Advertisement
JBC

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


     

Originally published In Press as doi:10.1074/jbc.M801282200 on May 7, 2008

J. Biol. Chem., Vol. 283, Issue 28, 19440-19447, July 11, 2008
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Supplemental Data
Right arrow All Versions of this Article:
283/28/19440    most recent
M801282200v1
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 Google Scholar
Google Scholar
Right arrow Articles by Ui, M.
Right arrow Articles by Tsumoto, K.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Ui, M.
Right arrow Articles by Tsumoto, K.
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?

How Protein Recognizes Ladder-like Polycyclic Ethers

INTERACTIONS BETWEEN CIGUATOXIN (CTX3C) FRAGMENTS AND ITS SPECIFIC ANTIBODY 10C9*Formula

Mihoko Ui{ddagger}1, Yoshikazu Tanaka{ddagger}, Takeshi Tsumuraya§, Ikuo Fujii§, Masayuki Inoue**, Masahiro Hirama, and Kouhei Tsumoto{ddagger}2||

From the {ddagger}Department of Medical Genome Sciences, Graduate School of Frontier Sciences, the University of Tokyo, Kashiwa 277-8562, Chiba, the **Graduate School of Pharmaceutical Sciences, the University of Tokyo, Bunkyo-ku 113-003, Tokyo, the §Department of Biological Science, Graduate School of Science, Osaka Prefecture University, Sakai 599-8570, Osaka, the Department of Chemistry, Graduate School of Science, Tohoku University, and ||Solution-Oriented Research for Science and Technology (SORST), Japan Science and Technology Agency, Sendai, 980-8578, Japan

Received for publication, February 19, 2008 , and in revised form, April 7, 2008.


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 REFERENCES
 
Ciguatoxins are a family of marine toxins composed of transfused polycyclic ethers. It has not yet been clarified at the atomic level on the pathogenic mechanism of these toxins or the interaction between a polycyclic ether compounds and a protein. Using the crystal structures of anti-ciguatoxin antibody 10C9 Fab in ligand-free form and in complexes with ABCD-ring (CTX3C-ABCD) and ABCDE-ring (CTX3C-ABCDE) fragments of the antigen CTX3C at resolutions of 2.6, 2.4, and 2.3Å, respectively, we elucidated the mechanism of the interaction between the polycyclic ethers and the antibody. 10C9 Fab has an extraordinarily large and deep binding pocket at the center of the variable region, where CTX3C-ABCD or CTX3C-ABCDE binds longitudinally in the pocket via hydrogen bonds and van der Waals interactions. Upon antigen-antibody complexation, 10C9 Fab adjusts to the antigen fragments by means of rotational motion in the variable region. In addition, the antigen fragment lacking the E-ring induces a large motion in the constant region. Consequently, the thermostability of 10C9 Fab is enhanced by 10 °C upon complexation with CTX3C-ABCDE but not with CTX3C-ABCD. The crystal structures presented in this study also show that 10C9 Fab recoginition of CTX3C antigens requires molecular rearrangements over the entire antibody structure. These results further expand the fundamental understanding of the mechanism by which ladder-like polycyclic ethers are recognized and may be useful for the design of novel therapeutic agents by antibodies, marine toxins, or new diagnostic reagents for the detection and targeting of members of the polycyclic ether family.


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 REFERENCES
 
Ciguatoxins are a family of ladder-like polycyclic ethers that causes ciguatera seafood poisoning in tropical and subtropical regions; more than 50,000 people suffer annually from this type of food poisoning (1–7). CTX3C, which contains 13 ether rings, is probably the best known the member of polycyclic ether family. Although CTX3C is known to bind to voltage-sensitive sodium channels, like other polycyclic ether toxins (8–10), the precise pathogenic mechanism remains to be elucidated, and methods for the treatment of ciguatera poisoning have not yet been developed (5).

Oguri et al. (11) established a direct sandwich enzyme-linked immunosorbent assay using two distinct monoclonal antibodies (10C9 and 3D11) to detect CTX3C at the parts per billion level with no cross-reactivity against other related marine toxins. 10C9 IgG is prepared by immunization of mice with a carrier protein (keyhole limpet hemocyanin (KLH)3) conjugated to the ABCDE-ring of CTX3C. 10C9 IgG binds not only to CTX3C-ABCDE but also to full-length CTX3C (Fig. 1), with dissociation constants of 0.8 and 2.8 nM, respectively. Because the skeletons of polycyclic ethers are so rigid and because the compounds are so large (12–15) (they extend beyond the binding surface of the paratope of the antibody (200–400 Å2) (16–21)), how the antibody recognizes the polycyclic ethers within the limited areas of the antigen-binding site for the compounds is not clear. Elucidation of the recognition mechanism is important for the development of therapeutic or diagnostic agents for ciguatera, and furthermore, analytical reagents and catalysts for the synthesis of other polyether-like compounds.

We focused our investigation on anti-ciguatoxin antibody 10C9 and the mechanism by which it recognizes the ABCD-ring and ABCDE-ring fragments of CTX3C (Fig. 1). For that purpose, we determined the crystal structures of 10C9 Fab in the ligand-free form and in complex with CTX3C-ABCD or CTX3C-ABCDE. Comparison of the three structures revealed that recognition of the CTX3C fragments by 10C9 Fab involves three features: a large, deep antigen-binding pocket that captures the fragments, polar residues that form hydrogen bonds with the antigens, and rotational conformational changes that stabilize the complexes.


Figure 1
View larger version (17K):
[in this window]
[in a new window]

  		FIGURE 1.
Structures of ciguatoxin fragments: CTX3C-ABCDE-KLH (1), CTX3C-ABCDE (2), CTX3C-ABCD (3), and CTX3C (4).

 

   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS AND DISCUSSION
 REFERENCES
 
Preparation of Antigen Fragments CTX3C-ABCD and CTX3C-ABCDE—The left-hand portions of CTX3C (CTX3C-ABCD and CTX3C-ABCDE) were synthesized according to previously described methods (22, 23). CTX3C-ABCDE was conjugated to KLH to yield ABCDE-KLH for immunization of mice to prepare the 10C9 monoclonal antibody (11).

Preparation of 10C9 Fab—10C9 IgG was prepared as described previously (11). 10C9 Fab was obtained by papain digestion of 10C9 IgG and purified on a Protein A column (Fab preparation kit; Pierce) followed by cation exchange chromatography (Resource S column; GE Healthcare) and size exclusion chromatography (HiLoadTM 26/60 Superdex 200; GE Healthcare). The purity and homogeneity of the fragment were evaluated by means of SDS-PAGE.

Crystallization of Ligand-free 10C9 Fab and Its Complexes with CTX3C-ABCD and CTX3C-ABCDE—Purified 10C9 Fab was dialyzed against a stock buffer (10 mM Tris-HCl, pH 8.0) and then concentrated to 8 mg ml-1 for crystallization trials. For preparation of the CTX3C-ABCD·10C9 Fab complex, the appropriate amount of a 500 µM CTX3C-ABCD solution in DMSO was added to the protein solution (final 10C9 Fab/CTX3C-ABCD molar ratio, 1:1.7), and for preparation of the CTX3C-ABCDE·10C9 Fab complex, a 5 mM CTX3C-ABCDE solution in DMSO was added to the protein solution (final 10C9 Fab/CTX3C-ABCDE molar ratio, 1:1.2). The mixtures were subjected to static incubation at room temperature for 1 h prior to crystallization experiment. The initial crystallization conditions were screened by the sparse matrix method at 20 °C using Crystal Screen and Crystal Screen 2 kits (Hampton Research) and Wizard I and Wizard II screening kits (Emerald Biostructures, Bainbridge Island, WA). Crystals of ligand-free 10C9 Fab suitable for data collection were grown from 0.1 M sodium cacodylate (pH 5.2) containing 0.2 M zinc acetate and 15% polyethylene glycol 8000 by the hanging drop vapor diffusion method. Crystals of the CTX3C-ABCD·10C9 Fab and CTX3C-ABCDE·10C9 Fab complexes were grown from 0.1 M sodium acetate (pH 4.6) containing 0.2 M ammonium sulfate and 30% polyethylene glycol monomethyl ether 2000 and from 0.1 M trisodium citrate dehydrate (pH 5.6) containing 20% (v/v) isopropyl alcohol and 20% polyethylene glycol 4000, respectively.

Diffraction Data Collection and Processing—X-ray diffraction experiments were performed at the Photon Factory (Tsukuba, Japan). The data set for crystals of ligand-free 10C9 Fab was collected on beamline NW12 under cryogenic conditions (100 K), and the data sets for the complexes with CTX3C-ABCD and CTX3C-ABCDE were collected on beamlines 6a and 5a, respectively. Diffraction data for the ligand-free fragment and the two complexes were collected to a maximum resolutions of 2.6, 2.4, and 2.3 Å, respectively. Data were indexed, integrated, and scaled with the HKL-2000 program package (24). Data collection and processing statistics are summarized in supplemental Table 1.

Structure Determination and Refinement—The structures of 10C9 Fab in the ligand-free state and in complexes with CTX3C-ABCD or CTX3C-ABCDE were determined by the molecular replacement method using the program PHASER (25). The heavy and light chains of 10C9 Fab (Protein Data Bank codes 1HQ4 and 1Z3G, respectively) were used as search models for the heavy and light chains of 10C9 Fab in complex with CTX3C-ABCD. Then, the heavy and light chains of 10C9 Fab in complex with CTX3C-ABCD were subsequently used as search models for 10C9 Fab in the ligand-free state and in complex with CTX3C-ABCDE. The structural coordinates of CTX3C-ABCD and CTX3C-ABCDE were modeled manually with the graphical program XtalView (26). The coordination file, the topology file, and the parameter file for CTX3C-ABCD and CTX3C-ABCDE were prepared with the Dundee PRODRG2 Server (available on the World Wide Web). To monitor the quality of refinement, a random 10% subset of all the reflections was set aside for calculation of the Rfree-factor. The positional and individual B-factor refinements of structures in the ligand-free form and in complex with CTX3C-ABCDE were carried out with the program CNS (27). The refinement of the structure in complex with CTX3C-ABCD was carried out with the program REFMAC5 (28). Finally, the water molecules were added automatically, and then ligand molecules were placed manually. The crystallographic R and Rfree values of 10C9 Fab in the ligand-free state and in complexes with CTX3C-ABCDE and CTX3C-ABCD converged to 22.0 and 27.0%, 22.8 and 26.9%, and 29.8 and 33.9%, respectively. The stereochemical quality of the final refined models was analyzed with the program PROCHECK (29). The refinement statistics are summarized in supplemental Table 1.

Structural Comparison between Ligand-free Fab and Its Complexes—The three-dimensional structures of the CTX3C-ABCD·10C9 Fab and CTX3C-ABCD·10C9 Fab complexes were superimposed on the structure of ligand-free 10C9 Fab with the LSQKAB program (part of the CCP4 suite of programs) (30). The root mean square (r.m.s.) deviations between the complexes and ligand-free 10C9 Fab were calculated with the COMPARE program (also part of the CCP4 suite) (31). The three-dimensional structures were generated by the PyMol program, which provides a high quality rendering of three-dimensional structure (available on the World Wide Web).

Isothermal Titration Microcalorimetry—The thermodynamic parameters of the interactions between 10C9 Fab and the two antigen fragments were determined by isothermal titration microcalorimetry with a VP-ITC Micro Calorimeter (MicroCal Inc.). The antigen fragments (~5 µM in phosphate-buffered saline) were titrated with a 50 µM solution of 10C9 Fab in the same buffer at 25 °C. Thermogram data were analyzed by means of the ORIGIN program (MicroCal).

Differential Scanning Calorimetry—The thermostabilities of 10C9 Fab in the ligand-free state and in complexes with CTX3C-ABCD and CTX3C-ABCDE were evaluated with a VP-DSC Capillary Cell Micro Calorimeter (MicroCal Inc.). 10C9 Fab (20 µM in 10 mM Tris-HCl buffer, pH 8.0) was used for the measurements. The antigen fragments were separately dissolved in DMSO, and the solutions were added to the protein solution (final 10C9 Fab/antigen molar ratio, 1:1.2). The temperature value at the peak partial molar heat capacity was taken as the Tm value for the species.


   RESULTS AND DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
REFERENCES
 
Crystal Structures of 10C9 Fab in the Ligand-free Form and in Complexes with CTX3C-ABCD and CTX3C-ABCDE—The crystal structure of ligand-free 10C9 Fab was determined by the molecular replacement method at a resolution of 2.6 Å (supplemental Table 1). The structure shows that 10C9 Fab possess an extraordinarily large and deep pocket (depth, 11 Å) in the interface between the variable heavy and light chains (VH and VL). Aromatic and hydrophobic amino acid residues, such as L-Tyr36,4 H-Ile37, L-Trp91, H-Trp47, and H-Trp103 are present on the surface of the pocket, providing a hydrophobic environment for CTX3C binding. The extraordinary large antigen-binding pocket probably stabilizes CTX3C by sequestering it away from solvent. There are also charged residues, such as L-His34, H-His35A, H-Asn58, L-Arg46, and L-Gln89, in the pocket. Two 10C9 Fab molecules are arranged in an asymmetric unit in the crystal. Some atoms of the side chains in the complementarity determining regions (CDRs) of the heavy chain (CDR-H1, from H-Thr30 to H-Gly32; CDR-H2, from H-His52 to H-Arg54; and CDR-H3, from H-Phe97 to H-Ser99) are disordered and could not be precisely located in either molecule of the asymmetric unit, although atoms of the main chains could be located. These results suggest that the some regions of the CDR loops are flexible in the ligand-free molecule (32, 33). In addition, residues from H-Ser128 to H-Asn133 in the CH1 domain were completely disordered due to the flexibility of the loop.

The crystal structures of the 10C9 Fab complexes with CTX3C-ABCD and CTX3C-ABCDE were determined at resolutions of 2.4 and 2.3 Å, respectively (supplemental Table 1). In the CTX3C-ABCD·10C9 Fab complex, two molecules are arranged in the asymmetric unit. The electron density map of the variable region of one of the molecules was too poorly defined to build a model due to the high flexibility of the protein. Nevertheless we were able to construct a model of the other molecule, although the residues from L-Ile150 to L-Gln156, from L-Thr180 to L-Tyr192, from H-Gly127 to H-Val136, and from H-Pro184 to H-Ser186 in the constant region of this molecule are also disordered. These structural flexibilities caused high R- and Rfree-factors in the structural refinement procedure. Although we could not build a complete structural model of the constant regions, one of the 10C9 Fab molecule in complex with CTX3C-ABCD was well defined. It is noteworthy to mention that the electron density maps of the CDR loops in the CTX3C-ABCD complex were well defined, whereas the electron density maps of the same regions in the ligand-free form were significantly no islet. This result suggests that the CDR loops are fixed upon binding to CTX3C-ABCD.

For the CTX3C-ABCDE·10C9 Fab complex, we constructed a structural model for all of the residues except for those in the region (from H-Gly127 to H-Asn133), which is completely disordered. Note that the electron density map of the CDRs was well defined in the CTX3C-ABCDE·10C9 Fab complex, as was the case for the CTX3C-ABCD complex. These results suggest that the flexible loops of the CDR are stabilized upon complexation with the antigen fragments.

The structures of 10C9 Fab in ligand-free form and in complexes with CTX3C-ABCD and with CTX3C-ABCDE are shown in Fig. 2. Both antigen fragments are longitudinally bound within the pocket with the A-ring directed toward the bottom of the pocket. CTX3C-ABCD and CTX3C-ABCDE are buried inside the pocket to a length of ~10 Å, and the surface areas buried within the cavity are 90 and 80%, respectively. A tail of CTX3C-ABCDE protrudes from the cavity. In ligand-free 10C9 Fab, the pocket is present at the center of the variable region prior to antigen binding, although the entrance of the pocket is very narrow. The antigen-binding pocket is expanded by flipping of the side chain of L-Trp91 at the entrance of the pocket. The crystal structures of both complexes indicate that the indole ring flips around to open the gate for the antigens and provide a large hydrophobic space for them to bind (supplemental Fig. 1).

Antigen-Antibody Interactions—In the CTX3C-ABCD·10C9 Fab and CTX3C-ABCDE·10C9 Fab complexes, the antigen fragments are positioned identically within the binding pocket (Fig. 3). The numbers of water molecules in the binding pocket are equal for the three crystal structures. Although the positions of the water molecules are not the same in the three 10C9 Fab structures, there are two hydrated water molecules at the interface between the antigen fragment and the antibody in the crystal structures of the complexes, and they bridge the space between the antigen fragments and 10C9 Fab by forming hydrogen bonds.


Figure 2
View larger version (48K):
[in this window]
[in a new window]

  		FIGURE 2.
Top views (top) and cross-section of the antigen-binding site (bottom) of variable regions of 10C9 Fab in ligand-free form (a), in complex with CTX3C-ABCDE (b), and in complex with CTX3C-ABCD (c). Green, heavy chain of 10C9 Fab; blue, light chain of 10C9 Fab; magenta, L-Trp91 residue in the ligand-free antibody; red, L-Trp91 residue in the complexes; yellow, carbon atoms in the antigens; red, oxygen atoms in the antigens. The figures were generated with the PyMol program.

 


Figure 3
View larger version (42K):
[in this window]
[in a new window]

  		FIGURE 3.
Close-up view of the binding site of 10C9 Fab in ligand-free form (a) in complex with CTX3C-ABCDE (b) and in complex with CTX3C-ABCD (c). Green, C{alpha} atom backbone of the heavy chain; cyan, backbone of the light chain; yellow, antigens. The carbon, oxygen, and nitrogen atoms in the residues surrounding the antigens are colored gray, red, and blue, respectively. Hydrogen bonds are indicated with dashed lines, and the numbers indicate distances between atoms.

 
In both complexes, the side chains facing the antigen fragments are almost identical in conformation and orientation, except for H-Asn58. Fig. 4 shows a close-up view of the entrance to the antigen-binding pocket in the two complexes. In the CTX3C-ABCDE·10C9 Fab complex, a hydrogen bond between Of of CTX3C-ABCDE and H-Asn58-NH of 10C9 Fab creates a network of hydrogen bonds between H-Asn58, H-Tyr59, and L-Asn94. These interactions are believed to stabilize the complex. In contrast, the network of hydrogen bonds between H-Asn58, H-Tyr59, and L-Asn94 is not present in the CTX3C-ABCD·10C9 Fab complex, due to rotation of the side chain of H-Asn58. The presence of the hydrogen bond between H-Asn58 and CTX3C-ABCDE may explain why the enthalpy of binding is more favorable with CTX3C-ABCDE than with CTX3C-ABCD (Table 1).


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

  		TABLE 1
Thermodynamic parameters of the interactions between 10C9 Fab and the antigen fragments N, stoichiometry; Ka, binding constant; {Delta}G, {Delta}H, and {Delta}S, changes in Gibbs energy, binding enthalpy, and entropy, respectively.

 
The interactions between atoms in CTX3C-ABCD and 10C9 Fab and between atoms in CTX3C-ABCDE and 10C9 Fab were evaluated on the basis of the three-dimensional structures (supplemental Fig. 2, supplemental Table 2). Van der Waals interactions and hydrogen bonds were involved in the direct antigen-antibody interactions in both complexes. The number of hydrogen bonds at the interface between CTX3C-ABCD and 10C9 Fab is nearly equal to the number at the interface between CTX3C-ABCDE and 10C9 Fab; there is a weak hydrogen bond between the nitrogen atom of the L-Arg46 side chain in 10C9 Fab and the Oa atom of CTX3C-ABCD (bond distance, 3.4 Å) that is not present in the other complex. Although the numbers of residues interacting with the antigen fragments in the two complexes are nearly identical, the total number of interactions of 10C9 Fab with CTX3C-ABCDE (70 interactions) is larger than the number of interactions with CTX3C-ABCD (55 interactions), which indicates that the conformation of the complex with CTX3C-ABCDE is more complementary than that of the complex with CTX3C-ABCD. In particular, the number of interactions with L-Tyr36 and L-Pro96 (located in framework L2 and CDR L3, respectively) is substantially higher for CTX3C-ABCDE than for CTX3C-ABCD. In contrast, the contribution of L-Trp91 to the antigen-antibody interaction is larger in the CTX3C-ABCD·10C9 Fab complex than in the CTX3C-ABCDE·10C9 Fab complex. L-Trp91 interacts mainly with the C- and D-rings of the antigen fragments through van der Waals interactions, which may contribute to keep each antigen fragment on the paratope.

Note that not all of the amino acid residues of the flexible loops of the CDRs, which are disordered in ligand-free 10C9 Fab, interact directly with either antigen fragment, although the residues at the bottom of the flexible loops do form direct interactions. Taking all these results together, we conclude that the two antigen fragments are bound in 10C9 Fab in an identical manner, although there are some differences in the number of interactions and in the extent of the hydrogen-bonding network.

To investigate how 10C9 was matured, we compared the rearranged sequence of 10C9 with that of the corresponding germ line gene (supplemental Fig. 3). There are seven (H-Gln3, H-Asn34, H-Arg54, H-Thr56, H-Thr61, H-Ser73, and H-Cys94) and eight mutated residues (L-Glu1, L-Leu2, L-Met4, L-Thr7, L-Ser32, L-Val33, L-Pro55, and L-Gly60) in variable regions of heavy and light chains, respectively. Among them, three residues (i.e. H-Asn34, H-Thr61, and L-Pro55) are located in the vicinity of the ligand binding site (supplemental Fig. 4). Although they do not interact with the antigen directly, it is possible that these residues contribute to ligand binding. H-Asn34 would play an important role in capturing the antigen via hydrogen bonding with a water molecule that interacts with both CTX3C-ABCD and CTX3C-ABCDE in the antigen binding pocket. Since H-Thr61 forms a hydrogen bond with L-Asn94, it is plausible that H-Thr61 contributes by stabilizing the VH-VL interaction. L-Pro55 seems to fix the loop structure of CDR-L2, which forms the external wall of the antigen binding pocket.

Stabilization of the 10C9 Fab Complexes—We carried out an isothermal titration calorimetry study of the associations between 10C9 Fab and CTX3C-ABCD and CTX3C-ABCDE (supplemental Fig. 5). Thermodynamic parameters calculated from the titration curves are presented in Table 1. The isothermal titration calorimetry study indicated that both CTX3C-ABCD and CTX3C-ABCDE bind to 10C9 Fab with a 1:1 stoichiometry, a result that agrees with the results of the x-ray crystallographic analysis. The association constants for CTX3C-ABCD and CTX3C-ABCDE were estimated to be 1.50 x 107 M-1 and 9.02 x 107 M-1, respectively, indicating that absence of the E-ring reduces the binding affinity by a factor of six. The binding enthalpies for CTX3C-ABCD and CTX3C-ABCDE were estimated to be -45.7 and -68.4 kJ/mol, respectively, suggesting that the interaction between 10C9 Fab and CTX3C-ABCDE is more enthalpically favorable than that between 10C9 Fab and CTX3C-ABCD. In contrast, the binding entropies for CTX3C-ABCD and CTX3C-ABCDE to 10C9 Fab were determined to be -0.015 and -0.076 kJ mol-1 K-1, respectively, suggesting that the interaction between 10C9 Fab and CTX3C-ABCDE is accompanied by a larger entropic loss due to binding. Although the interactions are enthalpy-driven in both cases, the binding enthalpy for the interaction with CTX3C-ABCDE is ~20 kJ/mol larger than that with CTX3C-ABCD, which suggests that the enthalpic contribution is significant for CTX3C-ABCDE. In general, favorable entropy increases are caused by the release of water molecules from the interface between two bound species (34). Considering that binding of the antigen fragments did not change the number nor the arrangement of water molecules in the binding pocket, the entropy difference appears to be due to differences in the conformational degrees of freedom in the two complexes (35) (i.e. there are more degrees of freedom in the complex with CTX3C-ABCD than in the complex with CTX3C-ABCDE). These results suggest that the enhanced binding affinity of the longer antigen (bearing the E-ring) is due to a favorable enthalpy change.


Figure 4
View larger version (42K):
[in this window]
[in a new window]

  		FIGURE 4.
Hydrogen-bonding interactions in the CTX3C-ABCDE·10C9 Fab (a) and CTX3C-ABCD·10C9 Fab (b) complexes. Green, heavy chain of 10C9 Fab; cyan, light chain of 10C9 Fab; red, oxygen atoms; blue, nitrogen atoms; yellow, CTX3C fragments. Hydrogen bonds are indicated with dashed lines, and the numbers indicate the distances between atoms.

 


Figure 5
View larger version (22K):
[in this window]
[in a new window]

  		FIGURE 5.
Superposition of complexed (blue) and ligand-free (red) 10C9 Fab; top views of the variable regions the CTX3C-ABCDE·10C9 Fab complex (a) and the CTX3C-ABCD·10C9 Fab complex (b).

 
To investigate the increases in stability upon ligand binding, we determined the thermostabilities of the CTX3C-ABCD·10C9 Fab and CTX3C-ABCDE·10C9 Fab complexes and ligand-free 10C9 Fab by differential scanning calorimetry (supplemental Fig. 6). Ligand-free 10C9 Fab exhibited a thermogram with a peak at ~78 °C, which is a typical Tm value for antigen-binding fragments (36, 37). The Tm value for the CTX3C-ABCD·10C9 Fab complex is nearly identical to that of the ligand-free form, whereas the Tm value for the CTX3C-ABCDE·10C9 Fab complex is substantially higher (86.8 °C). These results indicate that the stability of 10C9 Fab is markedly increased upon binding of CTX3C-ABCDE but not CTX3C-ABCD. The interaction of 10C9 Fab with the E-ring may be what leads to the increased thermostability of 10C9 Fab.

Induced Fitting Mechanism for the Antigen Fragments—We investigated the conformational changes that 10C9 Fab undergoes upon complexation with the antigen fragments to clarify the difference between the thermostabilities of the two complexes. Fig. 5 shows the structural differences in the variable regions of 10C9 Fab in the ligand-free state and in complexes with the antigen fragments. 10C9 Fab clearly undergoes rotational motion to accommodate the large, rigid polycyclic ethers, and there is a major difference in the degree of movement for the two complexes. In fact, 10C9 Fab seems to require a large motion to capture CTX3C-ABCD. To evaluate the difference between the conformational changes for the two complexes, we calculated the r.m.s. deviations by superimposing the C{alpha} atoms of the VH, VL, CH1, and CL domains of ligand-free 10C9 Fab with those of the CTX3C-ABCD·10C9 Fab and CTX3C-ABCDE·10C9 Fab complexes (supplemental Table 3). When the VH domain of the CTX3C-ABCD·10C9 Fab complex was superimposed on that of the ligand-free form, the r.m.s. deviation of the VL domain was ~2.5 times that of the VH domain itself, which indicates marked conformational changes in the variable region. In contrast, when the VH domain of the CTX3C-ABCDE·10C9 Fab complex was superimposed on that of the ligand-free form, the r.m.s. deviation of the VL domain was only 1.9 times that of the VH domain itself. These results indicate that 10C9 Fab undergoes a large rotational movement of the entire variable region to recognize CTX3C-ABCD but only a moderate movement to recognize CTX3C-ABCDE.

To investigate the overall conformational changes in 10C9 Fab caused by binding to the antigen fragments, we also investigated the constant regions of 10C9 Fab. When the CH1 and CL domains of the CTX3C-ABCD·10C9 Fab complex were superimposed on the respective domains of the ligand-free form, the r.m.s. deviation of the complementary domain was more than 2 times that of the corresponding domain itself. We believe that the structural change in the variable region of 10C9 Fab upon complexation with CTX3C-ABCD causes a large conformational change in the constant region. In contrast, in the case of the CTX3C-ABCDE·10C9 Fab complex, no significant increase in the r.m.s. deviations of the complementary domains was observed. This result suggests that the lower thermostability of the CTX3C-ABCD·10C9 Fab complex relative to that of the CTX3C-ABCDE·10C9 Fab complex is due to the marked conformational change that occurs in the constant region as a result of the large movement in the variable region of 10C9 Fab upon complexation with CTX3C-ABCD.

Contribution of the E-ring to the Stability of the Antigen-Antibody Complex—The crystallographic temperature factors of the C{alpha} atoms in the heavy chains of ligand-free 10C9 Fab and the CTX3C-ABCD·10C9 Fab and CTX3C-ABCDE·10C9 Fab complexes were investigated (supplemental Fig. 7). Because the constant region in the light chain of the CTX3C-ABCD·10C9 Fab complex is disordered, the temperature factor of the light chain could not be compared with the factors for ligand-free 10C9 Fab and the CTX3C-ABCDE·10C9 Fab complex. In both complexes, the B-factor values of the variable regions are lower than the value for the ligand-free 10C9 Fab. In ligand-free 10C9 Fab, there are several regions whose temperature factors are particularly high; all of the regions are CDRs: CDR-H1, Leu20–Thr30; CDR-H2, Tyr50–Arg54, and CDR-H3, Asp95–Asp101. This fact indicates that these loops are flexible (38). However, these temperature factors are markedly reduced in both complexes, which suggests that the CDRs are well stabilized upon complexation. A large difference in B-factor values between the CTX3C-ABCD·10C9 Fab and CTX3C-ABCDE·10C9 Fab complexes was observed in the constant region (i.e. the temperature factors in the constant region of the CTX3C-ABCD·10C9 Fab complex are markedly increased upon ligand binding, whereas those in the CTX3C-ABCDE complex are decreased). These results suggest that the entire constant region in the CTX3C-ABCD·10C9 Fab complex is perturbed and destabilized compared with the identical region of ligand-free 10C9 Fab, although the constant region of the CTX3C-ABCDE·10C9 Fab complex is well stabilized. The difference between the thermostabilities of CTX3C-ABCD and CTX3C-ABCDE complexes (supplemental Fig. 6) is due to destabilization in the constant region of the CTX3C-ABCD complex. Therefore, it is plausible that the structural difference between the antigen fragments (i.e. the presence or absence of the E-ring) induces large differences in the structure and stability of the complex.

Recognition Mechanism for Ladder-like Polycyclic Ether Compounds—Because 10C9 Fab was prepared by immunization of mice with KLH-conjugated CTX3C-ABCDE (Fig. 1) (11), the ligand-binding pocket of 10C9 Fab can be expected to be more suitable for CTX3C-ABCDE than for CTX3C-ABCD, and this expectation is confirmed with the results of our thermodynamic analyses. There are five common features of the interactions between 10C9 Fab and the two antigen fragments: the extraordinarily large and deep antigen-binding pocket, some hydrogen bonds and numerous van der Waals interactions in the antigen-binding site, the contribution of favorable enthalpy upon binding, rotational movement of the variable region upon complexation, and lowering of the B-factor values of the variable region of the complexes. The difference between the two antigen fragments with regard to the antigen-antibody interactions upon complexation is in the degree of structural rearrangement of 10C9 Fab. Upon complexation with CTX3C-ABCD or CTX3C-ABCDE, the entrance to the binding pocket of 10C9 Fab expands, and the structure of the protein adjusts to the antigen fragments by means of a rotational motion in the variable region, which extends to the constant region in the case of CTX3C-ABCD. Conformational rearrangement of 10C9 Fab provides the well established interactions with the antigen fragments. The hydrogen bond between the E-ring of CTX3C-ABCDE and H-Asn58 of 10C9 Fab is important for stabilizing the antigen-antibody complex; the hydrogen bond forms part of an intramolecular hydrogen-bonding network that contributes to stable complex formation (Fig. 4). In fact, the thermostability of 10C9 Fab is enhanced by 10 °C upon complexation with CTX3C-ABCDE, and the B-factor values of the heavy chain decrease over the whole structure. In contrast, the incompleteness of the interaction with CTX3C-ABCD, due to the absence of the E-ring, induces a large structural rearrangement of 10C9 Fab in both the variable region and the constant region, which results in increases in the B-factor values of the constant region of the heavy chain. The contribution of the enthalpy upon binding is smaller for CTX3C-ABCD than for CTX3C-ABCDE, and this difference may originate from the distortion of the steric structure of 10C9 Fab in complex with CTX3C-ABCD.


Figure 6
View larger version (37K):
[in this window]
[in a new window]

  		FIGURE 6.
The mechanism by which 10C9 Fab recognizes CTX3C-ABCDE (upper) and CTX3C-ABCD (lower). Pink, CTX3C-ABCDE; orange, CTX3C-ABCD; green, variable region of 10C9 Fab; blue, constant region of 10C9 Fab.

 
Taking all of our results together, we propose that 10C9 Fab recognizes ladder-like polycyclic ethers by means of the mechanism outlined in Fig. 6. 10C9 Fab contains a large, deep pocket to accommodate its large antigens. When CTX3C-ABCDE is inserted into the pocket, it imposes a rotational movement of the variable region of 10C9 Fab, and this movement stabilizes the complex. To bind to the smaller antigen (CTX3C-ABCD), 10C9 Fab must undergo a larger movement in the variable region to recognize the antigen, and this larger movement results in additional structural changes in the constant region, which destabilize the constant region. Our elucidation of the mechanism of the interaction between 10C9 Fab and antigen fragments should contribute to the fundamental understanding of the molecular recognition of polycyclic ether compound by antibodies.

Conclusion—In this report, we present the first atomic level description of the mechanical system by which the antibody for polycyclic ether compounds recognizes its antigens. Anti-ciguatoxin antibody 10C9 Fab undergoes a conformational change in the variable region with rotational movement to accommodate the antigens, and this rotational movement sometimes induces additional motion in the constant region, depending on the antigen. The keys for the recognition of a polycyclic ether compound possessing a large, rigid structure like that of CTX3C are the presence of a suitable pocket that can form contacts with the majority of the surface of the antigens, and the presence of polar residues that can form hydrogen bonds with the antigens. We expect our findings to be useful for the development of novel therapeutic agents, analytical reagents, and detection systems for ciguatoxin-like polycyclic ether compounds and for the development of functional antibody-based materials in the fields of chemistry and protein engineering.


   FOOTNOTES
 
The atomic coordinates and structure factors (codes 2Z91, 2Z92, and 2Z93) have been deposited in the Protein Data Bank, Research Collaboratory for Structural Bioinformatics, Rutgers University, New Brunswick, NJ (http://www.rcsb.org/).

* The research was in part supported by the Solution-Oriented Research for Science and Technology project from the Japan Science and Technology Agency, Japan. 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 Tables 1–3 and Figs. 1–7. Back

1 Supported by Research Fellowships of the Japan Society for the Promotion of Science for Young Scientists. Back

2 To whom correspondence should be addressed. Tel.: 81-4-7136-3600; Fax: 81-4-7136-3601; E-mail: tsumoto{at}k.u-tokyo.ac.jp.

3 The abbreviations used are: KLH, keyhole limpet hemocyanin; Fab, antigen-binding fragment of immunoglobulin; CDR, complementarity-determining region; VH, variable region of immunoglobulin heavy chain; VL, variable region of immunoglobulin light chain; CH1, constant region of immunoglobulin heavy chain; CL, constant region of immunoglobulin light chain; r.m.s., root mean square. Back

4 Residues are denoted in the following manner: H-Asn58, the asparagine 58 (Kabat number) of the 10C9 Fab heavy chain. Back


   ACKNOWLEDGMENTS
 
We thank Dr. Sato and the Cell Science and Technology Institute, Inc. staff for valuable technical assistance.



   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 REFERENCES
 

  1. Yasumoto, T., and Murata, M. (1993) Chem. Rev. 93, 1897-1909[CrossRef]
  2. Yasumoto, T. (2001) Chem. Rec. 1, 228-242[CrossRef][Medline] [Order article via Infotrieve]
  3. Hirama, M. (2005) Chem. Record 5, 240-250[CrossRef]
  4. Levine, D. Z. (1995) J. Am. Osteopath. Assoc. 95, 193-198[Abstract]
  5. Isbister, G. K., and Kiernan, M. C. (2005) Lancet Neurol. 4, 219-228[CrossRef][Medline] [Order article via Infotrieve]
  6. Lewis, R. J. (2006) Toxicon 48, 799-809[Medline] [Order article via Infotrieve]
  7. Lewis, R. J. (2001) Toxicon 39, 97-106[Medline] [Order article via Infotrieve]
  8. Poli, M. A., Mende, T. J., and Baden, D. G. (1986) Mol. Pharmacol. 30, 129-135[Abstract]
  9. Lombet, A., Bidard, J. N., and Lazdunski, M. (1987) FEBS Lett. 219, 355-359[CrossRef][Medline] [Order article via Infotrieve]
  10. Wang, S. Y., and Wang, G. K. (2003) Cell. Signal. 15, 151-159[CrossRef][Medline] [Order article via Infotrieve]
  11. Oguri, H., Hirama, M., Tsumuraya, T., Fujii, I., Maruyama, M., Uehara, H., and Nagumo, Y. (2003) J. Am. Chem. Soc. 125, 7608-7612[CrossRef][Medline] [Order article via Infotrieve]
  12. Satake, M., Murata, M., and Yasumoto, T. (1993) J. Am. Chem. Soc. 115, 361-362[CrossRef]
  13. Satake, M., Murata, M., and Yasumoto, T. (1993) Tetrahedron Lett. 34, 1975-1978[CrossRef]
  14. Shimizu, Y., Chou, H. N., Bando, H., Vanduyne, G., and Clardy, J. C. (1986) J. Am. Chem. Soc. 108, 514-515[CrossRef]
  15. Lin, Y. Y., Risk, M., Ray, S. M., Vanengen, D., Clardy, J., Golik, J., James, J. C., and Nakanishi, K. (1981) J. Am. Chem. Soc. 103, 6773-6775[CrossRef]
  16. Davies, D. R., Padlan, E. A., and Sheriff, S. (1990) Annu. Rev. Biochem. 59, 439-473[CrossRef][Medline] [Order article via Infotrieve]
  17. Arevalo, J. H., Taussig, M. J., and Wilson, I. A. (1993) Nature 365, 859-863[CrossRef][Medline] [Order article via Infotrieve]
  18. Jeffrey, P. D., Schildbach, J. F., Chang, C. Y., Kussie, P. H., Margolies, M. N., and Sheriff, S. (1995) J. Mol. Biol. 248, 344-360[Medline] [Order article via Infotrieve]
  19. Tanaka, F., Kinoshita, K., Tanimura, R., and Fujii, I. (1996) J. Am. Chem. Soc. 118, 2332-2339[CrossRef]
  20. Romesberg, F. E., Spiller, B., Schultz, P. G., and Stevens, R. C. (1998) Science 279, 1929-1933[Abstract/Free Full Text]
  21. Xu, J. A., Deng, Q. L., Chen, J. G., Houk, K. N., Bartek, J., Hilvert, D., and Wilson, I. A. (1999) Science 286, 2345-2348[Abstract/Free Full Text]
  22. Hirama, M., Oishi, T., Uehara, H., Inoue, M., Maruyama, M., Guri, H., and Satake, M. (2001) Science 294, 1904-1907[Abstract/Free Full Text]
  23. Oishi, T., Tanaka, S., Ogasawara, Y., Maeda, K., Oguri, H., and Hirama, M. (2001) Synlett 2001, 952-954[CrossRef]
  24. Otwinowski, Z., and Minor, W. (1997) Macromol. Crystallogr. A 276, 307-326[CrossRef]
  25. Read, R. J. (2001) Acta Crystallogr. Sect. D Biol. Crystallogr. 57, 1373-1382[CrossRef][Medline] [Order article via Infotrieve]
  26. McRee, D. E. (1999) J. Struct. Biol. 125, 156-165[CrossRef][Medline] [Order article via Infotrieve]
  27. 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. 54, 905-921
  28. Murshudov, G. N., Vagin, A. A., and Dodson, E. J. (1997) Acta Crystallogr. Sect. D Biol. Crystallogr. 53, 240-255[CrossRef][Medline] [Order article via Infotrieve]
  29. Laskowski, R. A., Macarthur, M. W., Moss, D. S., and Thornton, J. M. (1993) J. Appl. Crystallogr. 26, 283-291[CrossRef]
  30. Kabsch, W. (1976) Acta Crystallogr. Sect. A 32, 922-923[CrossRef]
  31. (1994) Acta Crystallogr. 50, 760-763
  32. Mian, I. S., Bradwell, A. R., and Olson, A. J. (1991) J. Mol. Biol. 217, 133-151[CrossRef][Medline] [Order article via Infotrieve]
  33. Vasudevan, S., Celikel, R., Ruggeri, Z. M., Varughese, K. I., and Kunicki, T. J. (2004) Blood Cells Mol. Dis. 32, 176-181[CrossRef][Medline] [Order article via Infotrieve]
  34. Braden, B. C., Goldman, E. R., Mariuzza, R. A., and Poljak, R. J. (1998) Immunol. Rev. 163, 45-57[CrossRef][Medline] [Order article via Infotrieve]
  35. Finkelstein, A. V., and Janin, J. (1989) Protein Engineering 3, 1-3[Free Full Text]
  36. Garber, E., and Demarest, S. J. (2007) Biochem. Biophys. Res. Commun. 355, 751-757[CrossRef][Medline] [Order article via Infotrieve]
  37. Demarest, S. J., Chen, G., Kimmel, B. E., Gustafson, D., Wu, J., Salbato, J., Poland, J., Elia, M., Tan, X., Wong, K., Short, J., and Hansen, G. (2006) Protein Eng. Des. Sel. 19, 325-336[Abstract/Free Full Text]
  38. Haruyama, H., Ito, S., Miyadai, K., Takahashi, T., Kawaida, R., Takayama, T., Hanzawa, H., Hata, T., Yamaguchi, J., Yoshida-Kato, H., Ichikawa, K., Ohsumi, J., Yonehara, S., and Serizawa, N. (2002) Biol. Pharm. Bull. 25, 1537-1545[CrossRef][Medline] [Order article via Infotrieve]

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
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Supplemental Data
Right arrow All Versions of this Article:
283/28/19440    most recent
M801282200v1
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 Google Scholar
Google Scholar
Right arrow Articles by Ui, M.
Right arrow Articles by Tsumoto, K.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Ui, M.
Right arrow Articles by Tsumoto, K.
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