Creation of Cross-Linked Crystals With Intermolecular Disulfide Bonds Connecting Symmetry-Related Molecules Allows Retention of Tertiary Structure in Different Solvent Conditions.

Protein crystals are generally fragile and sensitive to subtle changes such as pH, ionic strength, and/or temperature in their crystallization mother liquor. Here, using T4 phage lysozyme as a model protein, the three-dimensional rigidification of protein crystals was conducted by introducing disulfide cross-links between neighboring molecules in the crystal. The effect of cross-linking on the stability of the crystals was evaluated by microscopic observation and X-ray diffraction. When soaking the obtained cross-linked crystals into a precipitant-free solution, the crystals held their shape without dissolution and diffracted to approximately 1.1 Å resolution, comparable to that of the non-cross-linked crystals. Such cross-linked crystals maintained their diffraction even when immersed in other solutions with pH values from 4 to 10, indicating that the disulfide cross-linking made the packing contacts enforced and resulted in some mechanical strength in response to changes in the preservation conditions. Furthermore, the cross-linked crystals gained stability to permit soaking into solutions containing high concentrations of organic solvents. The results suggest the possibility of obtaining protein crystals for effective drug screening by introducing appropriate cross-linked disulfide bonds.

An insight into the thermodynamic characteristics of human thrombopoietin complexation with TN1 antibody.

Human thrombopoietin (hTPO) primarily stimulates megakaryocytopoiesis and platelet production and is neutralized by the mouse TN1 antibody. The thermodynamic characteristics of TN1 antibody-hTPO complexation were analyzed by isothermal titration calorimetry (ITC) using an antigen-binding fragment (Fab) derived from the TN1 antibody (TN1-Fab). To clarify the mechanism by which hTPO is recognized by TN1-Fab the conformation of free TN1-Fab was determined to a resolution of 2.0 Å using X-ray crystallography and compared with the hTPO-bound form of TN1-Fab determined by a previous study. This structural comparison revealed that the conformation of TN1-Fab does not substantially change after hTPO binding and a set of 15 water molecules is released from the antigen-binding site (paratope) of TN1-Fab upon hTPO complexation. Interestingly, the heat capacity change (ΔCp) measured by ITC (-1.52 ± 0.05 kJ mol(-1)  K(-1) ) differed significantly from calculations based upon the X-ray structure data of the hTPO-bound and unbound forms of TN1-Fab (-1.02 ∼ 0.25 kJ mol(-1)  K(-1) ) suggesting that hTPO undergoes an induced-fit conformational change combined with significant desolvation upon TN1-Fab binding. The results shed light on the structural biology associated with neutralizing antibody recognition.

Structural basis for acceptor-substrate recognition of UDP-glucose: anthocyanidin 3-O-glucosyltransferase from Clitoria ternatea.

UDP-glucose: anthocyanidin 3-O-glucosyltransferase (UGT78K6) from Clitoria ternatea catalyzes the transfer of glucose from UDP-glucose to anthocyanidins such as delphinidin. After the acylation of the 3-O-glucosyl residue, the 3'- and 5'-hydroxyl groups of the product are further glucosylated by a glucosyltransferase in the biosynthesis of ternatins, which are anthocyanin pigments. To understand the acceptor-recognition scheme of UGT78K6, the crystal structure of UGT78K6 and its complex forms with anthocyanidin delphinidin and petunidin, and flavonol kaempferol were determined to resolutions of 1.85 Å, 2.55 Å, 2.70 Å, and 1.75 Å, respectively. The enzyme recognition of unstable anthocyanidin aglycones was initially observed in this structural determination. The anthocyanidin- and flavonol-acceptor binding details are almost identical in each complex structure, although the glucosylation activities against each acceptor were significantly different. The 3-hydroxyl groups of the acceptor substrates were located at hydrogen-bonding distances to the Nε2 atom of the His17 catalytic residue, supporting a role for glucosyl transfer to the 3-hydroxyl groups of anthocyanidins and flavonols. However, the molecular orientations of these three acceptors are different from those of the known flavonoid glycosyltransferases, VvGT1 and UGT78G1. The acceptor substrates in UGT78K6 are reversely bound to its binding site by a 180° rotation about the O1-O3 axis of the flavonoid backbones observed in VvGT1 and UGT78G1; consequently, the 5- and 7-hydroxyl groups are protected from glucosylation. These substrate recognition schemes are useful to understand the unique reaction mechanism of UGT78K6 for the ternatin biosynthesis, and suggest the potential for controlled synthesis of natural pigments.

Crystal structure of UDP-glucose:anthocyanidin 3-O-glucosyltransferase from Clitoria ternatea.

Flowers of the butterfly pea (Clitoria ternatea) accumulate a group of polyacylated anthocyanins, named ternatins, in their petals. The first step in ternatin biosynthesis is the transfer of glucose from UDP-glucose to anthocyanidins such as delphinidin, a reaction catalyzed in C. ternatea by UDP-glucose:anthocyanidin 3-O-glucosyltransferase (Ct3GT-A; AB185904). To elucidate the structure-function relationship of Ct3GT-A, recombinant Ct3GT-A was expressed in Escherichia coli and its tertiary structure was determined to 1.85 Å resolution by using X-ray crystallography. The structure of Ct3GT-A shows a common folding topology, the GT-B fold, comprised of two Rossmann-like ß/α/ß domains and a cleft located between the N- and C-domains containing two cavities that are used as binding sites for the donor (UDP-Glc) and acceptor substrates. By comparing the structure of Ct3GT-A with that of the flavonoid glycosyltransferase VvGT1 from red grape (Vitis vinifera) in complex with UDP-2-deoxy-2-fluoro glucose and kaempferol, locations of the catalytic His-Asp dyad and the residues involved in recognizing UDP-2-deoxy-2-fluoro glucose were essentially identical in Ct3GT-A, but certain residues of VvGT1 involved in binding kaempferol were found to be substituted in Ct3GT-A. These findings are important for understanding the differentiation of acceptor-substrate recognition in these two enzymes.

Highly efficient method of preparing human catalytic antibody light chains and their biological characteristics.

The ultimate goal of catalytic antibody research is to develop new patient therapies that use the advantages offered by human catalytic antibodies. The establishment of a high-throughput method for obtaining valuable candidate catalytic antibodies must be accelerated to achieve this objective. In this study, based on our concept that we can find antibody light chains with a high probability of success if they include a serine protease-like catalytic triad composed of Ser, His, and Asp on a variable region of the antibody structure, we amplified and cloned DNAs encoding human antibody light chains from germline genes of subgroup II by seminested PCR using two primer sets designed for this purpose. Seven DNA fragments encoding light chains in 17 clones were derived from germline gene A18b, 6 DNA fragments from A3/A19, 2 DNA fragments from A17, and a clone DNA fragment from A5 and O11/O1. All light chains expressed in Escherichia coli and highly purified under nondenaturing conditions exhibited amidolytic activity against synthetic peptides. Some of the light chains exhibited unique features that suppressed the infectious activity of the rabies virus. Furthermore, the survival rate of mice in which a lethal level of the rabies virus was coinoculated directly into the brain with light chain 18 was significantly improved. In the case of humans, these results demonstrate that high-throughput selection of light chains possessing catalytic functions and specificity for a target molecule can be attained from a light-chain DNA library amplified from germline genes belonging to subgroup II.

Distinct structural requirements for interleukin-4 (IL-4) and IL-13 binding to the shared IL-13 receptor facilitate cellular tuning of cytokine responsiveness.

Both interleukin-4 (IL-4) and IL-13 can bind to the shared receptor composed of the IL-4 receptor alpha chain and the IL-13 receptor alpha1 chain (IL-13Ralpha1); however, the mechanisms by which these ligands bind to the receptor chains are different, enabling the principal functions of these ligands to be different. We have previously shown that the N-terminal Ig-like domain in IL-13Ralpha1, called the D1 domain, is the specific and critical binding unit for IL-13. However, it has still remained obscure which amino acid has specific binding capacity to IL-13 and why the D1 domain acts as the binding site for IL-13, but not IL-4. To address these questions, in this study we performed mutational analyses for the D1 domain, combining the structural data to identify the amino acids critical for binding to IL-13. Mutations of Lys-76, Lys-77, or Ile-78 in c' strand in which the crystal structure showed interaction with IL-13, and those of Trp-65 and Ala-79 adjacent to the interacting site, resulted in significant impairment of IL-13 binding, demonstrating that these amino acids generate the binding site. Furthermore, mutations of Val-35, Leu-38, or Val-42 at the N-terminal beta-strand also resulted in loss of IL-13 binding, probably from decreased structural stability. None of the mutations employed here affected IL-4 binding. These results demonstrate that the D1 domain of IL-13Ralpha1 acts as an affinity converter, through direct cytokine interactions, that allows the shared receptor to respond differentially to IL-4 and IL-13.

Structure of HIV-1 protease in complex with potent inhibitor KNI-272 determined by high-resolution X-ray and neutron crystallography.

HIV-1 protease is a dimeric aspartic protease that plays an essential role in viral replication. To further understand the catalytic mechanism and inhibitor recognition of HIV-1 protease, we need to determine the locations of key hydrogen atoms in the catalytic aspartates Asp-25 and Asp-125. The structure of HIV-1 protease in complex with transition-state analog KNI-272 was determined by combined neutron crystallography at 1.9-A resolution and X-ray crystallography at 1.4-A resolution. The resulting structural data show that the catalytic residue Asp-25 is protonated and that Asp-125 (the catalytic residue from the corresponding diad-related molecule) is deprotonated. The proton on Asp-25 makes a hydrogen bond with the carbonyl group of the allophenylnorstatine (Apns) group in KNI-272. The deprotonated Asp-125 bonds to the hydroxyl proton of Apns. The results provide direct experimental evidence for proposed aspects of the catalytic mechanism of HIV-1 protease and can therefore contribute substantially to the development of specific inhibitors for therapeutic application.

Expression of the extracellular region of the human interleukin-4 receptor alpha chain and interleukin-13 receptor alpha1 chain by a silkworm-baculovirus system.

The receptor binding to interleukin (IL)-13 is composed of the IL-13 receptor alpha1 chain (IL-13Ralpha1) and the IL-4 receptor alpha chain (IL-4Ralpha). In order to investigate the interaction of IL-13 with IL-13Ralpha1 and IL-4Ralpha, the DNA fragments coding the extracellular regions of human IL-13Ralpha1 and the IL-4Ralpha (containing a cytokine receptor homologous region) were fused with mouse Fc and expressed by a silkworm-baculovirus system. The expressed receptors were successfully purified by affinity chromatography using protein A, and the Fc region was removed by thrombin digestion. After further purification with anion-exchange chromatography, these receptors were used to investigate the ligand-receptor interaction. Size exclusion chromatography and SPR analysis revealed that mixture of IL-13 and IL-13Ralpha1 showed predominant affinity to IL-4Ralpha, although neither detectable affinity of IL-13 nor IL-13Ralpha1 was observed against IL-4Ralpha. Combining these data with the moderate affinity of IL-13 to IL-13Ralpha1, this indicates that IL-13 first binds to IL-13Ralpha1 and recruits consequently to IL-4R.

Mutagenesis of the crystal contact of acidic fibroblast growth factor.

An attempt has been made to improve a crystal contact of human acidic fibroblast growth factor (haFGF; 140 amino acids) to control the crystal growth, because haFGF crystallizes only as a thin-plate form, yielding crystals suitable for X-ray but not neutron diffraction. X-ray crystal analysis of haFGF showed that the Glu81 side chain, located at a crystal contact between haFGF molecules, is in close proximity with an identical residue related by crystallographic symmetry, suggesting that charge repulsion may disrupt suitable crystal-packing interactions. To investigate whether the Glu residue affects the crystal-packing interactions, haFGF mutants in which Glu81 was replaced by Ala, Val, Leu, Ser and Thr were constructed. Although crystals of the Ala and Leu mutants were grown as a thin-plate form by the same precipitant (formate) as the wild type, crystals of the Ser and Thr mutants were grown with increased thickness, yielding a larger overall crystal volume. X-ray structural analysis of the Ser mutant determined at 1.35 A resolution revealed that the hydroxy groups of Ser are linked by hydrogen bonds mediated by the formate used as a precipitant. This approach to engineering crystal contacts may contribute to the development of large protein crystals for neutron crystallography.

'Crystal lattice engineering,' an approach to engineer protein crystal contacts by creating intermolecular symmetry: crystallization and structure determination of a mutant human RNase 1 with a hydrophobic interface of leucines.

A protein crystal lattice consists of surface contact regions, where the interactions of specific groups play a key role in stabilizing the regular arrangement of the protein molecules. In an attempt to control protein incorporation in a crystal lattice, a leucine zipper-like hydrophobic interface (comprising four leucine residues) was introduced into a helical region (helix 2) of the human pancreatic ribonuclease 1 (RNase 1) that was predicted to form a suitable crystallization interface. Although crystallization of wild-type RNase 1 has not yet been reported, the RNase 1 mutant having four leucines (4L-RNase 1) was successfully crystallized under several different conditions. The crystal structures were subsequently determined by X-ray crystallography by molecular replacement using the structure of bovine RNase A. The overall structure of 4L-RNase 1 is quite similar to that of the bovine RNase A, and the introduced leucine residues formed the designed crystal interface. To characterize the role of the introduced leucine residues in crystallization of RNase 1 further, the number of leucines was reduced to three or two (3L- and 2L-RNase 1, respectively). Both mutants crystallized and a similar hydrophobic interface as in 4L-RNase 1 was observed. A related approach to engineer crystal contacts at helix 3 of RNase 1 (N4L-RNase 1) was also evaluated. N4L-RNase 1 also successfully crystallized and formed the expected hydrophobic packing interface. These results suggest that appropriate introduction of a leucine zipper-like hydrophobic interface can promote intermolecular symmetry for more efficient protein crystallization in crystal lattice engineering efforts.

Homodimeric cross-over structure of the human granulocyte colony-stimulating factor (GCSF) receptor signaling complex.

A crystal structure of the signaling complex between human granulocyte colony-stimulating factor (GCSF) and a ligand binding region of GCSF receptor (GCSF-R), has been determined to 2.8 A resolution. The GCSF:GCSF-R complex formed a 2:2 stoichiometry by means of a cross-over interaction between the Ig-like domains of GCSF-R and GCSF. The conformation of the complex is quite different from that between human GCSF and the cytokine receptor homologous domain of mouse GCSF-R, but similar to that of the IL-6/gp130 signaling complex. The Ig-like domain cross-over structure necessary for GCSF-R activation is consistent with previously reported thermodynamic and mutational analyses.

Characterization of the interaction between interleukin-13 and interleukin-13 receptors.

Interleukin-13 (IL-13) possesses two types of receptor: the heterodimer, composed of the IL-13Ralpha1 chain (IL-13Ralpha1) and the IL-4Ralpha chain (IL-4Ralpha), transducing the IL-13 signals; and the IL-13Ralpha2 chain (IL-13Ralpha2), acting as a nonsignaling "decoy" receptor. Extracellular portions of both IL-13Ralpha1 and IL-13Ralpha2 are composed of three fibronectin type III domains, D1, D2, and D3, of which the last two comprise the cytokine receptor homology modules (CRHs), a common structure of the class I cytokine receptor superfamily. Thus far, there has been no information about the critical amino acids of the CRHs or the role of the D1 domains of IL-13Ralpha1 and IL-13Ralpha2 in binding to IL-13. In this study, we first built the homology modeling of the IL-13.hIL-13 receptor complexes and then predicted the amino acids involved in binding to IL-13. By incorporating mutations into these amino acids, we identified Tyr-207, Asp-271, Tyr-315, and Asp-318 in the CRH of human IL-13Ralpha2, and Leu-319 and Tyr-321 in the CRH of human IL-13Ralpha1, as critical residues for binding to IL-13. Tyr-315 in IL-13Ralpha2 and Leu-319 in IL-13Ralpha1 are positionally conserved hydrophobic amino acid residues. Furthermore, by using D1 domain-deleted mutants, we found that the D1 domain is needed for the expression of IL-13Ralpha2, but not IL-13Ralpha1, and that the D1 domain of IL-13Ralpha1 is important for binding to IL-13, but not to IL-4. These results provide the basis for a precise understanding of the interaction between IL-13 and its receptors.

Crystallization of a 2:2 complex of granulocyte-colony stimulating factor (GCSF) with the ligand-binding region of the GCSF receptor.

The granulocyte-colony stimulating factor (GCSF) receptor receives signals for regulating the maturation, proliferation and differentiation of the precursor cells of neutrophilic granulocytes. The signalling complex composed of two GCSFs (GCSF, 19 kDa) and two GCSF receptors (GCSFR, 34 kDa) consisting of an Ig-like domain and a cytokine-receptor homologous (CRH) domain was crystallized. A crystal of the complex was grown in 1.0 M sodium formate and 0.1 M sodium acetate pH 4.6 and belongs to space group P4(1)2(1)2 (or its enantiomorph P4(3)2(1)2), with unit-cell parameters a = b = 110.1, c = 331.8 A. Unfortunately, this crystal form did not diffract beyond 5 A resolution. Since the heterogeneity of GCSF receptor appeared to prevent the growth of good-quality crystals, the GCSF receptor was fractionated by anion-exchange chromatography. Crystals of the GCSF-fractionated GCSF receptor complex were grown as a new crystal form in 0.2 M ammonium phosphate. This new crystal form diffracted to beyond 3.0 A resolution and belonged to space group P3(1)21 (or its enantiomorph P3(2)21), with unit-cell parameters a = b = 134.8, c = 105.7 A.

Thermodynamic analysis of the activation mechanism of the GCSF receptor induced by ligand binding.

The granulocyte colony-stimulating factor receptor (GCSFR), containing the Ig-like domain (Ig) and cytokine receptor homologous region (CRH), was prepared as a preformed dimer (Ig-CRH-Fc)(2) after fusion to the mouse Fc region via an eight-residue linker (approximately 55 A). Monomer Ig-CRH was also prepared after the Fc region was removed from (Ig-CRH-Fc)(2). GCSF binding to Ig-CRH and (Ig-CRH-Fc)(2) was investigated using light scattering and isothermal titration calorimetry. The average molecular mass determined by light scattering showed that both Ig-CRH and (Ig-CRH-Fc)(2) formed a 2:2 dimer with GCSF. Moreover, isothermal titration calorimetry showed that the thermodynamic parameters upon binding of GCSF to Ig-CRH and (Ig-CRH-Fc)(2) were comparable, suggesting a similar binding stoichiometry and interface [including similar buried surface area (5700-6000 A(2))] despite the presence of the eight-residue linker. The buried surface area is much larger than that calculated from our previous report of the crystal structure of the GCSF-CRH complex [Aritomi, M., et al. (1999) Nature 401, 713-717], suggesting a substantial contribution of the Ig domain to GCSF binding. The data also indicate that the distance (55 A) between two CRH domains in the 2:2 complex is much shorter than in our previous model (approximately 90 A) predicted from the same crystal structure of the GCSF-CRH complex.

Genomic clones encoding two isoforms of pokeweed antiviral protein in seeds (PAP-S1 and S2) and the N-glycosidase activities of their recombinant proteins on ribosomes and DNA in comparison with other isoforms.

Pokeweed antiviral proteins (PAPs) are single-chain ribosome-inactivating proteins (RIPs) isolated from several organs of Phytolacca americana (Pokeweed) that are characterized by their ability to depurinate not only ribosomes but also various nucleic acids. PAP-S is one of the isoforms found in seeds. In this study, we obtained three different genomic clones encoding two forms of PAP-S (here designated as PAP-S1 and PAP-S2) and alpha-PAP after PCR using a pair of degenerated primers based on the known N- and C-terminal amino acid sequences of PAP-S. The nucleotide sequences of the genomic clones contained no introns. The deduced amino acid sequences of PAP-S1 and PAP-S2, which showed 83% identity to each other, were found to correspond to sequences reported independently for PAP-S protein and cDNA, respectively, demonstrating that at least two forms of PAP-S actually exist in seeds of the same plant. The recombinant PAP-S1, PAP-S2, alpha-PAP, and PAP I (a form appearing in spring leaves) exhibit the same level of depurinating activity on rat ribosomes, while their efficiencies on Escherichia coli ribosomes and salmon sperm DNA differ substantially from one another in the order of PAP I > alpha-PAP > PAP-S1 > PAP-S2 and alpha-PAP > PAP I > PAP-S1 > PAP-S2. Structural comparisons suggest that the large difference in ribosome recognition between PAP-S1 (or S2) and PAP I is caused by the alteration of residues adjacent to the adenine-binding site.

Real-time kinetic analyses of the interaction of ricin toxin A-chain with ribosomes prove a conformational change involved in complex formation.

Ricin toxin A-chain (RTA), a ribosome-inactivating protein from seeds of the castor bean plant (Ricinus communis), inactivates eukaryotic ribosomes by hydrolyzing the N-glycosidic bond of a single adenosine residue in a highly conserved loop of 28S rRNA, but does not act on prokaryotic ribosomes. We investigated the interaction of rat liver 80S ribosomes with RTA using an optical biosensor based on surface plasmon resonance (BIAcore instrument), which allows real-time recording of the interaction. RTA was coupled to the dextran gel matrix on the sensor chip surface through a single thiol group that is not involved in the enzymatic action. The interaction of rat ribosomes with RTA, which was greatly affected by the Mg(2+) concentration and ionic strength, was usually measured at 5 mM Mg(2+), 50 mM KCl, and pH 7.5. The modes of interaction of intact and RTA-depurinated rat liver ribosomes with the immobilized RTA were virtually the same, while no considerable interaction was observed for Escherichia coli ribosomes. The interaction was not influenced by the presence of 5 mM adenine, which is higher than the reported dissociation constant (1 mM) for the adenine-RTA complex. These results demonstrate that binding of the target adenine with the active site of RTA does not contribute much to the total interaction of ribosomes and RTA. Global analyses of association and dissociation data with several binding models, taking account of mass transport, allowed us to conclude that the data were unable to fit a simple 1:1 binding model, but were best described by a model including a conformational change involved in high affinity complex formation.