| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
J. Biol. Chem., Vol. 281, Issue 25, 17061-17068, June 23, 2006
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
2-Macroglobulin Abrogate Binding of Platelet-derived Growth Factor-BB and Transforming Growth Factor-
1*
1
From the
Department of Pathology, University of California San Diego, La Jolla, California 92093 and Department of Pathology, University of Virginia, Charlottesville, Virginia 22908
Received for publication, March 9, 2006
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
2-Macroglobulin (2M) inhibits diverse extracellular proteases, binds growth factors such as platelet-derived growth factor-BB (PDGF-BB) and transforming growth factor-
1 (TGF-1), and carries -amyloid peptide. 2M may also trigger cell signaling by binding to the low density lipoprotein receptor-related protein (LRP-1) and/or other cell surface receptors. Based on studies with recombinant 2M fragments expressed in bacteria and synthetic peptides, we previously localized a growth factor-binding site near the center of the 2M subunit. However, because intact 2M forms a hollow cylinder structure, an alternative model for growth factor binding involves nonspecific entrapment within the 2M core. To distinguish between these two models, we engineered mutations in the putative growth factor binding sequence of full-length 2M. These mutations did not perturb the tetrameric structure of 2M, reaction with proteases, the thiol ester bonds, or binding to LRP-1. A single mutation (E730R) completely blocked binding of platelet-derived growth factor-BB to intact 2M. E730R did not alter TGF-1 binding; however, this mutation in combination with mutations at Glu714 and Asp719 eliminated the increase in TGF-1 binding associated with 2M conformational change. These studies demonstrate that growth factor binding to intact 2M is specific, involving a defined region of the 2M subunit. The exact sequences required for binding different growth factors may be non-identical, mimicking the model of the bait region in which different proteases target adjacent and sometimes overlapping sequences. |
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
2-Macroglobulin (2M)2 is a 718-kDa homotetrameric glycoprotein that functions as a protease inhibitor and as a carrier of growth factors in the blood and in other extracellular spaces (1). Proteases react with 2M by cleaving any of a number of target peptide bonds near the center of the 2M subunit, in the "bait region" (2). Cleavage in the bait region induces a major conformational change in 2M that entraps the reacting protease within the hollow core of the cylindrical 2M structure (2–4). 2M conformational change reveals the recognition site for the 2M receptor/low density lipoprotein receptor-related protein (LRP-1), which is cryptic in the native 2M conformation (5, 6). For this reason, conformationally transformed 2M is referred to as "activated." 2M thiol ester bonds, which are formed by the side chains of Cys949 and Glu952, also become exposed during 2M conformational change and may react with lysine residues of the attacking protease to form covalent linkages; however, these linkages are not necessary to stabilize the 2M-protease complex because the trapping mechanism is essentially irreversible under physiological conditions (7, 8). 2M conformational change may be induced in the absence of proteases by aminolysis of the thiol ester bonds with small primary amines such as methylamine (MA) (9, 10).
One growth factor reported to interact with 2M is platelet-derived growth factor (PDGF), a potent stimulator of mesenchymal cell mitogenesis and chemotaxis and a regulator of inflammation and wound healing in vivo (11). Biologically active PDGF is formed by homodimerization of A-chains or B-chains or by heterodimerization of the A- and B-chains to form PDGF-AB (11). Recently, PDGF C-chains and D-chains also have been described (12). The PDGFs mediate biological activities by binding to the PDGF -receptor or -receptor; the specificity of this interaction is dictated by the subunit composition of the PDGF (12). The interaction of PDGF with 2M was first discovered by analysis of PDGF carrier proteins in the plasma (13, 14). Subsequent studies demonstrated binding of purified PDGF-BB and PDGF-AB to 2M, but not PDGF-AA (15). The interaction of PDGF-BB with 2M blocks PDGF-BB binding to its receptor (16, 17).
TGF- binding to 2M was also originally discovered by analysis of TGF- carrier proteins in the plasma (18). Because TGF- in complex with 2M lacks biological activity, the complex was originally referred to as "latent complex," a term now reserved for the secreted complex in which TGF- retains its pro-region, the latency-associated peptide, and binds latent TGF--binding protein (19). Like PDGF, growth factors in the TGF- family regulate inflammation and the response to injury (20). These factors are also involved in development (20). The latent TGF- complex, which is released from cells, may be dissociated by proteases, chaotropic agents, or decreases in pH, yielding active growth factor (21). Because liberated active TGF- associates rapidly with native 2M, which is not LRP-1 recognized, the TGF- is stabilized against renal clearance and may be reversibly released at sites of tissue remodeling or inflammation (18, 22).
Binding of TGF-1 and PDGF-BB is regulated by 2M conformational change; both growth factors bind with higher affinity to 2M that is activated by methylamine or by proteases (22, 23). When native 2M and various forms of activated 2M are denatured, TGF-1 binding becomes equivalent, confirming the role of 2M conformation in determining growth factor binding affinity (17, 24, 25). The ability of 2Mto bind growth factors after denaturation also suggests that linear sequences in the structure of 2M may be responsible for the observed interaction.
To identify candidate growth factor-binding sites in the structure of 2M, we screened a library of overlapping glutathione S-transferase (GST) fusion proteins and defined a binding site for TGF-1 between amino acids 700 and 738 in the mature 2M subunit (24, 25). GST fusion proteins containing the same sequence also bound PDGF-BB and nerve growth factor-, suggesting that these binding sites may be identical or overlapping (17). In subsequent studies, we screened synthetic peptides and further narrowed a putative binding site for TGF- and PDGF-BB to between amino acids 714 and 733 (25, 26). The growth factor binding peptides were notable for their high content of hydrophobic and negatively charged residues. Importantly, the peptides blocked binding of TGF-1 to its receptors, suggesting that the peptides and receptors bind to equivalent or overlapping sites on the growth factor surface (26).
Genetic engineering studies with full-length recombinant 2M (r2M) have contributed considerably to our understanding of the structure and function of this protein. When Cys949 is mutated to serine, precluding formation of thiol esters, 2M is expressed in its conformationally transformed state (27). Mutation of Asn1065 also alters thiol ester formation and its reaction with nucleophiles (28). Truncation of the 2M bait region yields 2M variants incapable of protease entrapment and induces defects in subunit association (29). Finally, mutagenesis studies targeting the LRP-1 recognition sequence demonstrated an essential role for Lys1370 but not Lys1374 (30).
Although work with GST fusion proteins and synthetic peptides provided evidence regarding the potential location of a growth factor binding sequence in the structure of 2M, a pitfall in the use of this strategy involved the possibility that the discovered sequence is cryptic in full-length 2M. This concern was heightened by the highly hydrophobic nature of the identified sequence. Given what is known about the structure of MA-modified 2M, which approximates the shape of a hollow cylinder with a substantial internal cavity, an alternative hypothesis regarding growth factor binding might involve nonspecific entrapment within the 2M central core. Thus, to validate the results of our experiments with GST fusion proteins and synthetic peptides, it was necessary to reproduce the results in intact 2M.
In this study, we expressed full-length r2M and three mutated forms of r2M in which acidic amino acids in the putative growth factor-binding site were subjected to charge reversal. We chose charge reversal as opposed to the more conventional approach of alanine mutagenesis to avoid possible collapse of what we hypothesized was already a highly hydrophobic surface patch in the structure of 2M. The mutated forms of r2M included r2M(E730R), r2M(E714R,D719R), and r2M(E714R,D719R,E730R). We have demonstrated that PDGF-BB binding is entirely eliminated in r2M(E730R). TGF-1 binding is not totally blocked by any of these three mutations; however, the substantial increase in TGF-1 binding that accompanies conversion of 2M to the MA-transformed conformation is eliminated in r2M(E714R,D719R,E730R). These studies have demonstrated that growth factor binding to 2M reflects a specific interaction involving defined amino acids within the 2M structure. Binding of different growth factors may involve non-identical but overlapping sites.
|
EXPERIMENTAL PROCEDURES |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
1 and PDGF-BB were from R&D Systems (Minneapolis, MN). Methylamine HCl (MA), dithiothreitol, Earles balanced salt solution, and bovine serum albumin (BSA) were from Sigma. Na125I was from MP Biochemicals (Aurora, OH). Polyvinylidene fluoride membranes were from Millipore Corp. (Bedford, MA). The QuikChange® XL site-directed mutagenesis kit was from Stratagene (La Jolla, CA). Bis(sulfosuccinimidyl) suberate (BS3) was purchased from Pierce (Rockford, IL). Dulbecco's modified Eagle's medium, Iscove's modified Dulbecco's medium, RPMI 1640, and trypsin-EDTA were from Invitrogen. Trypsin was purchased from Worthington (Lakewood, NJ). The concentration of trypsin was determined by active site titration with p-nitrophenyl p'-guanidino benzoate (31). Trypsin was radioiodinated with IODO-BEADS (Pierce) according to the manufacturer's instructions. 125I-Trypsin was collected by gel filtration into 1 mM HCl for storage. TGF-1 and PDGF-BB were radioiodinated according to the procedure of Ruff and Rizzino (32). 125I-TGF-1 and 125I-PDGF-BB were subjected to SDS-PAGE and autoradiography to confirm integrity and stored at 4 °C for no longer than 2 weeks. The specific activity of 125I-TGF-1 was 100–200 µCi/µg. The specific activity of 125I-PDGF-BB was 3–5 µCi/µg.
Human 2 M—2M was purified from outdated human plasma according to the method of Imber and Pizzo (33). The purity of 2M was confirmed by SDS-PAGE and the concentration determined by the absorbance at 280 nm using an A1%, 1.0 cm of 8.93 (34). To form MA-activated 2M, native 2M was dialyzed against 200 mM MA in 50 mm Tris-Cl, pH 8.2, overnight at 22 °C, followed by extensive dialysis against 20 mM sodium phosphate, 150 mM NaCl, pH 7.4 (PBS) at 4 °C. Reaction of native 2M with MA was confirmed by the characteristic increase in electrophoretic mobility in nondenaturing PAGE (9, 10).
The full-length human 2M cDNA in pAT153 (24) was cloned into the expression vector, pSecTag2/Hygro, from Invitrogen. Site-directed mutagenesis was performed using the QuikChange® XL site-directed mutagenesis kit according to the manufacturer's instructions. Mutated forms of r2M, including r2M(E730R); r2M(E714R,D719R), also called r2M(RR); and r2M(E714R,D719R,E730R), also called r2M(RRR), were confirmed by sequence analysis.
K-562 cells were obtained from the American Type Culture Collection (ATCC, Manassas, VA) and maintained in Iscove's modified Dulbecco's medium with 10% heat-inactivated fetal bovine serum, 4 mM L-glutamine, and 1% penicillin/streptomycin. K-562 cells do not express detectable levels of endogenous 2M, as previously described (30). Constructs encoding wild-type r2M and the various mutated forms of r2M were transfected by electroporation (270 V, 1180 µF) into 5 x 106 K-562 cells. The cells were selected in 200 µg/ml of Hygromycin-B (Invitrogen) and single cell cloned to identify high expressing cells. R2M expression and secretion into serum-free medium were determined by immunoblot analysis, using polyclonal rabbit anti-human 2M IgG from DAKO (Carpinteria, CA). r2Ms were purified from conditioned medium by the method of Imber and Pizzo (33) and analyzed as described for plasma 2M (34). MA-activated r2Ms were radioiodinated using IODO-BEADS. The specific activity of 125I-r2M(RRR) was 1–5 µCi/µg.
Analysis of r2M Structural Integrity and Conformational Change—In plasma-derived 2M, the four subunits are linked into pairs by disulfide bonds and into intact tetramers by noncovalent interactions (3). To assess subunit integrity and the pattern of disulfide bonds, wild-type and mutated forms of r2M were subjected to SDS-PAGE. In some experiments, r2M variants were treated with a 2-fold molar excess of active trypsin for 2 min. Reactions were terminated with p-nitrophenyl p'-guanidino benzoate and the products analyzed by SDS-PAGE. In plasma-derived 2M, trypsin cleaves only the bait region, converting the 180-kDa subunits into 90-kDa products (3). To assess r2M conformational change in response to trypsin or MA, samples were subjected to nondenaturing PAGE on 5% slabs, according to the method of Van Leuven et al. (35). Proteases and MA induce nearly identical conformational change in plasma 2M (9, 10). Direct binding of trypsin to plasma-derived 2M, wild-type r2M, and mutated forms of r2M was determined using 125I-trypsin. Each 2M derivative was incubated with a 2-fold molar excess of 125I-trypsin for 2 min at 22 °C. Samples were subjected to nondenaturing PAGE and autoradiography to detect radio-iodinated trypsin co-migrating with 2M. In control experiments, we demonstrated that free 125I-trypsin does not migrate near the mobility of 2M.
2 M Receptor Binding Experiments—RAW 264.7 cells were obtained from the ATCC and maintained in RPMI 1640 medium supplemented with 10% fetal bovine serum and 1% penicillin/streptomycin at 37 °C, 5% CO2, and 95% humidity. For receptor binding experiments, cells were plated at 5 x 105/well in 24-well plates and cultured until nearly confluent. The cells were then washed twice with Earles balanced salt solution containing 10 mM HEPES, pH 7.4, and 1.0 mg/ml BSA (EHB) and equilibrated in the same solution.
MA-activated 125I-r2M(RRR) (0.1–5.0 nM) was incubated with cells for 4 h at 4 °C, alone or in the presence of nonradiolabeled MA-activated plasma-derived 2M (0.2 µM). The wells were washed two times with EHB and once with Earles balanced salt solution. Cell extracts were prepared in 0.1 M NaOH, 1.0% SDS. Cell-associated radioactivity was determined in a 1470 Wizard Gamma Counter (PerkinElmer). Cellular protein was measured by bicinchoninic acid assay (Sigma). Specific binding was then determined as the fraction of total binding that was inhibited by the nonradiolabeled MA-activated plasma 2M (typically 75–95%). Each experiment was performed at least in triplicate. Binding data were analyzed by nonlinear regression, comparing single-site and two-site models.
Analysis of PDGF-BB Binding to 2 M—125I-PDGF-BB (5 nM) was incubated with native or MA-activated plasma 2M, wild-type r2M, or mutated forms of r2M (0.1 µM) in PBS containing 15 µM BSA for 60 min at 37 °C. Samples were subjected to nondenaturing PAGE for 1 h at 150 V. 125I-PDGF-BB binding to 2M was determined by autoradiography. In control experiments, we demonstrated that free 125I-PDGF-BB does not migrate near 2M.
We also analyzed 125I-PDGF-BB binding to 2M by co-immunoprecipitation. Various concentrations of MA-activated plasma 2M or r2M(E730R) were coupled to rabbit anti-human 2M polyclonal antibody immobilized on Protein A-Sepharose (Amersham Biosciences). The coupling time was 1 h at 22°C. The beads were washed and then incubated with 125I-PDGF-BB (1 nM) for 1 h at 37°C in PBS with 1 mg/ml of BSA. The beads were washed three times again, and the amount of 125I-PDGF-BB associated with the beads was determined by measuring radioactivity in a 
counter. Some of the beads were treated with SDS sample buffer under reducing conditions to dissociate immunoprecipitated proteins. Samples were subjected to SDS-PAGE, and the gels were Coomassie stained and dried. Autoradiography was performed to detect 125I-PDGF-BB. As a control, 125I-PDGF-BB was incubated with 2M-specific antibody coupled to Protein A-Sepharose in the absence of 2M. Nonspecific binding was subtracted as background.
As a third method to study 125I-PDGF-BB binding to 2M, we applied a cross-linking strategy. MA-activated plasma 2M and r2M(E730R) (0.1 µM) were incubated with 125I-PDGF-BB for 1 h at 37°C in PBS with 1 mg/ml of BSA. The samples were then divided in two. One aliquot was denatured in SDS without further modification, so that only covalent interactions were preserved. The other aliquot was treated with the homobifunctional cross-linker BS3 (5 mM) for 1 min. Pulse exposure to BS3 covalently stabilizes a fraction of the non-covalent PDGF-BB-2M complex. Because BS3 is reacted with the protein mixture under pseudo first-order conditions, the fraction of covalently stabilized 2M-PDGF-BB complex is constant and independent of the total amount of non-covalent complex (22). Cross-linking was terminated by acidification with HCl and immediate addition of buffered SDS (2.0% v/v). Samples were subjected to 4–15% gradient SDS-PAGE under non-reducing conditions. The gels were stained with Coomassie Blue and dried. Autoradiography was performed to detect 125I-PDGF-BB in association with 2 M.
PDGF Receptor Phosphorylation—Murine embryonic fibroblasts from wild-type C57/BL6 mice were seeded in 12-well tissue culture plates in Dulbecco's modified Eagle's medium containing 10% fetal bovine serum and cultured until nearly confluent. The cultures were washed three times with PBS and then serum starved for 18 h. Cells were treated for 10 min at 37 °C with PDGF-BB (2 ng/ml), plasma 2M (1 µM), r2M(E730R) (1 µM), PDGF-BB plus plasma r2M, or PDGF-BB plus r2M(E730R). The 2M was MA activated. When PDGF-BB and the various forms of 2M were added to the same cultures, the two proteins were preincubated for 15 min at 37 °C. Treated cells were washed with ice-cold PBS and extracted in radioimmune precipitation assay buffer. The extracts were subjected to immunoblot analysis to detect tyrosine-phosphorylated proteins using the mouse monoclonal anti-phosphotyrosine antibody from BD Biosciences. As a control for load, extracts were immunoblotted for total ERK1/2 using the rabbit polyclonal anti-ERK1/2 antibody from Upstate (Charlottesville, VA).
Analysis of TGF-1 Binding to 2 M—125I-TGF-1(1nM) was incubated with native or MA-activated plasma 2M, wild-type r2M, and mutated forms of r2M (0.1 µM) in PBS containing 15 µM BSA for 60 min at 37 °C. Samples were subjected to nondenaturing PAGE for 1 h at 150 V. 125I-TGF-1 binding to 2M was determined by autoradiography. In control experiments, we demonstrated that free 125I-TGF-1 does not migrate near 2M.
TGF-1 binding to 2M in solution was also determined by competition for binding to immobilized MA-activated plasma-derived 2Mas previously described (36). Plasma 2M (1.0 µgin100 µl) was incubated in each well of 96-well microtiter plates for 4 h at 22 °C. This procedure results in immobilization of 
90 fmol 2M (36). The wells were washed three times with PBS, 0.1% Tween 20 (PBST) and blocked with PBST for 16 h at 4 °C. As a control, some wells were blocked with PBST without first immobilizing 2M. 125I-TGF-1 (0.1 nM) was incubated with the immobilized 2M in the presence of 15 µM BSA and 250 nM native or MA-activated plasma 2M or r2M (RRR) for 1 h at 22°C. The wells were then washed three times with PBST. 125I-TGF-1 that was associated with the immobilized phase was recovered in 0.1 M NaOH, 2% SDS and quantitated in a -counter.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
2 M—We previously localized a specific region in the 2M subunit that includes putative binding sites for several growth factors, including TGF-1, PDGF-BB, and nerve growth factor- (24, 25). We refer to this site as "protein interaction domain 1" (PID-1); however, the strict definition of "domain" may not apply. Fig. 1 shows the relationship of PID-1, in the primary sequence of the 2M subunit, to other functional domains. The second non-covalent protein interaction domain (PID-2) serves as a binding site for -amyloid peptide (37).
Mutation of PID-1 in the r2M cDNA was executed according to the results of our previous studies with GST fusion proteins and synthetic peptides (26). The three mutated r2M variants generated are shown in Fig. 1. All of the r2Ms were subjected to SDS-PAGE under reducing and non-reducing conditions. In the presence of reductant, the intact 180-kDa 2M subunit was consistently observed (Fig. 2A). In the absence of reductant, disulfide-linked subunit dimers (Mr 360,000) were observed, indicating that the mutations did not disrupt the normal pattern of inter-subunit disulfide bonding.
Plasma-derived 2M undergoes conformational change when reacted with proteases or MA (9, 10), and this conformational change is readily observed as an increase in electrophoretic mobility by nondenaturing PAGE (35). As shown in Fig. 2B, the native forms of plasma-derived 2M and wild-type r2M demonstrated equivalent mobility in nondenaturing PAGE and equivalent increases in mobility when treated with MA. R2M(E730R) demonstrated slightly slower mobility, as anticipated given the replacement of an acidic amino acid with a basic amino acid (Fig. 2B). However, r2M(E730R) demonstrated increased electrophoretic mobility when treated with MA or trypsin. Mobility in nondenaturing PAGE is determined by charge, molecular mass, and conformation. The mobility of r2M(E730R) is consistent with conservation of tetrameric structure. The increase in electrophoretic mobility in response to MA suggests that the thiol ester bonds are intact. The increase in electrophoretic mobility in response to trypsin suggests that the bait region is available for cleavage by trypsin. Like r2M (E730R), both r2M(RR) and r2M(RRR) demonstrated slightly decreased mobility by nondenaturing PAGE (Fig. 2C). Additionally, both proteins underwent conformational change, as judged by a shift in mobility, when treated with MA.
|
|
2M(E730R) and r2M(RRR) were treated with trypsin and subjected to SDS-PAGE under reducing conditions. In both cases, the r2M subunits were cleaved only in the bait region, yielding the anticipated 90-kDa fragments (Fig. 3A). Additional trypsin cleavage sites were not observed. To confirm that the mutated r2Ms not only are cleaved by protease but also retain protease binding activity, we incubated plasma 2M, r2M, and r2M(RRR) with a 2-fold molar excess of 125I-trypsin. The identical incubation was conducted after treatment with MA. The preparations were subjected to nondenaturing PAGE and autoradiography to detect 2M-associated 125I-trypsin. As shown in Fig. 3B, native plasma-derived 2M and wild-type r2M bound 125I-trypsin but not after reaction with MA. This result was anticipated because MA-induced 2M conformational change markedly inhibits reaction with proteases (3). Identical results were obtained with r2M (RRR). These results confirm that the mutated r2Ms function as protease inhibitors without perturbation. Furthermore, the thiol ester bonds in these proteins react with MA as anticipated.
As a final test of the integrity of the structure and function of the mutated r2Ms, we examined binding of r2M(RRR) to receptors in RAW 264.7 cells. In these cells, the only detectable 2M receptor is LRP-1 (38). Binding of activated 2M to LRP-1 is conserved across species lines (39). MA-activated r2M(RRR) was radiolabeled and incubated with the RAW 264.7 cells for 4 h at 4°C. Specific binding was determined as the fraction of total binding that was inhibited by a large molar excess of nonradiolabeled MA-activated plasma-derived 2M. A representative binding isotherm is shown in Fig. 3C. The KD and Bmax were determined from the results of three separate experiments. The KD was 0.25 ± 0.05 nM, which is in excellent agreement with values determined previously for MA-activated plasma 125I-2M (33, 35, 38) and wild-type 125I-r2M (30). The Bmax was 220 ± 10 fmol/mg of cell protein. These results confirm that the triple mutation in the PID-1 region did not affect the LRP-1 recognition sequence in 2M.
|
2M and wild-type r2M (Fig. 4A). The samples were subjected to non-denaturing PAGE. Binding of 125I-PDGF-BB was determined by autoradiography. Plasma-derived native 2M bound low levels of PDGF-BB and greatly increased amounts of PDGF-BB after MA-induced conformational change (Fig. 4A), as anticipated given the substantially higher binding affinity of MA-activated 2M for PDGF-BB (22). Equivalent results were obtained with wild-type r2M. Once again, significantly increased 125I-PDGF-BB binding was observed following reaction with MA.
Next, we compared binding of 125I-PDGF-BB to plasma-derived 2M and r2M(E730R). As determined by nondenaturing PAGE and autoradiography, mutation of a single amino acid, Glu730, was sufficient to entirely block 125I-PDGF-BB binding to r2M (Fig. 4B). The Coomassie-stained gel in Fig. 4B confirms that the absence of PDGF-BB binding was not due to underloading of the gel. By contrast, mutation of Glu714 and Asp719, in r2M(RR), failed to significantly affect PDGF-BB binding (Fig. 4C). Binding of 125I-PDGF-BB to r2M(RRR), which includes mutation of Glu730, was not detected, consistent with the results obtained with r2M(E730R).
|
2M(E730R), we immobilized MA-activated plasma 2M and r2M(E730R) on 2M-specific antibody coupled to Protein A-Sepharose. 125I-PDGF-BB (1 nM) was incubated for 1 h with the beads or, as a control, with beads that were not preloaded with 2M. Binding was determined by precipitation of radioactivity. 125I-PDGF-BB precipitated with plasma 2M, and the amount of precipitated radioactivity increased proportionately to the amount of immobilized 2M present (Fig. 5A). By contrast, precipitated radioactivity was extremely limited with MA-activated r2M(E730R).
Precipitated proteins were eluted from the beads with SDS and dithiothreitol and subjected to SDS-PAGE. Co-precipitation of 125I-PDGF-BB with MA-activated plasma 2M was confirmed by autoradiography (Fig. 5B). In samples from beads loaded with MA-activated r2M(E730R), almost no 125I-PDGF-BB was detected. Approximately even loading of the beads with plasma 2M and r2M(E730R) was confirmed by Coomassie staining of the eluates.
|
2M, we utilized the method of Crookston et al. (22), in which pulse-exposure to BS3 is used to stabilize non-covalent complex prior to SDS-PAGE analysis. The cross-linking reaction occurs under pseudo-first order conditions, and the amount of cross-linked complex is proportional to the total amount of non-covalent complex available. Fig. 6 compares the binding of 125I-PDGF-BB to MA-activated plasma 2M and r2M(E730R). In the absence of BS3, a low level of SDS-stable 125I-PDGF-BB plasma 2M complex was observed, probably representing covalent stabilization due to thiol-disulfide exchange (22). Pulse-exposure to BS3 increased the amount of 125I-PDGF-BB associated with plasma 2M substantially. By contrast, 125I-PDGF-BB binding to r2M(E730R) was not observed with or without BS3.
Effects of 2M on PDGF-BB-induced Receptor Phosphorylation—In murine embryonic fibroblasts, PDGF-BB binds to the PDGF -receptor and induces transient tyrosine phosphorylation of the receptor, which is the primary event detected by immunoblot analysis with pan-phosphotyrosine-specific antibody (40). We previously demonstrated that MA-activated plasma 2M binds PDGF-BB and thereby prevents binding of the growth factor to a purified PDGF -receptor fusion protein (17). In new studies, we compared the ability of plasma 2M and r2M (E730R) to inhibit PDGF -receptor tyrosine phosphorylation in response to PDGF-BB. Murine embryonic fibroblasts were treated with PDGF-BB for 10 min, alone or in the presence of MA-activated plasma 2M or r2M(E730R). Cell extracts were subjected to immunoblot analysis to detect tyrosine phosphorylation and, as a control for load, total ERK1/2 (Fig. 7).
|
|
2M, PDGF-BB induced tyrosine phosphorylation of a major band with an apparent mass of 190 kDa, consistent with the known mass of mature PDGF -receptor (Fig. 6). MA-activated plasma 2M and r2M(E730R) failed to independently alter the pattern of immunoreactivity of phosphotyrosine-specific antibody. When PDGF-BB was preincubated with MA-activated plasma 2M, PDGF -receptor tyrosine phosphorylation was greatly decreased. Complete inhibition of tyrosine phosphorylation was not anticipated because 2M-PDGF-BB complexes are noncovalent and reversible (41). MA-activated r2M(E730R) did not inhibit PDGF -receptor tyrosine phosphorylation, providing further evidence that this mutated form of r2M does not bind PDGF-BB.
TGF-1 Binding to Mutated r2 Ms—TGF-1 binds to plasma-derived native 2M and, with increased affinity, to MA-activated 2M (22). Our evidence, collected in studies with intact plasma 2M, GST fusion proteins, and synthetic peptides, suggested that the binding sites for PDGF-BB and TGF-1 are equivalent; however, we could not rule out the possibility that these sites are overlapping but distinct (25). Fig. 8A compares the binding of 125I-TGF-1 to native and MA-activated plasma 2M and r2M(E730R). Binding was detected by nondenaturing PAGE and autoradiography. 125I-TGF-1 bound to both native and MA-activated plasma 2M; the level of binding was greatly increased with MA-activated 2M, consistent with our previous findings (22). 125I-TGF-1 also bound to the native and MA-activated forms of r2M(E730R). The extent of binding was equivalent to that observed with plasma 2M. Nondenaturing PAGE may be insufficiently sensitive to detect modest changes in binding affinity; however, these results indicate that TGF-1 binding to r2M is not neutralized by the same single mutation at Glu730, which blocks PDGF-BB binding.
|
1 to r2M(RR). Once again, binding was unchanged compared with that observed with plasma-derived 2M or wild-type r2M (Fig. 8B). This was unanticipated because mutation of Glu714 and Asp719, in the context of the 2M peptide GST fusion protein FP3 significantly decreased 125I-TGF-1 binding (26). Finally, we examined the binding of TGF-1 to r2M(RRR), which combines mutation of Glu730 with Glu714 and Asp719. As shown in Fig. 8C, in its native form, r2M(RRR) bound 125I-TGF-1 comparably with native plasma 2M; however, r2M(RRR) failed to demonstrate the characteristic increase in TGF-1 binding that accompanies reaction with MA. The studies presented in Figs. 2C and 3B demonstrated that MA induces r2M(RRR) conformational change without perturbation. Thus, we interpret the nondenaturing PAGE studies as indicating that mutation of the three specified amino acids directly affects the function of the binding site for TGF-1 in 2M that had undergone conformational change.
To confirm our results regarding the binding of TGF-1 to MA-activated r2M(RRR) in a second model system, we immobilized MA-activated plasma 2M in microtiter wells. 125I-TGF-1 was added to the wells in the presence and absence of native or MA-activated plasma 2M or r2M(RRR). As is shown in Fig. 8D, native plasma 2M inhibited binding of 125I-TGF-1 to the immobilized 2M by 30 ± 3%. Binding was inhibited 92 ± 4% by MA-activated plasma 2M. The native form of r2M(RRR) inhibited 125I-TGF-1 binding by 34 ± 5%; however, MA failed to increase the ability of r2M(RRR) to inhibit TGF-1 binding to the immobilized 2M (28 ± 6%), confirming the results presented in Fig. 8C.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
2M subunit includes the bait region, thiol ester bond (42–44), PID-1, which mediates the binding of several growth factors (17, 24–26), PID-2, which mediates binding of -amyloid peptide (37, 45), and the recognition sequence for LRP-1 (30, 46–49). Studies with 2M fragments expressed in bacteria suggest that PID-1, PID-2, and the LRP-1 recognition site function independently (24, 26, 37, 46, 47); however, relationships between these functional domains may be evident only in intact 2M. Studies with mutated full-length r2M have already yielded results that are non-identical with those obtained by studying 2M fragments. For example, in the isolated receptor-binding domain of human 2M, both Lys1370 and Lys1374 are involved in LRP-1 binding, whereas in intact 2M, Lys1374 may be mutated without an apparent change in LRP-1 binding affinity (30, 46). In this study, we have shown that essential amino acids in PID-1 may be mutated without affecting the overall structure of 2M, subunit association, disulfide bonds, the function of the thiol ester, protease inhibition, conformational change, or receptor binding.
Based on our studies with 2M peptide GST fusion proteins and synthetic peptides (17, 24–26), we hypothesized that PID-1 represents a hydrophobic surface patch, functioning as a basic binding platform for many proteins, with superimposed negatively charged amino acids that provide specificity in the range of non-covalent interactions. Previous studies suggested that, in activated 2M, the hydrophobic surface patch may be accessible from within the central cavity of 2M. For example, when 2M binds two mol/mol of trypsin, the central cavity of r2M becomes largely occupied and growth factor binding is precluded (50).
Based on our hypothesis regarding the structure of PID-1, we selected a mutagenesis strategy targeting acidic residues within the PID-1 sequence. This strategy proved successful. Mutation of Glu730 completely blocked PDGF-BB binding. Reversing the charge on residue 730, as opposed to conversion to alanine, increased our confidence that the mutation did not simply cause PID-1 to become buried. This conclusion was supported by our finding that 2M(E730R) bound TGF-1.
The TGF-1-binding site in full-length tetrameric 2M is more complex than our studies with GST fusion proteins and synthetic peptides had led us to believe. In our library of synthetic peptides, the shortest 2M-derived peptide that binds TGF-1 includes Glu714 and Asp719. Mutation of these residues in the fusion protein FP3 decreases TGF-1 binding (26); however, mutation of Glu714 and Asp719 in intact r2M(RR) failed to alter TGF-1 binding. When Glu714 and Asp719 were mutated alongside Glu730, the binding of TGF-1 to MA-activated r2M(RRR) was substantially inhibited. Intriguingly, the same mutation failed to affect the binding of TGF-1 to native 2M. These results suggest that PID-1 may play a selective role in mediating the binding of TGF-1 to the activated conformation of r2M. Our mutagenesis work has not provided an explanation for the low affinity binding of TGF-1 to native 2M (KD of 0.33 µM). The mechanism by which native 2M binds TGF-1 remains an important problem, given that the vast majority of TGF-1 in the plasma is associated with native 2M and thereby stabilized against renal filtration (18). The concentration of 2M in the plasma is 3–5 µM (2), well above the KD for TGF-1 binding.
The most important conclusion of this study is that binding of growth factors, such as PDGF-BB, to 2M reflects a specific interaction as opposed to nonspecific steric entrapment. In addition, our studies in which we have mutated 2M to neutralize a specific function (growth factor binding) offer an important opportunity to dissect the mechanism of action of 2M in cell culture and in vivo. For example, by binding to LRP-1 and by regulating the interaction of LRP-1 with the intracytoplasmic adaptor protein, Jun kinase interaction protein, activated 2M may regulate cellular apoptosis (51). Alternatively, 2M regulates cell physiology by binding endogenously produced growth factors such as TGF-1 and by disrupting autocrine signaling pathways involving these growth factors (52). R2M(RRR) may represent an excellent tool for distinguishing between these two pathways. Similar experiments are feasible with r2M(K1370A), which fails to bind LRP-1 when activated (30). Mutated forms of full-length 2M may also be helpful to dissect mechanisms responsible for abnormalities in the 2M gene knock-out mouse (53) and for determining why specific forms of 2M demonstrate potent anti-inflammatory activity when administered to endotoxin-challenged mice (54).
|
FOOTNOTES |
|---|
1 To whom correspondence should be addressed: Dept. of Pathology, University of California San Diego, 9500 Gilman Dr., La Jolla, CA 92093. Tel.: 858-534-1887; Fax: 858-534-0414; E-mail: sgonias{at}ucsd.edu.
2 The abbreviations used are: 2M, 2-macroglobulin; r2M, recombinant human 2M; r2M(RR), r2M in which Glu714 and Asp719 are converted into Arg; r2M(RRR), r2M in which Glu714, Asp719, and Glu730 are converted into Arg; PDGF-BB, platelet-derived growth factor-BB; TGF-1, transforming growth factor-1; LRP-1, low density lipoprotein-related protein-1; MA, methylamine; GST, glutathione S-transferase; BSA, bovine serum albumin; BS3, bis(sulfosuccinimidyl) suberate; PBS, phosphate-buffered saline; ERK, extracellular signal-regulated kinase.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
