Substrate distortion in the catalysis of orotidine monophosphate decarboxylase

One way for enzymes to affect reactions they catalyze is through transition state stabilization. Another factor to be considered is the contribution of substrate distortion, although it has been thoroughly described for only a few enzymes. We have a longstanding interest in the reaction mechanism of orotidine monophosphate decarboxylase (ODCase) and determined various crystal structures bound with distorted substrates at around 1.5 Å resolution. The enzyme is known as one of the most proficient enzymes, which accelerates the decarboxylation of orotidine 5'-monophosphate (OMP) to form uridine 5'-monophosphate (UMP) by 17 orders of magnitude. One argument against the contribution of substrate distortion to the ODCase reaction is the weak affinity of UMP. The distortions observed so far all appear at the C6-substituent of the pyrimidine ring, which corresponds to the carboxylate of OMP. Since the carboxylate is removed by the reaction, the product UMP should bind more tightly to ODCase than OMP, if the distortion of C6-substituent contributes to the catalysis. In order to investigate this inconsistency, we determined the crystal structure of ODCase with UMP at atomic resolution (1.03 Å). The structure showed an unfavorable interaction between UMP and the catalytic residue K72, an interaction considered to be absent in the OMP complex. Surface plasmon resonance analysis indicated that UMP binds stronger to the K72A mutant than to the wild-type enzyme by 5 orders of magnitude. These analyses invalidate the argument against a contribution of substrate distortion to ODCase catalysis. Finally, we estimated how much the distortion contributes to the catalysis using computational simulation methods. The results indicated that 10-15% decrease of the ΔΔG‡ value is contributed by substrate distortion.

Crystal structure of archaeal phosphopantothenate synthetase

Bacteria/eukaryotes share a common pathway for coenzyme A biosynthesis which involves two enzymes, pantothenate synthetase and pantothenate kinase, to convert pantoate to 4'-phosphopantothenate. These two enzymes are absent in almost all archaea. Recently, it was reported that two novel enzymes, pantoate kinase (PoK) and phosphopantothenate synthetase (PPS), are responsible for this conversion in archaea[1]. In archaea, pantoate is phosphorylated by PoK to produce 4-phosphopantoate (PPo), and then condensation of PPo and β-alanine is catalyzed by PPS, generating 4'-phosphopantothenate. Here, we report the crystal structure of PPS from the hyperthermophilic archaeon, Thermococcus kodakarensis and its complexes with ATP, and ATP and 4-phosphopantoate (PPo). PPS forms an asymmetric homodimer, in which two monomers composing a dimer, deviated from the exact 2-fold symmetry, displaying 4⁰-13⁰ distortion. Two active sites in PPS dimer are located near the rotation axis. Due to the asymmetricity of PPS dimer molecule, two active sites in PPS dimer are not equivalent. The structural features are consistent with the mutagenesis data and the results of biochemical experiments previously reported. Based on the structures of PPS, PPS/ATP complex, and PPS/ATP/PPo complex, we discuss the catalytic mechanism by which PPS produces phosphopantoyl adenylate (PPA), which is thought to be a reaction intermediate.

Structural studies of myo-inositol kinase

The TK2285 protein from a hyperthermopilic archaeon Thermococcus kodakarensis is a myo-inositol kinase. Only two myo-inositol kinases have been identified so far. One is the TK2285 protein and the other is an enzyme from Zea mays. Both of them synthesize myo-inositol monophosphate that shows enantiomerism. Because it is too difficult to discriminate enantiomers by NMR or chromatography analysis, it has not been identified which of the six hydroxyls is phosphorylated by these enzymes. Also, little is known about the substrate recognition of myo-inositol kinase, since only the unliganded crystal structure of TK2285 has been reported. In order to reveal the substrate-binding mechanism of myo-inositol kinase and identify the phosphorylated hydroxyl group of the product, we determined the crystal structures of TK2285 as the substrate-complex and the product-complex. The substrate-complex of TK2285 was prepared by using the TK2285, myo-inositol and AMP-PCP, and the products-complex was prepared by incubating the TK2285 with myo-inositol and ATP. The substrate-complex structure showed that all of the six hydroxyls of myo-inositol interacted with TK2285. This coincides with the fact that the Km value for myo-inositol is 100-1000 fold lower than those for other sugars. Also 3-hydroxyl group of myo-inositol, which the gamma-phosphate of AMP-PCP was nearest to, was thought to be phosphorylated by this enzyme. This was proved by the product-complex structure that had ADP and myo-inositol 3-phosphate. Site-directed mutagenesis and structure comparison with TK2285 homologs also provided information about the substrate-binding mechanism of myo-inositol kinase.

Partial oxygen-dissociation of crystalline giant hemoglobin

Allosteric oxygen-binding of hemoglobin (Hb) has been widely discussed based on the quaternary structural changes elucidated by the crystal structures of the oxygenated and deoxygenated states. However, it remains to be determined the structure of intermediate states between the oxy and deoxy forms without any artificial modification of the Hb molecule. A tubeworm, Lamellibrachia satsuma has extracellular giant hemoglobins with a molecular mass of about 400 and 3,600 kDa. Recently, we have determined the crystal structure of the 400 kDa Hb (V2Hb) in the oxy state, and then we successfully obtained the deoxygenated crystals of V2Hb from oxy crystals by the soaking methods [1]. These findings encourage us to initiate structural studies for the intermediate states between the oxy and deoxy forms of V2Hb, which should provide a more accurate understanding of the allosteric mechanism of Hbs. The deoxy crystals of V2Hb were obtained from oxy crystals through the soaking in a solution containing 50 mM sodium hydrosulfite, and incubated for a few minutes. We tested various soaking times from 3 s to 180 s and then immediately flash-frozen under a nitrogen gas stream. The obtained structures reveal that in the case of the soaking time was longer than 10 s, the electron densities of the oxygen molecules at some heme pockets (oxygen binding sites) were very week or disappeared. These `intermediate' structures show almost the same quaternary structure as that of the oxy structure. This fact suggests that quaternary rearrangement of V2Hb might arise just before a complete dissociation of all the oxygen molecules from all the subunits.

Structural basis for the biosynthesis of the CN ligand of [NiFe] hydrogenase

[NiFe] hydrogenases carry a NiFe(CN)₂CO center at the active site, catalyzing the reversible H₂ oxidation. The complex NiFe center is biosynthesized and inserted into the enzyme by six specific maturation proteins: Hyp proteins (HypABCDEF). HypE and HypF are involved in biosynthesis of cyanide ligands, which are attached to the Fe atom in the NiFe center. First, HypF catalyzes a transfer reaction of the carbamoyl moiety of carbamoylphosphate to the C-terminal cysteine residue of HypE. Then, HypE catalyzes an ATP-dependent dehydration of the carbamoylated C-terminal cysteine of HypE to thiocyanate. Although structures of HypE proteins have been determined, there has been no structural evidence to explain how HypE dehydrates thiocarboxamide into thiocyanate. In order to elucidate the catalytic mechanism of HypE, we have determined the crystal structures of the carbamoylated and cyanated states of HypE from Thermococcus kodakarensis in complex with nucleotides at 1.53 Å and 1.64 Å resolution, respectively [1]. Carbamoylation of the C-terminal cysteine (Cys338) of HypE by chemical modification is clearly observed in the present structures. A conserved glutamate residue (Glu272) is close to the thiocarboxamide nitrogen atom of Cys338. However, the configuration of Glu272 is less favorable for proton abstraction. On the other hand, the thiocarboxamide oxygen atom of Cys338 interacts with a conserved lysine residue (Lys134) through a water molecule. Interestingly, a conserved arginine residue makes close contact with Lys134 and lowers the pKa of Lys134, suggesting that Lys134 functions as a proton acceptor. These observations suggest that the dehydration of thiocarboxamide into thiocyanate is catalyzed by a two-step deprotonation process, in which Lys134 and Glu272 function as the first and second bases, respectively.

Crystallographic studies for the folding of an extending peptide

Full-length proteins can fold into thermodynamically stable structures at an exceptionally fast rate as shown by in vitro experiments. In contrast, it takes much more time to finish nascent protein folding than full-length protein folding, because nascent protein folding depends on the rate of ribosome biosynthesis in the living cell. Therefore nascent polypeptide chains in vivo fold co-translationally in different manners from the full-length proteins. However, the transient structures and the co-translational folding pathway are not well understood. In order to reveal the atomic details of nascent protein folding, we studied the hPin1 WW domain, which consists of two beta-hairpins between the three-stranded beta-sheets. Here we report a series of WW domain N-terminal fragment structures with increasing amino acid length by using circular dichroism spectroscopy and X-ray crystallography. In crystallization, maltose-binding protein was fused just behind the WW domain fragments to fix the C-terminus as nascent proteins are anchored to the ribosome. Co-translational folding of beta-sheet-rich proteins is discussed based on our finding that intermediate-length fragments unexpectedly take a helical conformation, even though the full-length protein has no helical regions. Furthermore, in a region of one of the loop structures of the full-length protein, these fragments take different formations. Our results suggest that the newly synthesized polypeptides adopt the most stable conformation during the course of peptide extension and fold into the native structures, eventually.

Ultra-high resolution structure of high-potential iron-sulfur protein

It is important for understanding the electron transfer reaction to include the information about valence shell electrons and hydrogen atoms into crystal structure refinement. High-potential iron-sulfur protein (HiPIP) possesses a Fe4S4 cluster which exhibits +2/+3 redox states and acts as an electron carrier from cytochrome bc1 complex to the reaction center complex in photosynthetic purple bacteria. We have reported the X-ray crystal structure of HiPIP from Thermochtomatium tepidum at 0.72 Å resolution (1). Recently, we have successfully collected 0.48 Å resolution data of HiPIP using high-energy X-rays (31 keV) in BL41XU beamline of SPring-8. We performed multipolar refinement with the MoPro program (2) to consider valence shell electrons in the structure refinement of HiPIP. Refinement of multipolar parameters was applied to atoms of single conformational residues, water molecules with two hydrogen atoms, and the Fe4S4 cluster. After multipolar refinement, the deformation map clearly displays the distribution of valence shell electrons such as lone-pair electrons of carbonyl oxygen atoms, bonding electrons in aromatic rings, and d-orbital electrons of Fe atoms in the Fe4S4 cluster. The deformation map also indicates electrostatic interactions between the S atoms of Fe4S4-(Cys-Sγ)4 and protein environment. In addition, we performed preliminary neutron diffraction experiment at iBIX beamline of Japan Proton Accelerator Research Complex (J-PARC) and observed diffraction spots up to 1.17 Å resolution using HiPIP crystal with the size of 2.3 mm3. In the multipolar refinement, the positions of hydrogen atoms were fixed to the standard bond distances derived from neutron crystal structures of small molecules and atomic displacement parameters of hydrogen atoms were constrained to 1.2 or 1.5 fold of their root atoms. Therefore, a high resolution neutron structure of HiPIP will improve the results obtained from the multipolar refinement.