding affinity of the S4c peptide to PIP 2 , as determined by band densitometry of the SDS-PAGE-resolved filter flowthrough samples, was significantly higher than that of the S4c-P peptide. At a peptide:PIP 2 molar ratio of 2:1, 50% of the S4c peptide that passed through the filter in the absence of PIP 2 was retained versus none of the S4c-P peptide (Fig. 3). Practically all the applied S4c peptide was retained by the filter at a peptide:PIP 2 molar ratio of 1:2, whereas as much as 50% of

Syndecan-4 Phosphorylation and PKC␣
the S4c-P peptide still passed through the filter under the same conditions. The apparent dissociation constants (K d ) calculated from the results shown in Fig. 3 were 2.4 M for the nonphosphorylated peptide (S4c), versus 232 M for the phosphorylated one (S4c-P). Thus, Ser 183 phosphorylation results in significant reduction in the ability of PIP 2 to bind to the cytoplasmic tail of syndecan-4.

Effect of Phosphorylation on Syndecan-4 Cytoplasmic Tail
Oligomerization-Previous studies have demonstrated that the cytoplasmic tail of syndecan-4 undergoes oligomerization in the presence of PIP 2 (4,8); furthermore this oligomerization appeared necessary for PKC␣ activation. The reduced affinity between the cytoplasmic tail and PIP 2 caused by phosphorylation could conceivably be accompanied by changes in the oligomerization properties of syndecan-4. To compare the oligomerization of the S4c peptide to that of the phosphorylated peptide S4c-P, both were incubated either in the presence or absence of PIP 2 as described under "Experimental Procedures" and passed through a size exclusion column. Both peptides eluted as a single peak when incubated in the absence of PIP 2 (Fig. 4, A and B). When incubated in the presence of PIP 2 , the S4c peptide eluted as two peaks, one of an approximate molecular mass of 7 kDa (Fig. 4C), and another heavier peak of a molecular mass greater than 17 kDa (the molecular mass of the heaviest molecular mass standard used in this experiment). The S4c-P peptide, on the other hand, eluted as a single peak of the same approximate molecular mass as the first peak of the S4c peptide (Fig. 4D). These results indicate that the cytoplasmic tail of syndecan-4 looses its capacity to form oligomers upon phosphorylation of Ser 183 . Based on the position of the first peaks of the S4c and the S4c-P peptides, it appears that both the S4c (as previously reported (4)) and the S4c-P peptides formed dimers, similar to the behavior observed in the PIP 2binding experiment (Fig. 3). The broader peaks observed with both peptides when incubated in the presence of PIP 2 , compared with the sharper ones obtained in the absence of PIP 2 , reflect a wider spread in molecular mass, probably resulting from the range of PIP 2 binding to the peptides. DISCUSSION This study presents three distinct findings concerning the role of the syndecan-4 core protein in signal transduction: (a) Phosphorylation of a single serine residue (Ser 183 ) located in the membrane-proximal part of the cytoplasmic tail of syndecan-4 reduces the affinity of the tail to the phosphoinositide PIP 2 . Upon phosphorylation, the cytoplasmic tail loses its capacity to (b) undergo multimerization and (c) activate PKC␣ in the presence of PIP 2 . These findings provide the first evidence for a functional role of the recently reported (7) phosphorylation of Ser 183 in the cytoplasmic tail of syndecan-4.
The capacities of the cytoplasmic tail of syndecan-4 to undergo multimerization and to activate PKC␣ were manifest only in the presence of PIP 2 . A recent NMR study reported on PIP 2 binding to a lysine-rich 9-amino acid-long variable region (LGKKPIYKK) from the middle of the cytoplasmic tail of syn- decan-4 (8). The interaction of PIP 2 with this region could conceivably be mediated in part by electrostatic bonds between the two phosphates attached to the PIP 2 inositol ring and the positively charged lysines of the variable region. This electrostatic interaction could be disrupted by the phosphorylation of Ser 183 located 3 residues upstream of the variable region. An interaction of a similar nature may take place between PIP 2 and several basic residues located in the N-terminal actinbinding domain of ␣-actinin (13). The specific binding of PIP 2 to the pleckstrin homology domains of several proteins also appears to be mediated by an electrostatic interaction between the PIP 2 phosphates and positively charged lysines in the pleckstrin homology domain (14). Moreover, the syndecan-4 variable region resembles the consensus sequences for PIP 2binding motifs (15). A reduction in PIP 2 binding following phosphorylation of serine residues, similar to the one we reported here, is thought to occur in a lysine-rich sequence of the myristoylated alanine-rich protein kinase C substrate (16).
The mechanism of PKC␣ activation by the cytoplasmic tail of syndecan-4 is thought to require formation of cytoplasmic tail multimers (4). Ser 183 phosphorylation prevents this oligomerization by inhibiting PIP 2 binding to the variable region of the syndecan-4 cytoplasmic tail. It follows, therefore, that the loss of PKC␣ activation by the cytoplasmic tail upon phosphorylation of Ser 183 is a direct consequence of the concomitant reduction in affinity to PIP 2 and impaired multimerization. Because the cytoplasmic tail of syndecan-4 did not activate PKC␦, this activation may be specific to PKC␣. On the other hand, Ser 183 phosphorylation had no effect on the capacity of the syndecan-4 cytoplasmic tail to bind PKC␣. The ability of syndecan-4 to activate PKC␣ signaling in endothelial cells, the regulation of this signaling by syndecan-4 phosphorylation, and the previously demonstrated bFGF-dependent regulation of the state of syndecan-4 cytoplasmic tail phosphorylation suggests the existence of a novel bFGF-dependent signaling pathway. The same signaling pathway may also be employed by other events affecting syndecan-4 phosphorylation.