HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Submit a Letter to Editor Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Matsushita, T. Articles by Sadler, J. E. Search for Related Content PubMed PubMed Citation Articles by Matsushita, T. Articles by Sadler, J. E. Social Bookmarking                      What's this?

J Biol Chem, Vol. 275, Issue 15, 11044-11049, April 14, 2000

Localization of von Willebrand Factor-binding Sites for Platelet Glycoprotein Ib and Botrocetin by Charged-to-Alanine Scanning Mutagenesis^*

Tadashi Matsushita^§, Dominique Meyer^¶, and J. Evan Sadler

From the  Howard Hughes Medical Institute, Departments of Medicine and of Biochemistry & Molecular Biophysics, Washington University School of Medicine, St. Louis, Missouri 63110 and ^¶ INSERM U.143, Hôpital Bicêtre, 94275 le Kremlin-Bicêtre, Cedex, France

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

At sites of vascular injury, von Willebrand factor (VWF) mediates platelet adhesion through binding to platelet glycoprotein Ib (GPIb). Previous studies identified clusters of charged residues within VWF domain A1 that were involved in binding GPIb or botrocetin. The contribution of 28 specific residues within these clusters was analyzed by mutating single amino acids to alanine. Binding to a panel of six conformation-dependent monoclonal antibodies was decreased by mutations at Asp⁵¹⁴, Asp⁵²⁰, Arg⁵⁵², and Arg⁶¹¹ (numbered from the N-terminal Ser of the mature processed VWF), suggesting that these residues are necessary for domain A1 folding. Binding of ¹²⁵I-botrocetin was decreased by mutations at Arg⁶²⁹, Arg⁶³², Arg⁶³⁶, and Lys⁶⁶⁷. Ristocetin-induced and botrocetin-induced binding to GPIb both were decreased by mutations at Lys⁵⁹⁹, Arg⁶²⁹, and Arg⁶³²; among this group the K599A mutant was unique because ¹²⁵I-botrocetin binding was normal, suggesting that Lys⁵⁹⁹ interacts directly with GPIb. Ristocetin and botrocetin actions on VWF were dissociated readily by mutagenesis. Ristocetin-induced binding to GPIb was reduced selectively by substitutions at positions Lys⁵³⁴, Arg⁵⁷¹, Lys⁵⁷², Glu⁵⁹⁶, Glu⁶¹³, Arg⁶¹⁶, Glu⁶²⁶, and Lys⁶⁴², whereas botrocetin-induced binding to GPIb was decreased selectively by mutations at Arg⁶³⁶ and Lys⁶⁶⁷. The binding of monoclonal antibody B724 involved Lys⁶⁶⁰ and Arg⁶⁶³, and this antibody inhibits ¹²⁵I-botrocetin binding to VWF. The crystal structure of the A1 domain suggests that the botrocetin-binding site overlaps the monoclonal antibody B724 epitope on helix 5 and spans helices 4 and 5. The binding of botrocetin also activates the nearby VWF-binding site for GPIb that involves Lys⁵⁹⁹ on helix 3.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Adhesion of platelets to vessel walls is a first step in normal hemostasis and also in thrombotic events such as the occlusion of atherosclerotic arteries. These events are mediated by at least two ligand-receptor interactions: von Willebrand factor (VWF)¹ binding to platelet glycoprotein Ib (GPIb) and fibrinogen binding to platelet GPIIb-IIIa. VWF does not bind spontaneously to platelets in blood, and its adhesive properties are induced in vivo upon the binding of VWF to subendothelial connective tissue, particularly under conditions of high shear stress that occur in the microcirculation (1).

VWF consists of disulfide-linked multimers that are assembled from subunits of  250 kDa. The VWF multimers range in size from dimers of 500 kDa to >10,000 kDa. VWF binds to an N-terminal domain of the GPIb subunit (2-4), and the GPIb-binding site on VWF corresponds approximately to the first of three repeated A domains in the VWF subunit. Domain A1 extends from Glu⁴⁹⁷ to Gly⁷¹⁶ and contains an intrachain disulfide loop that is defined by the disulfide bond Cys⁵⁰⁹-Cys⁶⁹⁵ (5, 6). Binding of VWF to GPIb in vitro can be induced by the antibiotic ristocetin or by the snake venom protein botrocetin. Ristocetin apparently can bind both to platelets and to VWF (7), whereas botrocetin binds to VWF domain A1 but not to GPIb (8).

By clustered charged-to-alanine scanning mutagenesis, we have found several clustered mutants of charged residues within VWF-A1 domain showing reduced or increased function (9). Several mutations in the two acidic segments Glu⁴⁹⁷-Arg⁵¹¹ and Arg⁶⁸⁷-Val⁶⁹⁸, which contain the Cys⁵⁰⁹-Cys⁶⁹⁵ disulfide, resulted in increased binding to GPIb (9). In addition, mutations within discontinuous segments including Glu⁵⁹⁶-Arg⁶¹⁶, Arg⁶²⁹-Arg⁶³², and Lys⁶⁴²-Lys⁶⁴⁵ decreased binding to GPIb, suggesting that several residues within these segments may interact with GPIb. We now have prepared 28 new mutants in which a single charged residue is changed to alanine, and these proteins were used to characterize the binding sites for botrocetin and GPIb. The results suggest that botrocetin and GPIb bind to adjacent sites on the VWF-A1 domain, and the amino acid residues required for ristocetin-induced binding to GPIb are distinct from those required for botrocetin-induced binding.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Materials-- Restriction enzymes were obtained from New England BioLabs (Beverly, MA). Taq DNA polymerase was from Perkin-Elmer. Highly purified two-chain botrocetin was provided by Dr. Yoshihiro Fujimura (Nara Medical University, Nara, Japan). Monoclonal antibody 6D1 against human platelet GPIb (10) was provided by Dr. Barry Coller (Mt. Sinai Medical Center, NY). Anti-VWF monoclonal antibody 33E12 was provided by Dr. Claudine Mazurier (Centre Regional de Transfusion Sanguine, Lille, France).

Monoclonal antibodies (mAbs) recognizing the VWF-A1 domain were kindly provided by the following researchers: NMC-4 (Dr. Midori Shima, Nara Medical University, Nara, Japan), AvWF3 (Dr. Philip Kroner, The Blood Center of Southeastern Wisconsin), CLB-RAG34 and CLB-RAG35 (Dr. J. A. van Mourik, Netherlands Red Cross Blood Transfusion Service), 211A6 (Dr. Claudine Mazurier), and 52K2 (Dr. Zaverio Ruggeri, Scripps Research Institute, La Jolla, CA).

Plasmid Constructs-- The strategy for mutagenesis was described previously (9). A polymerase chain reaction method was used to introduce mutations into plasmid pGEM-4ZNK (11), and each mutation was confirmed by DNA sequencing using a dideoxy termination method (Sequenase 2.0, U. S. Biochemical Corp.). The mutated NgoMI-KpnI fragments then were cloned into pSVHVWF1.1 (9).

Expression and Characterization of Recombinant VWF-- Human 293T cells (12) were grown in Dulbecco's modified Eagle's medium (Life Technologies, Inc.) supplemented with 10% fetal bovine serum (Life Technologies, Inc.). Cells were transfected by a calcium-phosphate method and recombinant VWF (rVWF) secreted in the medium was harvested and concentrated by ultrafiltration as described (9). VWF antigen was measured by an ELISA using polyclonal rabbit anti-human VWF antibody 082 and peroxidase-conjugated rabbit anti-human VWF antibody 226 (DAKO) (13). Multimer analysis was performed as described (14).

Binding of VWF to mAb Panel-- The IgG fractions of mAbs NMC-4, AVWF3, 52K2, CLB-RAG34, and CLB-RAG35 were isolated from ascites fluid by chromatography on recombinant protein A-agarose (Amersham Pharmacia Biotech). Binding was assessed by ELISA using microtiter plates with U-shaped bottoms (Coster, Cambridge, MA). Plates were coated with 25 µl of each mAb (7.5 µg/ml) in 0.1 M sodium carbonate, pH 9.6, for 24 h at 4 °C. The wells were washed with phosphate-buffered saline (PBS) containing 0.1% Tween 20 and then incubated with various concentrations of wild type or mutant rVWF diluted in PBS containing 3% bovine serum albumin (Sigma). The wells were washed again and incubated with antibody 226 diluted in PBS containing 3% bovine serum albumin color development with o-phenylenediamine, and the absorbance at 490 nm was determined. Binding of rVWF mutants was determined at a fixed concentration of rVWF (500 ng/ml) and normalized to the value obtained for wild type rVWF in paired assays. Negative control assays were performed by using concentrated conditioned medium from mock-transfected 293T cells.

¹²⁵I-Botrocetin Binding Assays-- Botrocetin binding to VWF was assayed according to methods of Fujimura et al. (15) as described previously (9). The assay was performed using anti-VWF mAb 33E12 that binds to the C-terminal region of the VWF subunit. It has no effect on VWF binding to platelets in the presence of either ristocetin or botrocetin (16). In brief, botrocetin was radioiodinated with ¹²⁵I using a chloramine-T method (17). Microtiter plates were coated with 25 µl of 33E12 (7.5 µg/ml) for 16 h at 4 °C. The wells were washed, blocked, and incubated with 15 µl of each rVWF mutant (5 µg/ml) for 3 h at room temperature. The wells were washed again, and 5 µl of ¹²⁵I-labeled botrocetin solution was added for 30 min at room temperature. Following rapid washing, air dried wells were excised and the bound radioactivity was measured by  spectroscopy. Nonspecific binding was obtained by assaying culture supernatant from mock-transfected cells, and specific binding was calculated by subtracting nonspecific from total binding. Values for mutant proteins at 3 µg/ml of ¹²⁵I-botrocetin were normalized to paired values for wild type rVWF.

mAb B724 was assayed for ability to inhibit ¹²⁵I-botrocetin binding to VWF. Microtiter plates coated with mAb 33E12 were incubated with wild type rVWF or mutant K599A. ¹²⁵I-botrocetin (3 µg/ml), and the indicated amount of mAb B724 or control mouse IgG (DAKO) was added and then incubated for 30 min at room temperature. After washing, the bound radioactivity was measured as described above. The observed botrocetin binding was normalized to the value obtained in the absence of competing IgG.

Platelet Binding Assays-- Assays were performed as described previously (9). To assay botrocetin-induced binding of VWF to GPIb, 570 ng/ml of rVWF was mixed with 2 × 10⁸/ml of lyophilized human platelets (Biodata, Hatboro, PA) and incubated with various concentrations (0-10 µg/ml) of botrocetin. After 30 min at room temperature, the VWF antigen present in the centrifuged supernatant was measured by ELISA. Data for botrocetin concentrations bracketing the midpoint of the dose response curve (2, 3, and 6 µg/ml) were pooled for statistical analysis. To assay ristocetin-induced binding of VWF to GPIb, 500 ng/ml of rVWF was mixed with 2 × 10⁸/ml of platelets, 4% bovine serum albumin, and various concentrations (0-1.5 mg/ml) of ristocetin (Helena Laboratories, Beaumont, TX). After 30 min at room temperature, the VWF in the centrifuged supernatant was measured by ELISA. Reaction mixtures without platelets were tested simultaneously to verify the absence of nonspecific VWF flocculation and sedimentation in the presence of ristocetin. The unbound VWF was expressed as a percentage of the values obtained with no modulators, and the percentage of bound VWF was calculated by subtraction from 100%. The percentage of bound VWF was normalized to the paired value obtained for wild type rVWF assayed concurrently. Data at 1.5 mg/ml ristocetin were used for statistical analysis.

Statistical Analysis-- Means and S.D. were calculated by standard methods. Confidence intervals (p < 0.05) for the means were estimated using the t distribution where the interval is given by the means ± S.D. × t/(n)^0.5.

Crystallographic Structure Representations-- Connelly surface plots of the VWF A1 domain were prepared with the program INSIGHT II (Molecular Simulations, San Diego, CA). Schematic drawings of secondary structure elements of domain A1 were prepared with the program MOLSCRIPT (18).

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Design and Expression of VWF A1 Domain Mutants-- In a previous study by charged-to-alanine scanning mutagenesis, we identified 6 single mutants and 13 clustered mutants with defects in binding to either botrocetin or platelet GPIb. From these data, the binding site for botrocetin or GPIb required amino acid residues within four segments of the VWF-A1 domain: 514-534, 549-552, 596-645, and 663-667 (Fig. 1) (9). Other reports suggested that binding to GPIb and heparin required residues within the segment 569-573 (19), and preliminary studies indicated that monoclonal antibody B724 (20) did not bind to construct (656-660)2A.

View larger version (38K): [in this window] [in a new window]   Fig. 1.   Amino acid residues of human VWF targeted for charged-to-alanine mutagenesis. The amino acid sequence shown includes a part of domain D3 (463-496) and the entire A1 domain (497-716). The secondary structure elements of the A1 domain as determined by Celikel et al. (27) are indicated below the sequence (, -helix; , -strand). Charged residues His, Arg, Lys, Glu, and Asp were targeted for the mutagenesis and are shown by characters in red. Previously studied clustered mutants have alanine mutations at more than one charged amino acid residue in a single construct (9) and are indicated by single underlines. The extent of the underlining indicates the residues mutated in a single construct. For example, construct (527-531)3A contained alanine substitutions at the three charged residues Glu527, Glu529, and Glu531. For the 11 clustered mutations shown, 28 corresponding single charged-to-alanine mutations were produced and are indicated by yellow boxes. Six additional single mutants that were constructed previously (9) are indicated by yellow boxes with double underlines.

We further studied these six segments by constructing 28 additional single charged-to-alanine mutations (Fig. 1). Human kidney 293T cells were transfected with each mutant construct, and serum-free media were analyzed for the expression of rVWF. All 28 mutants were expressed and secreted efficiently. The multimer distribution of all mutant proteins was similar to that of wild type rVWF and plasma VWF; in every case, at least 12 multimer bands were detected (data not shown).

rVWF Binding to Monoclonal Antibodies-- The folding of the A1 domain with each rVWF protein was evaluated with a panel of six conformation-dependent monoclonal antibodies: NMC-4 (21, 22), AvWF3 (23), CLB-RAG34, CLB-RAG35 (24), B724 (20), and 211A6 (25). Recognition of domain A1 by these antibodies is impaired by reduction or denaturation. Antibody 52K2 reacts with both reduced and nonreduced forms of VWF (21) and was used as a control. For each antibody, the absorbance value obtained for each mutant rVWF was expressed as a percentage of the value obtained for wild type rVWF (Fig. 2). All rVWF proteins displayed normal binding to mAb 52K2. Concentrated conditioned medium from mock transfected cells gave no signal with any of these antibodies.

View larger version (105K): [in this window] [in a new window]   Fig. 2.   Binding of rVWF to monoclonal antibodies against the human VWF-A1 domain. Each monoclonal antibody (indicated at the top of each histogram) was coated onto plastic microtiter wells. Binding was determined at a fixed concentration of rVWF (500 ng/ml) and normalized to the value obtained for wild type rVWF as described under "Experimental Procedures." Except for 52K2, all the antibodies bind to unreduced but not to reduced VWF A1 domain. 52K2 reacts with both reduced and unreduced VWF. The mutant rVWF proteins are indicated at the left. Each column represents the mean and S.D. obtained for at least duplicate assays. The results of NMC4 binding for 33 mutants were reported previously (9), and for comparison these results are indicated by open columns, representing the mean ± S.D. of values obtained in at least two independent sets of duplicate assays. Asterisks indicate values that are significantly different from 100% (p < 0.05).

Mutants D514A, D520A, R552A, and R611A showed markedly decreased binding to all six conformation-dependent mAbs (Fig. 2), suggesting that these substitutions cause significant misfolding of the A1 domain. Therefore, although further ligand binding data were obtained for these proteins, the results were excluded from additional interpretation. In fact, the binding of this class of mutants to botrocetin and GPIb was reduced >70% in all assays (data not shown). Mutant K534A showed decreased binding to four mAbs (AvWF3, CLB-RAG34, CLB-RAG35, and B724), although it retained nearly normal binding to two others (NMC-4 and 211A6). Thus, Lys⁵³⁴ is at least required for the presentation of epitopes for several antibodies and may be important to maintain the normal conformation of the A1 domain.

Other substitutions caused relatively selective defects in mAb binding. For example, at a concentration of 0.5 µg/ml mutant rVWF, binding to mAb B724 was decreased markedly by the mutations K660A and R663A (Fig. 2). These mutants showed normal binding to all other antibodies. Dose response binding curves illustrate the substantial decrease in affinity caused by the substitution K660A or R663A (Fig. 3) and suggest that the epitope of mAb B724 includes the side chains of Lys⁶⁶⁰ and Arg⁶⁶³. Similarly, binding of mAb NMC4 was reduced selectively by the substitution R632A, suggesting that its epitope contains the side chain of Arg⁶³².

View larger version (13K): [in this window] [in a new window]   Fig. 3.   Binding of VWF to mAb B724. mAb B724 was immobilized on microtiter plates, and VWF binding was determined as described under "Experimental Procedures" for wild type rVWF (closed circles), mutant K660A (open circles), and mutant R663A (closed squares). Nonspecific binding (open squares) was determined with concentrated conditioned medium from mock-transfected 293T cells and was undetectable. Each data point represents the mean ± S.D. of values obtained in at least two independent sets of duplicate assays. The symbols are larger than the S.D. range in most cases.

Binding of Botrocetin to rVWF-- Previous charged-to-alanine mutagenesis studies suggested that several amino acid residues of VWF domain A1 participate directly in binding to ¹²⁵I-botrocetin, including Arg⁶³², Arg⁶³⁶, and residues in the segments Lys⁶⁴²-Lys⁶⁴⁵ and Arg⁶⁶³-Lys⁶⁶⁷ (9). In addition, clustered mutations in other segments decreased binding to botrocetin, but their significance was uncertain because these mutations also impaired binding to mAb NMC4 so that protein misfolding could not be excluded (9).

To address this uncertainty, additional rVWF mutants were constructed and assayed for binding to 3 µg/ml ¹²⁵I-botrocetin (Fig. 4). At this concentration of botrocetin, binding to VWF is approximately half of the maximum (9, 15). Four previously constructed proteins that had shown decreased binding to GPIb or botrocetin were included as well: K534A, E626A, R632A, and R636A.

View larger version (68K): [in this window] [in a new window]   Fig. 4.   Histogram of 125I-botrocetin binding to VWF. The binding of radiolabeled botrocetin (3 µg/ml) to each mutant rVWF was measured and normalized to the values obtained for wild type rVWF as described under "Experimental Procedures." Data for five mutants (D514A, K534A, E626A, R632A, and R636A) were reported previously (9). Each shaded bar represents the mean ± S.D. of values obtained for up to four independent assays. Asterisks indicate values that are significantly different from 100% (p < 0.05).

Four mutants exhibited decreased binding <60% of wild type rVWF: R629A, R632A, R636A, and K667A. As discussed below, the mutations R629A and R632A also disrupt binding to GPIb induced by ristocetin, whereas the mutations R636A and K667A do not. Therefore, the side chains of Arg⁶³⁶ and Lys⁶⁶⁷ are specifically required for binding to botrocetin. As reported previously for plasma VWF (20), mAb B724 inhibited the binding of botrocetin to immobilized wild type rVWF (data not shown).

Although botrocetin binding was markedly reduced by the clustered mutation of four lysine residues in construct (642-645)4A (9), the individual mutation of each lysine had a modest effect on binding to botrocetin (Fig. 4), suggesting that no one of these residues makes a substantial contribution to this interaction. Similarly, the clustered mutant (613-616)2A exhibited markedly impaired binding to botrocetin (9), but the corresponding single mutants E613A and R616A bound with slightly decreased affinity to botrocetin (Fig. 4).

Binding of rVWF to GPIb-- Binding to GPIb was assessed by quantitating the rVWF that bound to formalin-fixed platelets in the presence of ristocetin (1.5 mg/ml) or botrocetin (2-6 µg/ml). In either case, binding to platelets was blocked completely by a mAb 6D1 to platelet GPIb (10) as reported previously (26). Values for each mutant protein were normalized to those obtained for control assays of wild type rVWF.

The VWF mutants R524A, E529A, E531A, K549A, K569A, D570A, R573A, K608A, D610A, K643A, K644A, K645A, H656A, K660A, R663A, and E666A displayed binding to GPIb that was >60% of wild type rVWF in the presence of ristocetin or botrocetin (Fig. 5). These mutants also showed essentially normal ¹²⁵I-botrocetin binding (Fig. 4), suggesting that the corresponding amino acid side chains are not directly involved in the interactions of VWF with GPIb, ristocetin, or botrocetin.

View larger version (61K): [in this window] [in a new window]   Fig. 5.   Histogram of VWF binding to platelet GPIb. Botrocetin-induced (A) and ristocetin-induced (B) binding of each rVWF mutant was determined as described under "Experimental Procedures." Platelet binding is expressed as a percentage relative to that determined for wild type rVWF. Data for five mutants (D514A, K534A, E626A, R632A, and R636A) were reported previously (9). Each bar represents the mean ± S.D. of values obtained. For botrocetin-induced binding the average number of independent determinations was six (range five to eleven); for ristocetin-induced binding the average number of independent determinations was four (range three to six). Asterisks indicate values that are significantly different from 100% (p < 0.05).

The mutants R636A and K667A showed decreased botrocetin-induced GPIb binding (Fig. 5) that is consistent with their decreased binding to ¹²⁵I-botrocetin (Fig. 4). However, these proteins retained normal ristocetin-induced GPIb binding (Fig. 5), indicating that the side chains of Arg⁶³⁶ and Lys⁶⁶⁷ interact with botrocetin but not with GPIb.

Alanine substitutions at eight residues (Lys⁵³⁴, Arg⁵⁷¹, Lys⁵⁷², Glu⁵⁹⁶, Glu⁶¹³, Arg⁶¹⁶, Glu⁶²⁶, and Lys⁶⁴²) decreased ristocetin-induced GPIb binding without affecting botrocetin-induced GPIb binding (Fig. 5) or ¹²⁵I-botrocetin binding (Fig. 4). Therefore, these sites appear to be required for ristocetin-dependent VWF modulation but not for direct interaction with GPIb.

Mutation at a residue that directly interacts with GPIb would be predicted to reduce both ristocetin-induced and botrocetin-induced GPIb binding, and three mutants had this phenotype: K599A, R629A, and R632A. Among them, only K599A retained normal binding to ¹²⁵I-botrocetin. Because GPIb and botrocetin can bind simultaneously to VWF, the selective loss of GPIb binding suggests that Lys⁵⁹⁹ interacts directly with GPIb and not with botrocetin. The mutants R629A and R632A decreased binding to both GPIb and ¹²⁵I-botrocetin, suggesting that these amino acid side chains could be involved in the binding of both ligands.

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Crystallographic structures were determined recently for two forms of the VWF A1 domain (27, 28). The A1 domain contains a central -sheet of five parallel strands and one antiparallel strand sandwiched between three -helices on each side (Fig. 6). A disulfide bond at one end of the -sheet links the N-terminal and C-terminal ends, and the sequence flanking the disulfide bond at the N terminus of the domain lies across its lower surface. These structures provide a framework for interpreting the effects of charged-to-alanine mutations on ligand binding.

View larger version (69K): [in this window] [in a new window]   Fig. 6.   Ligand-binding sites on the VWF A1 domain. On the left is a surface representation of the VWF A1 domain based on the coordinates of Emsley et al. (28). The locations are indicated of amino acid residues involved in the binding of platelet GPIb (red), mAb NMC-4 (orange), botrocetin (green), and mAb B724 (blue). Residue Arg636 is colored in both green and orange to indicate that it interacts with both NMC-4 and botrocetin. On the right is a schematic drawing of the A1 domain in the same orientation. The coloring of amino acid side chains is the same as in the left panel. Only a partial side chain is shown for Lys660 because atoms more distal than C were not present in the structure (28). Helices are colored for clarity: 3 in red, 4 in orange, and 5 in blue. Helices are numbered according to Celikel et al. (27). The Cys509-Cys695 disulfide bond is shown at the bottom right, with S atoms in yellow.

Binding Sites for GPIb, Botrocetin, and Monoclonal Antibodies-- K599A was the only VWF A1 mutant that retained normal binding to botrocetin and lost both ristocetin-induced and botrocetin-induced binding to GPIb. This specific effect suggests that the GPIb-binding site involves Lys⁵⁹⁹, which is located in the middle of helix 3 (Fig. 6). The mutations R629A and R632A in the adjacent helix 4 impaired binding to GPIb but also decreased botrocetin binding, suggesting that Arg⁶²⁹ or Arg⁶³² may interact with both GPIb and botrocetin. However, the interpretation of this dual defect is complicated by the strong positive cooperativity between botrocetin and GPIb binding. Because of this linkage, a single mutation in domain A1 might reduce the affinity for both ligands by interfering with allosteric changes rather than by disrupting a common binding site. The mutations R636A and K667A have a simpler phenotype, impairing botrocetin binding (Fig. 4) but not ristocetin-induced binding to GPIb (Fig. 5). The selective effect of these mutations indicates that the botrocetin-binding site is, in fact, adjacent to the GPIb-binding site and spans both helix 4 and helix 5 (Fig. 6).

This position for the botrocetin site is supported by the localization of epitopes for two monoclonal antibodies that compete for botrocetin binding, NMC-4 (29) and B724 (20). The crystallographic structure of the complex (27) shows that the NMC-4 Fab fragment interacts directly with the amino acid side chains of Arg⁶²⁹, Arg⁶³², and Arg⁶³⁶ of helix 4 in the VWF A1 domain (Fig. 6). The mutation R632A decreased the binding of NMC-4 (9) (Fig. 2) and is compatible with the location of the NMC-4 epitope. The binding sites for NMC-4 and botrocetin therefore both appear to contain Arg⁶³⁶, and this overlap explains their competitive interaction with VWF (Fig. 6). The proximity of the NMC-4 epitope to the proposed GPIb site involving Lys⁵⁹⁹ also is consistent with the ability of NMC-4 to inhibit the binding of VWF to platelets (22, 29).

In contrast to NMC-4, antibody B724 does not prevent the binding of VWF to platelets in the presence of ristocetin (20), suggesting that its epitope is distant from the GPIb site but close to the botrocetin site. This conclusion is supported by the results of mutagenesis (Fig. 2), which indicate that the B724 epitope contains residues Lys⁶⁶⁰ and Arg⁶⁶³ on helix 5 (Fig. 6). The adjacent residue Lys⁶⁶⁷ contributes to the botrocetin-binding site and probably accounts for the ability of B724 to inhibit botrocetin binding to VWF (20).

These results suggest a model for the location of several binding sites on the VWF A1 domain (Fig. 6). Platelet GPIb interacts with a site that includes Lys⁵⁹⁹ within helix 3. Botrocetin binding requires residues in both helices 4 and 5 that are within or adjacent to the epitopes for NMC-4 and B724, respectively. NMC-4 binds to helix 4 at a site that is sufficiently close to the GPIb site to prevent the binding of VWF to platelets. Antibody B724 binds to helix 5 at a site too remote to interfere with GPIb binding. All of these binding sites involve helices that are on the same side of the central -sheet within the VWF A1 domain.

Selective Loss of Ristocetin-induced Binding to GPIb-- Mutations that abolish the modulation of VWF activity by ristocetin frequently preserve modulation by botrocetin, and this phenotype indicates that ristocetin and botrocetin promote VWF binding to GPIb by fundamentally different mechanisms. Eight VWF mutations (K534A, R571A, K572A, E596A, E613A, R616A, E626A, and K642A) caused a selective decrease in ristocetin-induced binding to platelets. A few patients with severe bleeding and von Willebrand disease type 2M have plasma VWF with a similar phenotype. In three such cases, mutations within the A1 domain have been characterized by expression studies: G561S, F606I, and I662F (26, 30). In other patients with VWD type 2M, candidate mutations were identified that involve residues also implicated by alanine mutagenesis, including deletion of Lys⁶⁴² and E596K (31).

The mechanism by which ristocetin induces VWF to bind platelets is not understood. All but three of the mutations that selectively impair ristocetin-induced binding to GPIb are on the same side of the central -sheet as the binding sites for GPIb and botrocetin. Because botrocetin remains capable of inducing high affinity binding to platelets, the affected amino acid residues are not required for binding to GPIb. Instead they may be required for ristocetin binding or contribute to binding under conditions of fluid shear stress, or mutations at these positions may prevent conformational changes induced by ristocetin. The five residues known to be mutated in patients with VWD type 2M (Gly⁵⁶¹, Glu⁵⁹⁶, Phe⁶⁰⁶, Lys⁶⁴², and Ile⁶⁶²) are particularly important for the biological function of the GPIb-binding site in vivo. They are distributed around Lys⁵⁹⁹ (Fig. 7), and mutations at these positions could prevent the normal exposure of the GPIb-binding site during platelet adhesion.

View larger version (49K): [in this window] [in a new window]   Fig. 7.   Location of amino acid residues required for ristocetin-induced binding of VWF to platelet GPIb. The stereo schematic drawing of the VWF A1 domain shows in yellow the location of amino acid side chains that, when mutated to alanine, result in the selective loss of ristocetin-induced binding to platelets. Side chains in blue indicate positions of mutations in patients with VWD type 2M that have been characterized by expression of recombinant mutant VWF. For reference, the side chain of Lys599 (red) is shown to indicate the approximate position of the binding site for GPIb.

Regulation of VWF Binding to GPIb-- A remarkable subtype of VWD, referred to as type 2B, is characterized by mutant VWF that has increased affinity for platelet GPIb. In vivo, the larger VWF multimers bind spontaneously to platelets and are cleared from the circulation. The remaining small VWF multimers do not function normally, and the patients bleed. Mutations that cause VWD type 2B are known to affect 15 amino acid residues in the VWF A1 domain (32-35), and they cluster on one side of the central -sheet (Fig. 8). Interestingly, all of these residues are located on the side opposite the GPIb-binding site, which includes Lys⁵⁹⁹. The cluster of VWD type 2B mutations appears to mark the location of a regulatory site that normally inhibits the binding of domain A1 to GPIb. Mutations affecting the regulatory site can relieve this inhibition and cause the constitutive binding that characterizes VWD type 2B (9, 36, 37). A challenge for the future is to determine how mutations on one side of domain A1 can activate a GBIb-binding site on the opposite side and whether this apparently allosteric regulation of VWF-platelet interactions is important for normal hemostatis.

View larger version (80K): [in this window] [in a new window]   Fig. 8.   VWD type 2B gain-of-function mutations. A, the left panel is a surface representation of the VWF A1 domain showing the locations of mutations that cause VWD type 2B (yellow) and the location of K599 in the GPIb-binding site (red). In the right panel, the A1 domain has been rotated 90° about the vertical axis. B, this stereo schematic drawing is in the same orientation as the left panel of A. The side chains are shown and numbered for amino acid residues that are mutated in patients with VWD type 2B. Blue, nitrogen; red, oxygen; yellow, sulfur; black, carbon.

	ACKNOWLEDGEMENTS

We thank Lisa Westfield for preparing oligonucleotides. We also thank Drs. Shuji Miura and Akira Katsumi at Washington University for helpful discussions. We thank Drs. Takayuki Nakayama, Tetsuhito Kojima, and Hidehiko Saito at Nagoya University for generous suggestions and discussions.

	FOOTNOTES

^* The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

^§ Present address: First Dept. of Internal Medicine, Nagoya University School of Medicine, Nagoya, Aichi 466-8550, Japan.

To whom correspondence should be addressed: Howard Hughes Medical Inst., Washington University School of Medicine, 660 South Euclid Ave., Box 8022, St. Louis, MO 63110. Tel.: 314-362-9029; Fax: 314-454-3012; E-mail: esadler@im.wustl.edu.

	ABBREVIATIONS

The abbreviations used are: VWF, von Willebrand factor; mAb, monoclonal antibody; PBS, phosphate-buffered saline; rVWF, recombinant von Willebrand factor; VWD, von Willebrand disease; GPIb, glycoprotein Ib; ELISA, enzyme-linked immunosorbent assay.

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

CiteULike

Complore

Connotea

Del.icio.us

Digg

Reddit

Technorati

This article has been cited by other articles:

				I. Marx, O. D. Christophe, P. J. Lenting, A. Rupin, M.-O. Vallez, T. J. Verbeuren, and C. V. Denis Altered thrombus formation in von Willebrand factor-deficient mice expressing von Willebrand factor variants with defective binding to collagen or GPIIbIIIa Blood, August 1, 2008; 112(3): 603 - 609. [Abstract] [Full Text] [PDF]

				I. Marx, P. J. Lenting, T. Adler, R. Pendu, O. D. Christophe, and C. V. Denis Correction of Bleeding Symptoms in von Willebrand Factor-Deficient Mice by Liver-Expressed von Willebrand Factor Mutants Arterioscler. Thromb. Vasc. Biol., March 1, 2008; 28(3): 419 - 424. [Abstract] [Full Text] [PDF]

				R. De Cristofaro, F. Peyvandi, R. Palla, S. Lavoretano, R. Lombardi, G. Merati, F. Romitelli, E. Di Stasio, and P. M. Mannucci Role of Chloride Ions in Modulation of the Interaction between von Willebrand Factor and ADAMTS-13 J. Biol. Chem., June 17, 2005; 280(24): 23295 - 23302. [Abstract] [Full Text] [PDF]

				A. Shimizu, T. Matsushita, T. Kondo, Y. Inden, T. Kojima, H. Saito, and M. Hirai Identification of the Amino Acid Residues of the Platelet Glycoprotein Ib (GPIb) Essential for the von Willebrand Factor Binding by Clustered Charged-to-Alanine Scanning Mutagenesis J. Biol. Chem., April 16, 2004; 279(16): 16285 - 16294. [Abstract] [Full Text] [PDF]

				N. Maita, K. Nishio, E. Nishimoto, T. Matsui, Y. Shikamoto, T. Morita, J. E. Sadler, and H. Mizuno Crystal Structure of von Willebrand Factor A1 Domain Complexed with Snake Venom, Bitiscetin: INSIGHT INTO GLYCOPROTEIN Ib{alpha} BINDING MECHANISM INDUCED BY SNAKE VENOM PROTEINS J. Biol. Chem., September 26, 2003; 278(39): 37777 - 37781. [Abstract] [Full Text] [PDF]

				A. Bonnefoy, H. Yamamoto, C. Thys, M. Kito, J. Vermylen, and M. F. Hoylaerts Shielding the front-strand beta 3 of the von Willebrand factor A1 domain inhibits its binding to platelet glycoprotein Ibalpha Blood, February 15, 2003; 101(4): 1375 - 1383. [Abstract] [Full Text] [PDF]

				E. G. Huizinga, S. Tsuji, R. A. P. Romijn, M. E. Schiphorst, P. G. de Groot, J. J. Sixma, and P. Gros Structures of Glycoprotein Ibalpha and Its Complex with von Willebrand Factor A1 Domain Science, August 16, 2002; 297(5584): 1176 - 1179. [Abstract] [Full Text] [PDF]

				T. Nakayama, T. Matsushita, Z. Dong, J. E. Sadler, S. Jorieux, C. Mazurier, D. Meyer, T. Kojima, and H. Saito Identification of the Regulatory Elements of the Human von Willebrand Factor for Binding to Platelet GPIb. IMPORTANCE OF STRUCTURAL INTEGRITY OF THE REGIONS FLANKED BY THE CYS1272-CYS1458 DISULFIDE BOND J. Biol. Chem., June 7, 2002; 277(24): 22063 - 22072. [Abstract] [Full Text] [PDF]

				A. S. Tait, J.-F. Dong, J. A. Lopez, I. W. Dawes, and B. H. Chong Site-directed mutagenesis of platelet glycoprotein Ibalpha demonstrating residues involved in the sulfation of tyrosines 276, 278, and 279 Blood, May 29, 2002; 99(12): 4422 - 4427. [Abstract] [Full Text] [PDF]

				A.-S. Ribba, L. Hilbert, J.-M. Lavergne, E. Fressinaud, C. Boyer-Neumann, C. Ternisien, I. Juhan-Vague, J. Goudemand, J.-P. Girma, C. Mazurier, et al. The arginine-552-cysteine (R1315C) mutation within the A1 loop of von Willebrand factor induces an abnormal folding with a loss of function resulting in type 2A-like phenotype of von Willebrand disease: study of 10 patients and mutated recombinant von Willebrand factor Blood, February 15, 2001; 97(4): 952 - 959. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) Submit a Letter to Editor Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Matsushita, T. Articles by Sadler, J. E. Search for Related Content PubMed PubMed Citation Articles by Matsushita, T. Articles by Sadler, J. E. Social Bookmarking                      What's this?