Advertisement
JBC

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Originally published In Press as doi:10.1074/jbc.M705195200 on December 26, 2007

J. Biol. Chem., Vol. 283, Issue 11, 7230-7241, March 14, 2008
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Supplemental Data
Right arrow All Versions of this Article:
283/11/7230    most recent
M705195200v1
Right arrow Submit a Letter to Editor
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrowRequest Permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Shao, C.
Right arrow Articles by Gilbert, G. E.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Shao, C.
Right arrow Articles by Gilbert, G. E.
Social Bookmarking
Add to CiteULike   Add to Complore   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Reddit   Add to Technorati  
What's this?

Crystal Structure of Lactadherin C2 Domain at 1.7Å Resolution with Mutational and Computational Analyses of Its Membrane-binding Motif*Formula

Chenghua Shao{ddagger}, Valerie A. Novakovic§, James F. Head{ddagger}, Barbara A. Seaton{ddagger}, and Gary E. Gilbert§1

From the {ddagger}Department of Physiology and Biophysics, Boston University School of Medicine, Boston, Massachusetts 02118 and the §Department of Medicine, Veterans Affairs Boston Healthcare System, Brigham and Women's Hospital and Harvard Medical School, Boston, Massachusetts 02132

Received for publication, June 25, 2007 , and in revised form, December 18, 2007.


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
 REFERENCES
 
Lactadherin is a phosphatidyl-L-serine (Ptd-L-Ser)-binding protein that decorates membranes of milk fat globules. The major Ptd-L-Ser binding function of lactadherin has been localized to its C2 domain, which shares homology with the C2 domains of blood coagulation factor VIII and factor V. Correlating with this homology, purified lactadherin competes efficiently with factors VIII and V for Ptd-L-Ser binding sites, functioning as a potent anticoagulant. We have determined the crystal structure of the lactadherin C2 domain (Lact-C2) at 1.7Å resolution. The bovine Lact-C2 structure has a β-barrel core that is homologous with the factor VIII C2 (fVIII-C2) and factor V C2 (fV-C2) domains. Two loops at the end of the β-barrel, designated spikes 1 and 3, display four water-exposed hydrophobic amino acids, reminiscent of the membrane-interactive residues of fVIII-C2 and fV-C2. In contrast to the corresponding loops in fVIII-C2 and fV-C2, spike 1 of Lact-C2 adopts a hairpin turn in which the 7-residue loop is stabilized by internal hydrogen bonds. Further, central glycine residues in two membrane-interactive loops may enhance conformability of Lact-C2 to membrane binding sites. Mutagenesis studies confirmed a membrane-interactive role for the hydrophobic and/or Gly residues of both spike 1 and spike 3. Substitution of spike 1 of fVIII-C2 into Lact-C2 also diminished binding. Computational ligand docking studies identified two prospective Ptd-L-Ser interaction sites. These results identify two membrane-interactive loops of Lact-C2 and provide a structural basis for the more efficient phospholipid binding of lactadherin as compared with factor VIII and factor V.


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
 REFERENCES
 
Lactadherin is a Mr 47,000 glycoprotein that was identified as a component of milk fat globules. Lactadherin has been known as PAS-6/7, indicating the two glycosylation variants (1), bovine-associated mucoprotein, BA-46, P47, and MFG-E8 (2). Lactadherin is secreted into milk by mammary epithelial cells of humans, cows, and mice (3). The protein is also secreted by some other cells, including aortic medial smooth muscle cells (4), the epithelia of the vas deferens (5, 6), stimulated macrophages (7), and stimulated endothelial cells (8).

Bovine lactadherin has a domain structure of EGF1-EGF2-C1-C2 (where EGF indicates epidermal growth factor homology domains). Human lactadherin lacks the first epidermal growth factor homology domain but shares 64% amino acid sequence identity with bovine lactadherin at the other three domains. The EGF2 domain displays an Arg-Gly-Asp motif (9) that binds to the {alpha}vβ5 and {alpha}vβ3 integrins (1, 10-12). The C domains of lactadherin share homology with each other and with the discoidin family of lectin domains, including the lipid-binding "C" domains of blood coagulation factor VIII and factor V (2) (homology reviewed by Macedo-Ribeiro et al. (13)). Despite similar nomenclature, these C domains do not share sequence homology with the Ca2+-binding C2 domains of synaptotagmin, protein kinase C, or phospholipase C. In lactadherin, the second C domain binds to phospholipids, particularly phosphatidyl-L-serine (Ptd-L-Ser)2 (12). The demonstrated functions of lactadherin relate to binding Ptd-L-Ser and/or the {alpha}vβ3 and {alpha}vβ5 integrins.

Lactadherin binds to the phospholipid bilayer that surrounds the central triglyceride droplet of milk fat globules, apparently stabilizing the bilayer (3). Lactadherin is also found within milk ductules of the mammary gland, localized on the apical portion of secretory epithelium (3). During breast involution, lactadherin bridges Ptd-L-Ser on fat globules and apoptotic epithelial cells to integrins on phagocytic cells, mediating apoptotic clearance (14). Lactadherin secreted by macrophages aids clearance of apoptotic lymphocytes, bridging exposed Ptd-L-Ser of the dying cells and macrophage integrin(s) (7, 15). Aging mice that are deficient in lactadherin develop splenomegaly and immune complex glomerulonephritis, apparently because of impaired phagocytosis of apoptotic lymphocytes (15). Recently, lactadherin has been found to be associated with accelerated athero-sclerosis in mice (16). Lactadherin of humans was also found to bind to and help the clearance of amyloid β-peptide, which implied a possible means of controlling overproduction of amyloid β-peptide, the hallmark of Alzheimer disease (17).

We have recently found that purified lactadherin functions as a potent anticoagulant (18) and as a reagent for detection of Ptd-L-Ser exposure early in apoptosis of immortalized leukemia cells (19). Homology between the lactadherin C domains and those of factor VIII and factor V correlates with the capacity of lactadherin to compete efficiently for membrane binding sites on Ptd-L-Ser-containing membranes. The capacity for effective competition is explained, in part, by stereoselective binding of Ptd-L-Ser (20). Lactadherin inhibits the factor Xase complex, in which factor VIII functions, and the prothrombinase complex, in which factor V functions. Interestingly, factor V and factor VIII do not compete efficiently with each other for membrane binding sites (21, 22), whereas lactadherin displaces both with half-maximal displacement at 1-4 nM (18). Lactadherin also competes for membrane binding sites of vitamin K-dependent coagulation proteins, inhibiting the factor VIIa-tissue factor complex. The coagulation inhibitory properties and Ptd-L-Ser-detecting properties of lactadherin contrast with those of annexin V and other tested Ptd-L-Ser-binding proteins. The chief difference is that lactadherin binds to membranes with Ptd-L-Ser content below the threshold for annexin V and competes for most or all of the membrane binding sites. Annexin V and other Ptd-L-Ser-binding proteins have a more limited capacity for competition (18). Thus, the capacity of lactadherin to interact with a range of Ptd-L-Ser-containing binding sites and to compete efficiently with coagulation proteins is an unusual feature.

Crystal structures have been published for the C2 domains of both factor V (fV-C2) (13) and factor VIII (fVIII-C2) (23). The common feature is a β-barrel core with three relatively long loops protruding from one end. Both fV-C2 and fVIII-C2 have 3-4 water-exposed hydrophobic residues protruding from long loops, leading to a hypothesis that membrane binding is mediated by insertion of these residues into the membrane. Functionality of these hydrophobic residues has been confirmed by site-directed mutagenesis (24, 25). The crystal structures have also provided a basis for speculation as to the amino acids that interact with the hydrophilic head group of Ptd-L-Ser. The mutagenesis experiments have not provided support for the proposed Ptd-L-Ser-binding head group binding sites. Sequence homology between lactadherin C2 domain (Lact-C2) and fV-C2/fVIII-C2 (Fig. 1A) predicts that Lact-C2 will also adopt a central β-barrel motif and have relatively long loops that display potential membrane-interactive amino acids.

Both fV-C2 and fVIII-C2 undergo conformational changes that are related to membrane binding. The conformational flexibility of fV-C2 is illustrated by the different conformation of the largest membrane-interactive loop in the three crystal structures (13). Some conformational flexibility of fVIII-C2 is illustrated by comparison of this crystal structure with the co-crystal of fVIII-C2 with a monoclonal antibody B02C11. When fVIII-C2 is in complex with B02C11, the two strands of the third loop on the membrane-interactive surface rotated ~70° (26). fVIII-C2 also undergoes a conformational change in response to ESH8, a monoclonal antibody with an epitope that does not overlap with B02C11 (27, 28). The functional consequences of ESH8-induced conformational change include increased affinity for von Willebrand factor and decreased affinity for phospholipid membrane (27, 29). The structural nature of this conformational change remains unknown. The conformational flexibility of both fVIII-C2 and fV-C2 suggests that the lactadherin C2 domain may also have conformational flexibility that could impact membrane-binding functionality.

We have determined the crystal structure of the Lact-C2. The results demonstrate a β-barrel core very similar to those of fV-C2 and fVIII-C2 and several hydrophobic amino acids on the membrane-interactive surface. However, one of the membrane-interactive loops of Lact-C2 has a markedly different structure than the corresponding loop structures of either fV-C2 or fVIII-C2. Strategically placed glycine residues appear likely to enhance local flexibility. The membrane binding involvement of the hydrophobic residues and glycines has been tested by mutagenesis studies followed by membrane binding assays. Conformational docking studies, utilizing the Autodock software package, have identified two hypothetical Ptd-L-Ser binding sites that differ from the previously proposed Ptd-L-Ser binding sites on either fVIII-C2 or fV-C2.


   EXPERIMENTAL PROCEDURES
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
RESULTS
 DISCUSSION
 REFERENCES
 
Cloning and Purification of the C2 Domain of Lactadherin—The C2 domain of bovine lactadherin was cloned by PCR from a bovine mammary gland library (Stratagene). Forward and reverse primer sequences were 5'-CCCTTGGAGACGAGTATGTGAGAC and 5'-CGGCCAGGCCTACCACTAACAG, respectively. The product was ligated into the Topo 2.1 TA cloning vector (Invitrogen) and then excised and cloned into the pET28 bacterial expression vector (Novagen). Transformed E. coli were grown in 400 ml of LB media in a 2.5-liter low profile culture flask and cultured at 37 °C with constant shaking until the A600 reached 1.0. Cultures were induced with 10 µM isopropyl 1-thio-β-D-galactopyranoside for 3 h at 25 °C with constant shaking at 175 rpm. Cells were pelleted and frozen at -20 °C overnight and then lysed in B-PER (Pierce) according to the manufacturer's protocol. Cell lysates were applied to a Ni2+-nitrilotriacetic acid Superflow column (Qiagen) and eluted with a step gradient of 4, 8, 30, 100, and 200 mM imidazole with 150 mM NaCl and 20 mM Tris, pH 7.5. Lact-C2 eluted primarily at 200 mM imidazole. Total protein of all fractions was analyzed with the micro-BCA assay (Pierce). The peak fractions were analyzed on a LabChip BioAnalyzer (Agilent Technologies, Inc.) for appropriately sized protein bands. Positive fractions were pooled and subjected to thrombin cleavage at room temperature for 2 h to remove the His6 tag (17.5 µg of thrombin/mg of total protein). Thrombin was inactivated by the addition of 3:1 molar ratio of PPACK (Calbiochem). Cleavage was assessed by LabChip BioAnalyzer analysis. The product was then concentrated on an Ultrafree-15, molecular weight cut-off 10,000 concentrator (Millipore, Inc.) and further purified by gel filtration over Sephadex G50, medium size (AP Biotech or Sigma). The resulting purified C2 domain eluted as a single peak with the LabChip BioAnalyzer with a molecular weight less than the uncleaved product by 3000, corresponding to the engineered cleavage peptide. The protein concentration was determined using the calculated molar extinction coefficient of 44,620 M-1 cm-1 at 280 nm.

Fluorescence Labeling of Lact-C2—One milligram of pure Lact-C2 was labeled with fluorescein-5-isothiocyanate (Molecular Probes) according to the manufacturer's instructions. Following the removal of free fluorescein by gel filtration followed by buffer exchange on an Ultrafree-15 concentrator, the labeling efficiency was determined by comparison of absorbance at 280 nm versus absorbance at 488 nm, correcting for the absorbance of fluorescein at 280 nm. The molar ratio of fluorescein/Lact-C2 was 1.4. The labeled Lact-C2 is referred to hereafter as Lact-C2FITC.

Liposphere Membrane Binding Assay—Glass microspheres of 1.6-µm nominal diameter (Duke Scientific, Palo Alto, CA) were cleaned, size-restricted, and covered with a phospholipid bilayer (lipospheres), as previously described (21). Lipospheres were washed three times in 0.15 M NaCl, 0.02 M Tris-HCl, 0.1% defatted bovine albumin, 10 µM egg PC as sonicated vesicles; stored at 4 °C; and used within 8 h of synthesis. Lact-C2FITC was incubated with lipospheres for 10 min at room temperature, and membrane-bound Lact-C2FITC was measured by flow cytometry. This procedure was performed on 150-µl aliquots with an approximate liposphere concentration of 1 x 106/ml using a BD Biosciences FACSCalibur flow cytometer. Data acquisition was triggered by forward light scatter with all photomultipliers in the log mode. Noise was reduced during analysis by eliminating events with forward and side scatter values different from those characteristic of the lipospheres. Geometric mean log fluorescence was converted to linear fluorescence for values depicted in the figures. Only experiments in which the fluorescence histogram indicated a log normal distribution, as judged by inspection, were analyzed quantitatively. Flow cytometry experiments were performed in 0.14 M NaCl, 0.02 M Trizma (Tris base)-HCl, 0.1% bovine albumin, pH 7.5.

Crystallographic Analysis—Crystals of lactadherin C2 were grown at 17 °C by vapor diffusion in a hanging drop against a reservoir solution containing 18% (w/v) polyethylene glycol 4000, 0.1 M Hepes, pH 7.5, 0.2 M MgCl2, and 10% isopropyl alcohol. The tiny crystals initially obtained were used subsequently as seeds to produce diffraction quality crystals.

Crystallographic data were collected on an R-axis IV image plate detector mounted on a Rigaku RU-300 rotating anode generator. Crystals were cooled to 80-100 K in a nitrogen gas stream prior to collection. The isopropyl alcohol content in the crystallization drop was sufficient for cryoprotection. Data were indexed, integrated, and scaled using the DENZO and SCALEPACK software packages (30).

Phases were provided by molecular replacement using the program EPMR (31) and coordinates from the C2 domain of factor V (Protein Data Bank entries 1CZT [PDB] and 1CZV) as the starting model. Model-building and refinement steps were carried out using the programs Modeler (32, 33), O (33), and CNS (34). Individual B factors were refined for all atoms. Data collection and refinement statistics of the crystallization work are presented in Table 1. Figures were all made with Pymol (W. L. DeLano; available on the World Wide Web).


View this table:
[in this window]
[in a new window]

  		TABLE 1
Crystallographic data and refinement statistics

 
Preparation of Lact-C2 Mutants—Oligonucleotide-directed mutagenesis was used to create an EcoRI-HindIII PCR fragment corresponding to Lact-C2 in which the codons targeted for substitution were mutated to GCX, predicting an amino acid substitution of alanine. For the third mutant, the primer pair 5'-TTTTCTCGAGCTACTACAAAAACA TGTTCGCTAGCTGGTTTCCCTAC and 5'-TTTTCTCGAGGCTGTGATCTGCTTGTTGGG was utilized to change the native sequence of WGLSAF from spike 1 to a predicted amino acid sequence of MFA. After ligation of the PCR fragments using XhoI, the mutagenized and EcoRI- and HindIII-digested fragments were then ligated into EcoRI- and HindIII-digested pET28a vector. The resulting mutants were expressed as described above. Purification differed in that the cells were sonicated in a lysis buffer of 0.7% Sarkosyl, 50 mM Na2HPO4, 17 mM NaH2PO4, 68 mM NaCl, pH 7.4; the elution buffer was 300 mM imidazole in the lysis buffer at pH 8.0. Buffer exchange was effected utilizing Ultrafree-15, molecular weight cut-off 10,000 concentrators into 20 mM ethanolamine, 25 mM NaCl, pH 9.0. Mutants were loaded onto a 1-ml HiTrap Q HP column (GE Healthcare) equilibrated in the same buffer and eluted utilizing a 20-ml continuous gradient to 1 M NaCl, 20 mM ethanolamine, pH 9.0. The mutants each eluted as a single major peak. Absorbance of ultrafiltrates from these samples at 234, 220, and 200 nm (see below) versus control spectra from Sarkosyl in the same buffer indicated that the free residual Sarkosyl concentrations were <10 µM, whereas the concentrations of all mutants were >10 µM. As a control for the purification conditions, purified Lact-C2 was transferred to the same elution buffer and repurified as described for the mutants. Lact-C2 prepared in this way was utilized for experiments depicted in Fig. 4.

Circular dichroism melting curves were obtained to confirm that the mutants were stable under experimental conditions. Buffer exchange into 50 mM NaCl, 10 mM Na2HPO4, pH 7.5, was effected utilizing Ultrafree-15, molecular weight cut-off 10,000 concentrators. Melting scans at 222 nm were performed over a temperature range of 15-80 °C in 0.1-cm path length quartz cuvettes on a Jasco J-810 spectropolarimeter equipped with a Peltier temperature regulator. The temperature change was at 1 °C/min, and data points were obtained at 1 °C increments with a 30-s equilibration time and 16-s data averaging. Melting curves were smoothed with the Savitzky-Golay algorithm, and apparent Tm values were obtained from the inflection points of melting curves.

Fluorescence Resonance Energy Membrane Binding Assay—Binding of Lact-C2 to small unilamellar vesicles was measured by fluorescence energy transfer as previously described (35). Briefly, sonicated vesicles containing 5% dansyl-phosphatidylethanolamine (Avanti%20Polar%20Lipids">Avanti Polar Lipids) were prepared by an established method (36) with the modifications that sonication was in a bath apparatus for Misonix Sonicator 3000 (6-cm sonication head) (37). Phospholipid concentration was measured as elemental phosphorous (38). Fluorescence resonance energy transfer measurements were performed utilizing an established method (35). Experiments were performed in a 5 x 5 x 30-mm quartz cuvette in a SPEX Fluorolog II fluorescence spectrophotometer at room temperature. The excitation wavelength was set at 280 nm, a PerkinElmer UV-35 filter was placed in the excitation pathway to reduce second order scatter, and a long pass filter (cut-off 450 nm) was placed in the emission pathway to further reduce scatter. Excitation and emission slit widths were set at 8 mm. Emission spectra were collected at 1-nm increments and processed electronically as indicated in the legend to Fig. 4. Energy transfer experiments with Lact-C2 mutants were performed as described for Lact-C2, except that for some experiments mutants were incubated with thrombin, as described above for Lact-C2, prior to dilution and performance of energy transfer experiments. For experiments depicted in Fig. 4B, loss of dansyl fluorescence due to sample dilution and photobleaching was measured by consecutive measurements of fluorescence emission from vesicles modified only by the addition of the volume of buffer correlating to Lact-C2. This loss, typically ~5% of fluorescence, was used as the base line for integration of the energy transfer spectra.


Figure 1
View larger version (24K):
[in this window]
[in a new window]

  		FIGURE 1.
Binding of fluorescein-labeled Lact-C2 to membranes containing Ptd-L-Ser. A, various concentrations of Lact-C2 labeled with fluorescein isothiocyanate were mixed with lipospheres for 15 min at room temperature. Binding to lipospheres was evaluated by flow cytometry. The curve represents the best fit of the hyperbolic binding function, assuming a single class of binding sites and an excess of free Lact-C2FITC versus binding sites (21). B, only vesicles containing Ptd-L-Ser compete for binding of Lact-C2 to lipospheres. Sonicated vesicles containing 5, 10, or 20% Ptd-L-Ser (PS), with the balance as PC, were mixed with lipospheres prior to the addition of 2 nM Lact-C2FITC. Binding to lipospheres was measured by flow cytometry. Data were fit using nonlinear least squares regression with Prism version 4.0 for Macintosh. Membrane composition for lipospheres was Ptd-L-Ser/phosphatidylethanolamine/PC 15:20:65. Data for A represent mean ± S.E. for two binding experiments, whereas B represents one experiment with vesicles of each lipid composition.

 
Analysis of Competition Binding Data—Competition for Lact-C2FITC by unlabeled Lact-C2 was analyzed according to the equation,

Formula(Eq.1)
where F represents the fluorescence from Lact-C2FITC binding, Fmax is the maximum fluorescence when Lact-C2FITC sites are saturated, KD-FITC is the dissociation constant for binding of Lact-C2FITC, and KD is the dissociation constant for binding of Lact-C2 or a Lact-C2 mutant to the same binding sites. Competition binding data were fitted to the equation using nonlinear least squares analysis using Prism 4.0 for Macintosh. The constants Fmax and KD-FITC were obtained from the binding analysis depicted in Fig. 1.

Computational Docking—The search for the putative Ptd-L-Ser head group binding site on Lact-C2 was performed by using the Autodock program, version 3.0 (39), and established docking approaches (40-42). Lact-C2 and Ptd-L-Ser structural files were prepared with the AutoDockTools package. For Ptd-L-Ser, all rotatable bonds were set free, and all torsion angles were released. The Lamarckian Genetic Algorithm (39) was used with random ligand starting position and default parameter settings in all docking trials. The initial search space (grid map) was confined to the region around the hydrophobic patch of Lact-C2 but of sufficient size to cover approximately half of the molecule. For each trial, at least 100 docking runs were performed, and their results were evaluated according to docking energy and frequency of recurrence. Resulting solutions in which the ligand conformation and location were within 2 Å r.m.s. deviation were treated as a single solution cluster. Histograms of cluster frequency against energy were obtained.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
DISCUSSION
 REFERENCES
 
Membrane Binding—Bovine Lactadherin C2 domain (Lact-C2) was expressed from an E. coli expression vector and thereafter processed and purified as described. Lact-C2 was labeled with fluorescein isothiocyanate (Lact-C2FITC) according to the manufacturer's directions.

The binding of Lact-C2FITC to membranes was evaluated (Fig. 1). Lact-C2FITC was incubated with lipospheres with membranes containing 15% Ptd-L-Ser. Lact-C2FITC bound saturably with a dissociation constant of 69 ± 7 nM. When the concentration of Ptd-L-Ser in the membrane was increased to 20% or decreased to 4%, the dissociation constant was 19.2 ± 1.3 and 170 ± 10 nM, respectively (data not shown), which indicated a positive relation between the membrane binding affinity of Lact-C2 and the content of Ptd-L-Ser in the membrane. Competition binding experiments indicated that the presence of Ptd-L-Ser is required for high affinity binding of Lact-C2 to phospholipid vesicles and confirmed that the affinity for binding sites on vesicles is related to Ptd-L-Ser content (Fig. 1B) as shown recently. Inclusion of phosphatidylinsositol as 5 or 10% of total phospholipids did not support Lact-C2 binding greater than PC alone (57), confirming specificity of Lact-C2 for Ptd-L-Ser.

Overall Structure—Lact-C2 was crystallized in the P212121 space group with one monomer in the asymmetric unit. High quality diffraction data were collected at 1.67 Å resolution (Table 1). Structural modeling and refinement effectively built the structure of all amino acids. Identifiable electron density was observed for all main chain atoms and most side chain atoms. The structure of Lact-C2 is very similar to the structures of fV-C2 and fVIII-C2 (13, 23), consistent with their sequence homology (Fig. 2). The Lact-C2 structure is composed largely of β strands, β turns, and random coils (Fig. 2B). Among the 19 antiparallel β strands, eight dominant strands (labeled A-H) form a distorted β-barrel. This β-barrel core motif has been conserved across the family of discoidin-type lectin homology domains from the bacterial galactose oxidase (43) to Lact-C2. The other short 11 β strands, with 3 residues each, pack either on the top of the barrel (strands 1, 5, 8, 9, and 11) or at the bottom (strands 2, 3, 4, 6, 7, and 10). The loops connecting the strands are of varying length and contribute the variable element of the structure at either end of the barrel, producing a relatively flat upper surface and an irregular lower surface (Fig. 2B). A disulfide bond connects the N-terminal Cys1 and C-terminal Cys158, analogous to the disulfide bonds in fV-C2 and fVIII-C2. The overall structure of Lact-C2 has a compact oval shape with dimensions of ~40 x 30 x 25 Å, with loop 8-E (the loop between strand 8 and E) and loop 2-3 protruding outward.

Lower Spikes—Three relatively large loops on the lower surface of Lact-C2 resemble the corresponding loops of the fV-C2 and fVIII-C2 at the same surface (Fig. 2, B and C). These loops, commonly referred to as "spikes" in the structures of fV-C2 and fVIII-C2, carry proven or hypothetical membrane-interactive, hydrophobic amino acids (13). Spike 1 (Tyr23-Trp33) of Lact-C2 includes antiparallel β strands 2 and 3 and the loop 2-3 between the two strands; spike 2 (Gln43-Asn47) is part of loop 3-4; and spike 3 (Ala78-Gln85) includes antiparallel β strands 6 and 7 and the loop 6-7 (Fig. 2, A and B). Spikes 2 and 3 of Lact-C2 adopt conformations similar to those of fV-C2 and fVIII-C2 (Fig. 2C). The other loops on the lower surface of Lact-C2 do not display solvent-exposed hydrophobic amino acids.

Alignment between Lact-C2 and FV-C2 gave main chain r.m.s. deviation values of 0.25 Å at spike 2 and 0.87 Å at spike 3. The main chain atoms in spike 1 have well defined density in the electron density map (Fig. 3), and the local conformation of spike 1 in Lact-C2 differs from that in fV-C2 and fVIII-C2 (Fig. 2C). When residues from the spike 1 are omitted from the superposition calculation, the main chain r.m.s. deviation between the structures of Lact-C2 and fV-C2 is only 0.67 Å, indicating that the significant structural differences are confined to spike 1. There are no close crystal lattice contacts on spike 1 of Lact-C2, and the spike is surrounded by some ordered solvent molecules, so the distinct conformation of this spike is not a consequence of crystal packing.

The overall structure of spike 1 (residues 23-33) is that of a β-hairpin (Figs. 2 and 3). Two hydrogen bonds from the antiparallel sheet structure and another four formed within the loop (Fig. 3) endow spike 1 of Lact-C2 with a more organized structure than that of either fV-C2 or fVIII-C2. The relative compactness of the loop gives Lact-C2 spike 1 overall dimensions similar to those of fV-C2 spike 1, which is 2 residues shorter (Fig. 2C). There are no close crystal lattice contacts on spike 1, and it is surrounded by ordered water molecules. Thus, the observed β-hairpin conformation is not a consequence of crystal packing.

Four hydrophobic residues extend from spikes 1 and 3 on the lower end of the β barrel, including Trp26, Leu28, Phe31, and Phe81 (Fig. 2B). These solvent-exposed hydrophobic residues probably induce the mobility of spikes 1 and 3, as indicated by the relatively high B factors for both the individual hydrophobic side chains and the two spikes. The average B factor for the solvent-exposed residues 25-31 of spike 1 is 35.4 for all atoms and 33.2 for main chain atoms, the highest values for any portion of the Lact-C2 structure with 5 or more consecutive residues, indicating greater flexibility of spike 1. The average B factors for spike 3 are the second highest, with values of 28.6 for all atoms and 24.6 for main chain atoms, versus an average B value of 19.3 for all protein atoms of the overall structure and 17.7 for overall protein main chain atoms. By comparison, spike 2 is less mobile, with an average B factor of 23.0 for all atoms and 21.5 for main chain atoms. All other solvent-exposed loops have B factors similar to or lower than that of spike 2. The location and mobility of these residues are similar to the solvent-exposed hydrophobic residues in the structures of fV-C2 and fVIII-C2. The side chains of these 4 residues in Lact-C2 form a hydrophobic patch (Fig. 5A). Another hydrophobic residue on top of spike 2 is not packed with this hydrophobic patch. The tips of spikes 1 and 3 contain 2 glycine residues, Gly27 and Gly82, respectively, that are not found in either fV-C2 or fVIII-C2 (Fig. 2B).

Evaluation of Membrane Binding Hydrophobic Spikes—Site-specific mutagenesis was utilized to evaluate the membrane binding contributions of amino acids on spike 1 and on spike 3 (Table 2). For the first two mutants, one or two hydrophobic residues on spike 1 or 3 were replaced by Ala residues, and the central glycine residues were also replaced by Ala. The third mutant was a chimera protein in which spike 1 of Lact-C2 was replaced by spike 1 of fVIII-C2. The mutants were purified with a modified protocol (see "Experimental Procedures") aimed at maximizing yield. Circular dichroism melting curves revealed apparent Tm values of 58, 53, 54, and 53 °C for wild type Lact-C2 and mutants 1-3, respectively. Thus, the melting temperatures were modestly perturbed by mutagenesis, and the mutants were not prone to thermal instability over the temperature range where experiments were performed.


View this table:
[in this window]
[in a new window]

  		TABLE 2
Lact-C2 mutations

 


Figure 2
View larger version (60K):
[in this window]
[in a new window]

  		FIGURE 2.
Structure of Lact-C2 and homology with fV-C2 and fVIII-C2. A, sequence alignment of bovine Lact-C2 with homologous C2 domains, made through the ESPript Web server (55). The β strands are indicated by labeled black arrows. White characters on a red background denote identical residues shared by all four sequences. Red characters on a white background indicate residues shared by two or more sequences. Green triangles identify positions where mutations in both factor VIII and factor V lead to impaired membrane binding, and blue triangles indicate impaired membrane binding only for factor VIII (24, 25). The red star denotes the residue in the conformationally disallowed region of the Ramachandran plot in crystal structures for Lact-C2, fV-C2, and fVIII-C2. Spike 1 (Tyr23-Trp33), spike 2 (Gln43-Asn47), and spike 3 (Ala78-Gln85) are underscored with black bars. The UniProtKB data base entries used in the sequence alignment are as follows: Q95114 for bovine lactadherin; Q08431 for human lactadherin; P12259 for human factor V; P00451 for human factor VIII. The indicated numbering starts with the first residue of the secreted proteins without signaling peptides. B, crystal structure of the lactadherin C2 domain, in schematic diagram representation and stereo view. Among the 19 β strands are eight major strands (magenta) of the β sandwich core and 11 small β strands (cyan) flanking the core, along with the irregular loops (green). The N-terminal Cys1 and C-terminal Cys158 are shown as sticks with the disulfide bond colored orange. The two glycine residues of spikes 1 and 3 are colored blue. Side chains of the hydrophobic residues extending from the lower spikes are depicted in yellow. C, structural alignment of Lact-C2 (magenta), fV-C2 (green; Protein Data Bank entry 1CZT), and fVIII-C2 (cyan; Protein Data Bank entry 1D7P), in backbone stereo representation. The side chain of Asn147, the sole residue with disallowed dihedral angles in Lact-C2, is shown.

 
Direct binding of Lact-C2 and mutants was detected by fluorescence resonance energy transfer (Fig. 4A). Intrinsic tryptophan residues of Lact-C2 transfer energy to dansyl-phosphatidylethanolamine incorporated into the membrane, as previously reported for membrane binding of factor VIII and factor V (35). The results confirmed that Lact-C2 binds to Ptd-L-Ser-containing vesicles. In contrast, the integrated fluorescence of the mutants was decreased by more than 90% (Fig. 4B). Interpretation of reduced energy transfer must be made in light of the fact that WGL and fVIII loop mutants each have only 5 of the 6 Trp residues of wild type Lact-C2 (Trp is the fluorescence donor; see "Discussion") These results imply that all three mutants have decreased membrane-binding affinity.

Competition binding experiments were performed to determine whether Lact-C2 mutants interact with the membrane binding sites of Lact-C2FITC (Fig. 4C). Lact-C2 competed with Lact-C2FITC with an implied dissociation constant of 48 ± 2 nM. The results indicate that the membrane binding of Lact-C2 is not substantially altered by labeling with fluorescein isothiocyanate. In contrast, none of the three mutants competed with Lact-C2FITC, even at high concentrations of 200 nM. Results from these competition assays indicate that the WGL sequence of spike 1 and the FG sequence of spike 3 contribute to high affinity membrane binding.


Figure 3
View larger version (53K):
[in this window]
[in a new window]

  		FIGURE 3.
Stereo view of Lact-C2 spike 1 structure and electron density (from simulated annealing composite omit map contoured at 1 {sigma}) with main chain atoms in CPK color and side chain atoms in solid light yellow. The magenta dashed lines indicate hydrogen bonds.

 
Computational Docking of Phosphatidyl-L-serine with Lact-C2—To search for the potential binding site(s) of the phosphatidyl-L-serine (Ptd-L-Ser) head group on Lact-C2, step-wise computational dockings were preformed utilizing Autodock 3.0 (39). Docking studies were carried out initially on Ptd-L-Ser with full-length (dipalmitoyl) acyl chains. However, the solutions were dominated by the strong hydrophobicity of the acyl chains, which would normally be immersed in the bilayer core. Models with shorter acyl chains therefore were used in subsequent docking computations. The best docking results were obtained with the short chain phospholipid, diacetyl-Ptd-L-Ser. To evaluate the overall reliability and the parameter setting of the docking approach to detect head group binding sites on phospholipid-binding proteins, docking trials with the same or a similar parameter setting were performed and assessed by using structural data from previously crystallized protein-phospholipid complexes. The results show that it is possible to predict observed head group binding sites by computationally docking model phosphatidylserine with two-carbon acyl moieties in place of longer acyl chains (see supplemental data).

The lower spikes of fV-C2 and fVIII-C2 participate in membrane binding (24, 25), and the hydrophobic patch of Lact-C2 appears in a similar position as that of the proven membrane-interactive residues in fV-C2 and fVIII-C2. Therefore, we centered the computational docking space for the Ptd-L-Ser head group on this identified membrane-interactive region of the molecule. The initial blind docking of diacetyl-Ptd-L-Ser was performed with ample search space (30 x 30 x 30 Å) covering the lower half of the Lact-C2 molecule. Solutions were found concentrated on two surfaces surrounding the hydrophobic patch of the four surfaces designated A-D (Fig. 5A). We identified 46% trial solutions on surface C and 47% on surface D in 100 docking runs.

Hence, subsequent docking runs were confined to these two surfaces with restricted search space (23 x 23 x 23 Å), including the identified membrane-interactive amino acids and surrounding surfaces. The best solutions for each of these two surfaces were designated SolC-1 and SolD-1, respectively. SolC-1 had the lowest docking energy at -8.85 cal/mol and recurred 54 times in 100 runs on surface C (Fig. 5B). In this solution, the docked ligand features six hydrogen bonds between the phosphoserine moiety of diacetyl-Ptd-L-Ser and 5 residues of Lact-C2 (Table 2). The side chains of Arg79 and Arg152 are also in position to form two very weak salt bridges with the carboxylate and phosphoryl groups of diacetyl-Ptd-L-Ser. When another 200 runs of docking were performed, the SolC-1 solution was consistently the outstanding solution with slight change in the values of energy and frequency. On surface D, the solution SolD-1 had the lowest docking energy of -7.22 cal/mol and occurred 10 times among 100 solutions. The docked ligand in SolD-1 is stabilized by two salt bridges, one between Arg148 and the phosphoryl oxygen and the other between the serine nitrogen and the side chain of Asp80, and four hydrogen bonds (Table 3 and Fig. 5C). Some other solutions, as indicated by columns 2-5 in the histogram of Fig. 5C were also examined. Among them, the solutions of columns 2 and 4 were similar to SolD-1. The solution of column 3 has slightly higher energy and a higher frequency, given the name of SolD-2. The alignment of the glycerol and phosphate moieties of SolD-2 is similar to SolD-1, and the salt bridge between Arg148 and the phosphate group is equivalent. Instead of forming another salt bridge between Asp80 and the protonated nitrogen atom of serine in SolD-1, the serine moiety of SolD-2 rotates so that its nitrogen atom forms a hydrogen bond with Asn147-O (not shown). The solution of column 5 adopts a different conformation from SolD-1, except that the protonated amino group binds at the same location as SolD-1 binds.


View this table:
[in this window]
[in a new window]

  		TABLE 3
Major polar interactions between Lact-C2 and the diacetyl-Ptd-L-Ser head group for the two best Autodock solutions

 
Ptd-D-Ser was also utilized as a docking ligand in order to evaluate the stereochemistry of the interactions (supplemental Table 1). Statistically infrequent solutions were identified on both surfaces C and D with lower docking energy than SolC-1 and SolD-1 (supplemental Figs. 2 and 3). These had markedly different docking orientations. The dominant low energy solutions had only slightly higher docking energy than SolC-1 and SolD-1, had docking orientations that were very similar, and interacted with the same amino acids. Thus, this docking approach has revealed consistent candidate docking sites for Ptd-L-Ser but has not provided a compelling explanation for stereoselective interaction of Ptd-L-Ser.


Figure 4
View larger version (20K):
[in this window]
[in a new window]

  		FIGURE 4.
Binding of Lact-C2 hydrophobic spike mutants versus Lact-C2 to membranes containing Ptd-L-Ser. A, emission spectra of phospholipid vesicles (PLV) alone (heavy solid line) or in the presence of 20 nM Lact-C2 (heavy dashed line). Control spectra are provided for buffer (dashed line) and buffer containing Lact-C2 (solid line). The dansyl emission of phospholipid vesicles increased ~4-fold in the presence of Lact-C2. The phospholipid concentration was 3.0 µM small unilamellar vesicles of composition Ptd-L-Ser/dansyl-phosphatidylethanolamine/PC 20:5:75. Experiments were performed in 150 mM NaCl, 50 mM Tris-HCl, pH 7.5, at room temperature. Results are representative of four experiments. B, integrated fluorescence emission for Lact-C2 and three mutants. Fluorescence of phospholipid vesicles alone, corrected for the measured rate of photobleaching, were subtracted from each emission spectra collected as in A. The residual emission spectra were integrated from 450 to 600 nm. The results indicate that transferred fluorescence energy for mutants WGL and FG was reduced more than 90% and more than 95% for the mutant FVIII loop. Results are from a single experiment representative of three experiments. C, competition of Lact-C2 and mutants for binding sites of Lact-C2FITC on liposphere membranes containing 15% Ptd-L-Ser. The Lact-C2FITC concentration was 24 nM. Competition by Lact-C2 (bullet) was fitted to a competition binding model (smooth line) as described. Competition by WGL ({diamond}), FG ({triangledown}), and fVIII loop ({triangleup}) was less than 10%. Depicted data are normalized mean ± S.E. for three experiments. WT, wild type.

 

   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
REFERENCES
 
We have found that Lact-C2 binds to Ptd-L-Ser-containing membranes with high affinity and that the high affinity binding correlates to a crystal structure that is very similar to those of fV-C2 and fVIII-C2. The core structures are nearly identical, and Lact-C2 also has three analogous surface loops that appear likely to participate in membrane binding. The largest difference between Lact-C2 and the other structures is in spike 1, where Lact-C2 has a longer, more structurally complex loop. Our mutagenesis data confirm that spike 1 and spike 3 participate in high affinity membrane binding.

The hydrophobic residues on the lower spikes of Lact-C2, like the lower spikes of fV-C2 and fVIII-C2, appeared likely to be involved in membrane binding (24, 25). Prior mutagenesis studies have confirmed the membrane-interactive role of four hydrophobic amino acids for factor VIII (25) and at least two amino acids for factor V (24, 44). The interactive amino acids are localized on spikes 1 and 3, with no demonstrated role for residues of spike 2. Our mutagenesis data confirm that the hydrophobic residues and/or Gly residues of spike 1 and spike 3 participate in high affinity membrane binding of Lact-C2. Thus, the lower spikes have functional similarity as well as structural similarity.

Alignment of the hydrophobic patch of Lact-C2 with the protruding hydrophobic residues of fV-C2 and fVIII-C2 suggested a structural basis for a common membrane-binding pattern (Fig. 2C). The preference for membranes with convex curvature (20) and for membranes composed of unsaturated acyl chains (45) correlates with the presence of solvent-exposed hydrophobic amino acids that may interact with lipid moieties of the bilayer. The lateral pressure in outer membrane leaflet is lowered by convex curvature or by the presence of unsaturated acyl chains (46), presumably allowing easier entry by these hydrophobic residues.

Mutants 1 and 3 each have only 5 of the 6 Trp residues of wild type Lact-C2. Thus, the capacity for fluorescence resonance energy transfer may be reduced. Indeed, the absent Trp, positioned at the apex of spike 1, may possibly have the optimal location for transfer of energy to dansyl moieties at the interfacial plane of the membrane (47). However, the dimensions of Lact-C2 predict that several Trp residues in a membrane-bound Lact-C2 domain would fall within the 21-Å Forster distance (48) from membrane dansyl groups. Thus, the extent of the decrease in energy transfer for the two mutants (Fig. 4A) suggests a significant decrease in membrane interaction even after allowing for the absence of a Trp donor residue. This interpretation is buttressed by the failure of WGL and the fVIII-loop to compete for membrane binding of Lact-C2FITC (Fig. 4C).

The structural differences between Lact-C2 spikes 1 and 3 as compared with those of fV-C2 and fVIII-C2 may explain some of the contrasting membrane-interacting properties of these proteins. A pronounced difference is seen between the membrane binding properties of Lact-C2 (Fig. 1) and fVIII-C2 (49). The affinity of Lact-C2 for membranes containing Ptd-L-Ser is ~100-fold lower than intact lactadherin (20). In contrast, the affinity of fVIII-C2 is more than 10-fold lower than Lact-C2 (49).3 Intact lactadherin is less fastidious in its binding site requirements than factor VIII or factor V and competes efficiently for Ptd-L-Ser binding sites of both proteins (18). Factor VIII and factor V do not compete efficiently with each other (21, 22), and neither competes efficiently for lactadherin binding sites.4 The reduced affinity of the Lact-C2 fVIII loop mutant suggests that the higher affinity of Lact-C2 is due to structural features unique to Lact-C2 spike 1. In particular, it is tempting to speculate that the Gly residue in spike 1 is poised to enable conformational flexibility of loop tips and hydrophobic residues in response to the configuration of membrane binding site moieties. Further studies will be required to clarify the contribution of individual amino acids and to better explain the difference in membrane binding between Lact-C2 and fVIII-C2.


Figure 5
View larger version (31K):
[in this window]
[in a new window]

  		FIGURE 5.
Computational docking solutions for diacetyl-Ptd-L-Ser on Lact-C2. A, diacetyl-Ptd-L-Ser, with hydrogen atoms removed, is shown fully extended. Shown below is the hydrophobic patch viewed from the proposed membrane-binding surface and surrounded by four surfaces, designated A-D in the docking studies. B and C, the two major solutions, SolC-1 and SolD-1, are depicted on the corresponding C and D surfaces. Shown at the top are histograms of solution frequency versus docking energy, with the outstanding solutions identified by red bars. Shown below are the corresponding conformations of the docked diacetyl-Ptd-L-Ser superimposed on the corresponding surface C or D of the Lact-C2 crystal structure. All interactions within 3.5 Å (Table 2) are indicated with orange dashed lines.

 
Amino acids that participate in hydrophilic interaction with the Ptd-L-Ser head group by fV-C2, fVIII-C2, and Lact-C2 have not been confirmed with experimental studies (24, 25). Computational docking studies in this report predict participation of two dominant interactive facets of Lact-C2. An additional test for the plausibility of these solutions is the implied location and orientation of each Ptd-L-Ser molecule relative to the membrane. We have investigated this criterion with a membrane-binding model (Fig. 6). The model is based upon the assumptions that the hydrophobic patch of Lact-C2 partitions into the bilayer. The model illustrates that both solutions would place the polar region of Ptd-L-Ser and interacting polar residues of lact-C2 in or adjacent to the interfacial region of the membrane. The Ptd-L-Ser molecules are oriented so that the acyl chains would be submerged in the hydrophobic membrane core. Thus, both solutions appear to be consistent with the usual location and orientation of Ptd-L-Ser. Attempts to test these predictions by diffusing analogues of Ptd-L-Ser into Lact-C2 crystals have, thus far, been unsuccessful. Site-directed mutagenesis studies are in progress.

The possibility that the Ptd-L-Ser-interactive sites might be conserved was considered in light of mutagenesis data from fV-C2 and fVIII-C2. For the residues involved in SolC-1, mutation of the Asn47 homologue (N44A of fVIII-C2; Fig. 2A), of the Arg152 homologue (R150A of fV-C2), and of the Arg79 homologue (K77A of fV-C2 and K76A of the fVIII-C2) failed to demonstrate a reduction in membrane binding (25). These mutagenesis data suggest that SolC-1 may not be a conserved Ptd-L-Ser interactive site for factor VIII and factor V. However, SolC-1 remains a candidate Ptd-L-Ser head group binding site for Lact-C2. Mutation of fVIII-C2 His142/Gln143, the homologues of residues Asn147/Arg148 of Lact-C2, led to decreased membrane binding.5 These data support the SolD-1 solution as possibly identifying a Ptd-L-Ser binding site that is conserved between factor VIII and lactadherin.


Figure 6
View larger version (47K):
[in this window]
[in a new window]

  		FIGURE 6.
Proposed model of Lact-C2 binding to phospholipid membrane. Lact-C2 is depicted as a green ribbon diagram. Side chains of the proposed membrane-interactive hydrophobic residues (Trp26, Leu28, Phe31, and Phe81) are displayed as stick diagrams. Ptd-L-Ser molecules are depicted as space-filling models interacting at both the Solc-1 (left) and SolD-1 (right) interactive sites. The core of the phospholipid bilayer is depicted as a light brown solid.

 


Figure 7
View larger version (32K):
[in this window]
[in a new window]

  		FIGURE 7.
Distorted turn involving Asn147. Left, Ramachandran plot for Lact-C2, with an arrow pointing to Asn147 in the disallowed region. Right, stereo view of the surrounding loop structure, with electron density from a simulated annealing composite omit map contoured at 1.2 {sigma}. The distance between the i {alpha}-carbon (Trp145) and the i + 3 {alpha}-carbon (Arg148) of the β turn is 6.24 Å (dashed green line). The hydrogen bond between Asn147-O and Asn54-N{delta}2 is shown by a red dashed line.

 
The computational docking experiments have identified several candidate residues for hydrophilic interaction with Ptd-L-Ser. The lack of strong stereoselective preference for Ptd-L-Ser versus Ptd-D-Ser has several possible explanations. First, the orientation or conformation of Ptd-L-Ser may be constrained by immersion of the long native acyl chains in the phospholipid membrane, with the resulting structure having a larger balance of favorable energy than the free diacetyl molecule. Second, the conformation of Lact-C2 may change slightly upon phospholipid interaction in a way that enhances the stereoselective preference for Ptd-L-Ser. A third possibility is that the C1 domain of lactadherin may contribute to the stereoselective interaction with Ptd-L-Ser. Interaction of the C1 domain with phospholipid membranes would be analogous to the interaction of the factor V C1 domain and the factor VIII C1 domain with phospholipid membranes (50, 51). The C1 domain may interact cooperatively with the C2 domain to enhance stereoselective preference or the C1 domain may contain determinants of stereoselectivity. Further experiments will be required to determine which alternative(s) account for the stereoselective engagement of Ptd-L-Ser by intact lactadherin.

In Lact-C2, Asn147, part of a distorted β turn (Trp145-Arg148), falls within the disallowed region of the Ramachandran plot (Fig. 7). The {varphi}, {psi} angles are 70°, -60°, respectively, and a hydrogen bond forms between Asn147-O and Asn54-N{delta}2. Analogous residues in fVIII-C2 (His142) and fV-C2 (Gln145) adopt conformations with almost the same disallowed dihedral angles (Fig. 2C). We speculate that this distorted conformation of Asn147 may play a functional role associated with membrane binding. In annexin V, for example, a transition in 2 residues from disallowed to allowed Ramachandran regions occurs as part of a conformational change associated with membrane binding (52, 53). In our SolD-1 docking solution, Arg148 is the major binding residue for Ptd-L-Ser, suggesting involvement of this turn in membrane binding. In SolD-2, a hydrogen bond forms with Asn147. It is possible that Ptd-L-Ser binding produces a more favorable local conformation involving this turn, enabling Asn147 to access an allowable conformation.

We have found that the isolated C2 domain retains a substantial fraction of the membrane-binding functionality of lactadherin. Thus, it is possible that Lact-C2 might be utilized as an anticoagulant or as a Ptd-L-Ser detection reagent. Indeed, we have found recently that Lact-C2 can identify the intracellular location of Ptd-L-Ser (57). Alternatively, Lact-C2 might be coupled with another moiety in a fusion protein to target a procoagulant or anticoagulant moiety to membranes containing Ptd-L-Ser. Because of these considerations, we are currently evaluating the membrane-binding properties of Lact-C2 in comparison with lactadherin. The crystal structure will enable rational mutagenesis of Lact-C2 to evaluate the importance of the amino acids predicted to contribute to membrane-binding properties.

During the interval between submission and publication of this report, another group published a structure of the bovine lactadherin C2 domain (54). Compared with the present structure, this published structure shows the same β-barrel core and similar overall conformation. There are a few minor conformational variations in the spike loops; however, given the difference in resolution (1.7 Å versus 2.4 Å, respectively) in these two structures it is difficult to assess whether these differences are significant or meaningful. In addition to the higher resolution data, the current report presents membrane-binding affinity, computational, and mutational data, which were not part of the previous study. In summary, the present crystal structure of Lact-C2, compared with those of fVIII-C2 and fV-C2, displays a common core structure and membrane-interactive spikes. Mutagenesis studies confirm the contribution of the membrane-interactive spikes and raise the possibility that the different conformation of Lact-C2 spike 1 may endow greater flexibility and membrane conformability than the corresponding loops of fVIII-C2 and fV-C2. Computational docking predicts the locations of two Ptd-L-Ser interactive sites. One is possibly a conserved binding site for Ptd-L-Ser on fVIII-C2. These predictions about the membrane-binding mechanism of lactadherin, factor VIII, and factor V appear to be testable utilizing site-directed mutagenesis and biophysical docking studies.


   FOOTNOTES
 
The atomic coordinates and structure factors (code 3BN6) have been deposited in the Protein Data Bank, Research Collaboratory for Structural Bioinformatics, Rutgers University, New Brunswick, NJ (http://www.rcsb.org/).

* This work was supported by a Veterans Affairs Merit Award (to G. E. G.). 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. Back

Formula The on-line version of this article (available at http://www.jbc.org) contains supplemental Table S1 and Figs. S1-S3. Back

1 To whom correspondence should be addressed: Veterans Affairs Boston Healthcare System, 1400 VFW Parkway, West Roxbury, MA 02132. Tel.: 857-203-5252; Fax: 857-203-5592; E-mail: ggilbert{at}rics.bwh.harvard.edu.

2 The abbreviations used are: Ptd-L-Ser, phosphatidyl-L-serine; Lact-C2, lactadherin C2 domain; fV-C2, factor V C2 domain; fVIII-C2, factor VIII C2 domain; SolC-1, Solution 1 on the C surface of Lact-C2; SolD-1, Solution 1 on the D surface; SolD-2, Solution 2 on the D surface; Lact-C1, lactadherin C1 domain; FITC, fluorescein isothiocyanate; PC, phosphatidylcholine. Back

3 G. E. Gilbert, D. B. Cullinan, and V. A. Novakovic, unpublished observations. Back

4 G. E. Gilbert and J. Shi, unpublished data. Back

5 G. Gilbert and S. W. Pipe, unpublished observations. Back


   ACKNOWLEDGMENTS
 
We are grateful to Profs. Christian W. Heegaard and Jan T. Rasmussen for helpful discussions and willingness to share reagents. We thank Jialan Shi for participation in the preliminary portion of this project and Patricia Price for excellent technical assistance. We also thank Jeffrey Brown for helpful advice on using Autodock.



   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
 REFERENCES
 

  1. Hvarregaard, J., Andersen, M. H., Berglund, L., Rasmussen, J. T., and Petersen, T. E. (1996) Eur. J. Biochem. 240, 628-636[Medline] [Order article via Infotrieve]
  2. Stubbs, J., Lekutis, C., Singer, K., Bui, A., Yuzuki, D., Srinivasan, U., and Parry, G. (1990) Proc. Natl. Acad. Sci. U. S. A. 87, 8417-8421[Abstract/Free Full Text]
  3. Butler, J. E., Pringnitz, D. J., Martens, C. L., and Crouch, N. (1980) Differentiation 17, 31-40[CrossRef][Medline] [Order article via Infotrieve]
  4. Haggqvist, B., Naslund, J., Sletten, K., Westermark, G. T., Mucchiano, G., Tjernberg, L. O., Nordstedt, C., Engstrom, U., and Westermark, P. (1999) Proc. Natl. Acad. Sci. U. S. A. 96, 8669-8674[Abstract/Free Full Text]
  5. Ensslin, M., Calvete, J. J., Thole, H. H., Sierralta, W. D., Adermann, K., Sanz, L., and Topfer-Petersen, E. (1995) Biol. Chem. Hoppe-Seyler 376, 733-738[Medline] [Order article via Infotrieve]
  6. Ensslin, M. A., and Shur, B. D. (2003) Cell 114, 405-417[CrossRef][Medline] [Order article via Infotrieve]
  7. Hanayama, R., Tanaka, M., Miwa, K., Shinohara, A., Iwamatsu, A., and Nagata, S. (2002) Nature 417, 182-187[CrossRef][Medline] [Order article via Infotrieve]
  8. Silvestre, J. S., Thery, C., Hamard, G., Boddaert, J., Aguilar, B., Delcayre, A., Houbron, C., Tamarat, R., Blanc-Brude, O., Heeneman, S., Clergue, M., Duriez, M., Merval, R., Levy, B., Tedgui, A., Amigorena, S., and Mallat, Z. (2005) Nat. Med. 11, 499-506[CrossRef][Medline] [Order article via Infotrieve]
  9. Couto, J. R., Taylor, M. R., Godwin, S. G., Ceriani, R. L., and Peterson, J. A. (1996) DNA Cell Biol. 15, 281-286[Medline] [Order article via Infotrieve]
  10. Andersen, M. H., Berglund, L., Rasmussen, J. T., and Petersen, T. E. (1997) Biochemistry 36, 5441-5446[CrossRef][Medline] [Order article via Infotrieve]
  11. Taylor, M. R., Couto, J. R., Scallan, C. D., Ceriani, R. L., and Peterson, J. A. (1997) DNA Cell Biol. 16, 861-869[Medline] [Order article via Infotrieve]
  12. Andersen, M. H., Graversen, H., Fedosov, S. N., Petersen, T. E., and Rasmussen, J. T. (2000) Biochemistry 39, 6200-6206[CrossRef][Medline] [Order article via Infotrieve]
  13. Macedo-Ribeiro, S., Bode, W., Huber, R., Quinn-Allen, M. A., Kim, S. W., Ortel, T. L., Bourenkov, G. P., Bartunik, H. D., Stubbs, M. T., Kane, W. H., and Fuentes-Prior, P. (1999) Nature 402, 434-439[CrossRef][Medline] [Order article via Infotrieve]
  14. Hanayama, R., and Nagata, S. (2005) Proc. Natl. Acad. Sci. U. S. A. 102, 16886-16891[Abstract/Free Full Text]
  15. Hanayama, R., Tanaka, M., Miyasaka, K., Aozasa, K., Koike, M., Uchiyama, Y., and Nagata, S. (2004) Science 304, 1147-1150[Abstract/Free Full Text]
  16. Ait-Oufella, H., Kinugawa, K., Zoll, J., Simon, T., Boddaert, J., Heeneman, S., Blanc-Brude, O., Barateau, V., Potteaux, S., Merval, R., Esposito, B., Teissier, E., Daemen, M. J., Leseche, G., Boulanger, C., Tedgui, A., and Mallat, Z. (2007) Circulation 115, 2168-2177[Abstract/Free Full Text]
  17. Boddaert, J., Kinugawa, K., Lambert, J. C., Boukhtouche, F., Zoll, J., Merval, R., Blanc-Brude, O., Mann, D., Berr, C., Vilar, J., Garabedian, B., Journiac, N., Charue, D., Silvestre, J. S., Duyckaerts, C., Amouyel, P., Mariani, J., Tedgui, A., and Mallat, Z. (2007) Am. J. Pathol. 170, 921-929[Abstract/Free Full Text]
  18. Shi, J., and Gilbert, G. E. (2003) Blood 101, 2628-2636[Abstract/Free Full Text]
  19. Shi, J., Shi, Y., Waehrens, L. N., Rasmussen, J. T., Heegaard, C. W., and Gilbert, G. E. (2006) Cytometry 69, 1193-1201[Medline] [Order article via Infotrieve]
  20. Shi, J., Heegaard, C. W., Rasmussen, J. T., and Gilbert, G. E. (2004) Biochim. Biophys. Acta 1667, 82-90[Medline] [Order article via Infotrieve]
  21. Gilbert, G. E., Drinkwater, D., Barter, S., and Clouse, S. B. (1992) J. Biol. Chem. 267, 15861-15868[Abstract/Free Full Text]
  22. Gilbert, G. E., Sims, P. J., Wiedmer, T., Furie, B., Furie, B. C., and Shattil, S. J. (1991) J. Biol. Chem. 266, 17261-17268[Abstract/Free Full Text]
  23. Pratt, K. P., Shen, B. W., Takeshima, K., Davie, E. W., Fujikawa, K., and Stoddard, B. L. (1999) Nature 402, 439-442[CrossRef][Medline] [Order article via Infotrieve]
  24. Kim, S. W., Quinn-Allen, M. A., Camp, J. T., Macedo-Ribeiro, S., Fuentes-Prior, P., Bode, W., and Kane, W. H. (2000) Biochemistry 39, 1951-1958[CrossRef][Medline] [Order article via Infotrieve]
  25. Gilbert, G. E., Kaufman, R. J., Arena, A. A., Miao, H., and Pipe, S. W. (2002) J. Biol. Chem. 277, 6374-6381[Abstract/Free Full Text]
  26. Spiegel, P. C., Jr., Jacquemin, M., Saint-Remy, J. M., Stoddard, B. L., and Pratt, K. P. (2001) Blood 98, 13-19[Abstract/Free Full Text]
  27. Saenko, E. L., Shima, M., Gilbert, G. E., and Scandella, D. (1996) J. Biol. Chem. 271, 27424-27431[Abstract/Free Full Text]
  28. Saenko, E. L., Shima, M., and Sarafanov, A. G. (1999) Trends Cardiovasc. Med. 9, 185-192[CrossRef][Medline] [Order article via Infotrieve]
  29. Saenko, E. L., Scandella, D., Yakhyaev, A. V., and Greco, N. J. (1998) J. Biol. Chem. 273, 27918-27926[Abstract/Free Full Text]
  30. Otwinowski, Z., and Minor, W. (1997) Methods Enzymol. 27, 307-326
  31. Kissinger, C. R., Gehlhaar, D. K., and Fogel, D. B. (1999) Acta Crystallogr. D Biol. Crystallogr. 55, 484-491[CrossRef][Medline] [Order article via Infotrieve]
  32. Marti-Renom, M. A., Stuart, A. C., Fiser, A., Sanchez, R., Melo, F., and Sali, A. (2000) Annu. Rev. Biophys. Biomol. Struct. 29, 291-325[CrossRef][Medline] [Order article via Infotrieve]
  33. Jones, T. A., Zou, J. Y., Cowan, S. W., and Kjeldgaard. (1991) Acta Crystallogr. A. 47, 110-119[CrossRef]
  34. Brunger, A. T., Adams, P. D., Clore, G. M., DeLano, W. L., Gros, P., Grosse-Kunstleve, R. W., Jiang, J. S., Kuszewski, J., Nilges, M., Pannu, N. S., Read, R. J., Rice, L. M., Simonson, T., and Warren, G. L. (1998) Acta Crystallogr. Sect. D Biol. Crystallogr. 54, 905-921[CrossRef][Medline] [Order article via Infotrieve]
  35. Gilbert, G. E., Furie, B. C., and Furie, B. (1990) J. Biol. Chem. 265, 815-822[Abstract/Free Full Text]
  36. Barenholz, Y., Gibbes, D., Litman, B. J., Goll, J., Thompson, T. E., and Carlson, F. D. (1977) Biochemistry 16, 2806-2810[CrossRef][Medline] [Order article via Infotrieve]
  37. Johnson, S. M., Bangham, A. D., Hill, M. W., and Korn, E. D. (1971) Biochim. Biophys. Acta 233, 820-826[Medline] [Order article via Infotrieve]
  38. Chen, P., Toribara, T., and Warner, H. (1956) Anal. Chem. 28, 1756-1758
  39. Morris, G. M., Goodsell, D. S., Halliday, R. S., Huey, R., Hart, W. E., Belew, R. K., and Olson, A. J. (1998) J. Comp. Chem. 19, 1639-1662[CrossRef]
  40. Fodor, K., Harmat, V., Hetenyi, C., Kardos, J., Antal, J., Perczel, A., Patthy, A., Katona, G., and Graf, L. (2005) J. Mol. Biol. 350, 156-169[CrossRef][Medline] [Order article via Infotrieve]
  41. Kovacs, M., Toth, J., Hetenyi, C., Malnasi-Csizmadia, A., and Sellers, J. R. (2004) J. Biol. Chem. 279, 35557-35563[Abstract/Free Full Text]
  42. Hetenyi, C., and van der Spoel, D. (2002) Protein Sci. 11, 1729-1737[CrossRef][Medline] [Order article via Infotrieve]
  43. Baumgartner, S., Hofmann, K., Chiquet-Ehrismann, R., and Bucher, P. (1998) Protein Sci. 7, 1626-1631[Medline] [Order article via Infotrieve]
  44. Peng, W., Quinn-Allen, M. A., Kim, S. W., Alexander, K. A., and Kane, W. H. (2004) Biochemistry 43, 4385-4393[CrossRef][Medline] [Order article via Infotrieve]
  45. Gilbert, G. E., and Arena, A. A. (1998) Biochemistry 37, 13526-13535[CrossRef][Medline] [Order article via Infotrieve]
  46. Cantor, R. S. (1999) Biophys. J. 76, 2625-2639[Medline] [Order article via Infotrieve]
  47. Abbott, A., and Nelsestuen, G. (1987) Biochemistry 26, 7994-8003[CrossRef][Medline] [Order article via Infotrieve]
  48. Wu, P., and Brand, L. (1994) Anal. Biochem. 218, 1-13[CrossRef][Medline] [Order article via Infotrieve]
  49. Takeshima, K., Smith, C., Tait, J., and Fujikawa, K. (2003) Thromb. Haemost. 89, 788-794[Medline] [Order article via Infotrieve]
  50. Saleh, M., Peng, W., Quinn-Allen, M. A., Macedo-Ribeiro, S., Fuentes-Prior, P., Bode, W., and Kane, W. H. (2004) Thromb. Haemost. 91, 16-27[Medline] [Order article via Infotrieve]
  51. Gilbert, G. E., Novakovic, V. A., Cullinan, D. B., and Jacquemin, M. (2007) J. Thromb. Haemost. 5, Suppl. 1, P-S-O42
  52. Concha, N. O., Head, J. F., Kaetzel, M. A., Dedman, J. R., and Seaton, B. A. (1993) Science 261, 1321-1324[Abstract/Free Full Text]
  53. Sopkova, J., Renouard, M., and Lewit-Bentley, A. (1993) J. Mol. Biol. 234, 816-825[CrossRef][Medline] [Order article via Infotrieve]
  54. Lin, L., Huai, Q., Huang, M., Furie, B., and Furie, B. C. (2007) J. Mol. Biol. 371, 717-724[CrossRef][Medline] [Order article via Infotrieve]
  55. Gouet, P., Courcelle, E., Stuart, D. I., and Metoz, F. (1999) Bioinformatics 15, 305-308[Abstract/Free Full Text]
  56. Laskowski, R. A., MacArthur, M. W., Moss, D. S., and Thornton, J. M. (1993) J. Appl. Crystallogr. 26, 283-291[CrossRef]
  57. Yeung, T., Gilbert, G. E., Shi, J., Silvius, J., Kapus, A., and Grinstein, S. (2008) Science 319, 210-213[Abstract/Free Full Text]

Add to CiteULike CiteULike   Add to Complore Complore   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us   Add to Digg Digg   Add to Reddit Reddit   Add to Technorati Technorati    What's this?


This article has been cited by other articles:


Home page
 J. Cell Sci.Home page
A. S. Raymond and B. D. Shur
A novel role for SED1 (MFG-E8) in maintaining the integrity of the epididymal epithelium
J. Cell Sci., March 15, 2009; 122(6): 849 - 858.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Supplemental Data
Right arrow All Versions of this Article:
283/11/7230    most recent
M705195200v1
Right arrow Submit a Letter to Editor
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrowRequest Permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Shao, C.
Right arrow Articles by Gilbert, G. E.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Shao, C.
Right arrow Articles by Gilbert, G. E.
Social Bookmarking
Add to CiteULike   Add to Complore   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Reddit   Add to Technorati  
What's this?

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
 All ASBMB Journals   Molecular and Cellular Proteomics 
 Journal of Lipid Research   ASBMB Today 
Copyright © 2008 by the American Society for Biochemistry and Molecular Biology.
Advertisement
spacer
Advertisement
Advertisement