Thiolase: A Versatile Biocatalyst Employing Coenzyme A-Thioester Chemistry for Making and Breaking C-C Bonds.

Thiolases are CoA-dependent enzymes that catalyze the thiolytic cleavage of 3-ketoacyl-CoA, as well as its reverse reaction, which is the thioester-dependent Claisen condensation reaction. Thiolases are dimers or tetramers (dimers of dimers). All thiolases have two reactive cysteines: (a) a nucleophilic cysteine, which forms a covalent intermediate, and (b) an acid/base cysteine. The best characterized thiolase is the Zoogloea ramigera thiolase, which is a bacterial biosynthetic thiolase belonging to the CT-thiolase subfamily. The thiolase active site is also characterized by two oxyanion holes, two active site waters, and four catalytic loops with characteristic amino acid sequence fingerprints. Three thiolase subfamilies can be identified, each characterized by a unique sequence fingerprint for one of their catalytic loops, which causes unique active site properties. Recent insights concerning the thiolase reaction mechanism, as obtained from recent structural studies, as well as from classical and recent enzymological studies, are addressed, and open questions are discussed.

De novo biosynthesis of sterols and fatty acids in the Trypanosoma brucei procyclic form: Carbon source preferences and metabolic flux redistributions.

De novo biosynthesis of lipids is essential for Trypanosoma brucei, a protist responsible for the sleeping sickness. Here, we demonstrate that the ketogenic carbon sources, threonine, acetate and glucose, are precursors for both fatty acid and sterol synthesis, while leucine only contributes to sterol production in the tsetse fly midgut stage of the parasite. Degradation of these carbon sources into lipids was investigated using a combination of reverse genetics and analysis of radio-labelled precursors incorporation into lipids. For instance, (i) deletion of the gene encoding isovaleryl-CoA dehydrogenase, involved in the leucine degradation pathway, abolished leucine incorporation into sterols, and (ii) RNAi-mediated down-regulation of the SCP2-thiolase gene expression abolished incorporation of the three ketogenic carbon sources into sterols. The SCP2-thiolase is part of a unidirectional two-step bridge between the fatty acid precursor, acetyl-CoA, and the precursor of the mevalonate pathway leading to sterol biosynthesis, 3-hydroxy-3-methylglutaryl-CoA. Metabolic flux through this bridge is increased either in the isovaleryl-CoA dehydrogenase null mutant or when the degradation of the ketogenic carbon sources is affected. We also observed a preference for fatty acids synthesis from ketogenic carbon sources, since blocking acetyl-CoA production from both glucose and threonine abolished acetate incorporation into sterols, while incorporation of acetate into fatty acids was increased. Interestingly, the growth of the isovaleryl-CoA dehydrogenase null mutant, but not that of the parental cells, is interrupted in the absence of ketogenic carbon sources, including lipids, which demonstrates the essential role of the mevalonate pathway. We concluded that procyclic trypanosomes have a strong preference for fatty acid versus sterol biosynthesis from ketogenic carbon sources, and as a consequence, that leucine is likely to be the main source, if not the only one, used by trypanosomes in the infected insect vector digestive tract to feed the mevalonate pathway.

The peroxisomal zebrafish SCP2-thiolase (type-1) is a weak transient dimer as revealed by crystal structures and native mass spectrometry.

The SCP2 (sterol carrier protein 2)-thiolase (type-1) functions in the vertebrate peroxisomal, bile acid synthesis pathway, converting 24-keto-THC-CoA and CoA into choloyl-CoA and propionyl-CoA. This conversion concerns the ß-oxidation chain shortening of the steroid fatty acyl-moiety of 24-keto-THC-CoA. This class of dimeric thiolases has previously been poorly characterized. High-resolution crystal structures of the zebrafish SCP2-thiolase (type-1) now reveal an open catalytic site, shaped by residues of both subunits. The structure of its non-dimerized monomeric form has also been captured in the obtained crystals. Four loops at the dimer interface adopt very different conformations in the monomeric form. These loops also shape the active site and their structural changes explain why a competent active site is not present in the monomeric form. Native mass spectrometry studies confirm that the zebrafish SCP2-thiolase (type-1) as well as its human homolog are weak transient dimers in solution. The crystallographic binding studies reveal the mode of binding of CoA and octanoyl-CoA in the active site, highlighting the conserved geometry of the nucleophilic cysteine, the catalytic acid/base cysteine and the two oxyanion holes. The dimer interface of SCP2-thiolase (type-1) is equally extensive as in other thiolase dimers; however, it is more polar than any of the corresponding interfaces, which correlates with the notion that the enzyme forms a weak transient dimer. The structure comparison of the monomeric and dimeric forms suggests functional relevance of this property. These comparisons provide also insights into the structural rearrangements that occur when the folded inactive monomers assemble into the mature dimer.

Structural characterization of a mitochondrial 3-ketoacyl-CoA (T1)-like thiolase from Mycobacterium smegmatis.

Thiolases catalyze the degradation and synthesis of 3-ketoacyl-CoA molecules. Here, the crystal structures of a T1-like thiolase (MSM-13 thiolase) from Mycobacterium smegmatis in apo and liganded forms are described. Systematic comparisons of six crystallographically independent unliganded MSM-13 thiolase tetramers (dimers of tight dimers) from three different crystal forms revealed that the two tight dimers are connected to a rigid tetramerization domain via flexible hinge regions, generating an asymmetric tetramer. In the liganded structure, CoA is bound to those subunits that are rotated towards the tip of the tetramerization loop of the opposing dimer, suggesting that this loop is important for substrate binding. The hinge regions responsible for this rotation occur near Val123 and Arg149. The Lα1-covering loop-Lα2 region, together with the Nß2-Nα2 loop of the adjacent subunit, defines a specificity pocket that is larger and more polar than those of other tetrameric thiolases, suggesting that MSM-13 thiolase has a distinct substrate specificity. Consistent with this finding, only residual activity was detected with acetoacetyl-CoA as the substrate in the degradative direction. No activity was observed with acetyl-CoA in the synthetic direction. Structural comparisons with other well characterized thiolases suggest that MSM-13 thiolase is probably a degradative thiolase that is specific for 3-ketoacyl-CoA molecules with polar, bulky acyl chains.

The crystal structure of human mitochondrial 3-ketoacyl-CoA thiolase (T1): insight into the reaction mechanism of its thiolase and thioesterase activities.

Crystal structures of human mitochondrial 3-ketoacyl-CoA thiolase (hT1) in the apo form and in complex with CoA have been determined at 2.0 Å resolution. The structures confirm the tetrameric quaternary structure of this degradative thiolase. The active site is surprisingly similar to the active site of the Zoogloea ramigera biosynthetic tetrameric thiolase (PDB entries 1dm3 and 1m1o) and different from the active site of the peroxisomal dimeric degradative thiolase (PDB entries 1afw and 2iik). A cavity analysis suggests a mode of binding for the fatty-acyl tail in a tunnel lined by the Nß2-Nα2 loop of the adjacent subunit and the Lα1 helix of the loop domain. Soaking of the apo hT1 crystals with octanoyl-CoA resulted in a crystal structure in complex with CoA owing to the intrinsic acyl-CoA thioesterase activity of hT1. Solution studies confirm that hT1 has low acyl-CoA thioesterase activity for fatty acyl-CoA substrates. The fastest rate is observed for the hydrolysis of butyryl-CoA. It is also shown that T1 has significant biosynthetic thiolase activity, which is predicted to be of physiological importance.

Structural enzymology studies with the substrate 3S-hydroxybutanoyl-CoA: bifunctional MFE1 is a less efficient dehydrogenase than monofunctional HAD.

Multifunctional enzyme, type-1 (MFE1) catalyzes the second and third step of the ß-oxidation cycle, being, respectively, the 2E-enoyl-CoA hydratase (ECH) reaction (N-terminal part, crotonase fold) and the NAD+-dependent, 3S-hydroxyacyl-CoA dehydrogenase (HAD) reaction (C-terminal part, HAD fold). Structural enzymological properties of rat MFE1 (RnMFE1) as well as of two of its variants, namely the E123A variant (a glutamate of the ECH active site is mutated into alanine) and the BCDE variant (without domain A of the ECH part), were studied, using as substrate 3S-hydroxybutanoyl-CoA. Protein crystallographic binding studies show the hydrogen bond interactions of 3S-hydroxybutanoyl-CoA as well as of its 3-keto, oxidized form, acetoacetyl-CoA, with the catalytic glutamates in the ECH active site. Pre-steady state binding experiments with NAD+ and NADH show that the kon and koff rate constants of the HAD active site of monomeric RnMFE1 and the homologous human, dimeric 3S-hydroxyacyl-CoA dehydrogenase (HsHAD) for NAD+ and NADH are very similar, being the same as those observed for the E123A and BCDE variants. However, steady state and pre-steady state kinetic data concerning the HAD-catalyzed dehydrogenation reaction of the substrate 3S-hydroxybutanoyl-CoA show that, respectively, the kcat and kchem rate constants for conversion into acetoacetyl-CoA by RnMFE1 (and its two variants) are about 10 fold lower as when catalyzed by HsHAD. The dynamical properties of dehydrogenases are known to be important for their catalytic efficiency, and it is discussed that the greater complexity of the RnMFE1 fold correlates with the observation that RnMFE1 is a slower dehydrogenase than HsHAD.

Structures of lactaldehyde reductase, FucO, link enzyme activity to hydrogen bond networks and conformational dynamics.

A group-III iron containing 1,2-propanediol oxidoreductase, FucO, (also known as lactaldehyde reductase) from Escherichia coli was examined regarding its structure-dynamics-function relationships in the catalysis of the NADH-dependent reduction of (2S)-lactaldehyde. Crystal structures of FucO variants in the presence or absence of cofactors have been determined, illustrating large domain movements between the apo and holo enzyme structures. Different structures of FucO variants co-crystallized with NAD+ or NADH together with substrate further suggest dynamic properties of the nicotinamide moiety of the coenzyme that are important for the reaction mechanism. Modelling of the native substrate (2S)-lactaldehyde into the active site can explain the stereoselectivity exhibited by the enzyme, with a critical hydrogen bond interaction between the (2S)-hydroxyl and the side-chain of N151, as well as the previously experimentally demonstrated pro-(R) selectivity in hydride transfer from NADH to the aldehydic carbon. Furthermore, the deuterium kinetic isotope effect of hydride transfer suggests that reduction chemistry is the main rate-limiting step for turnover which is not the case in FucO catalysed alcohol oxidation. We further propose that a water molecule in the active site - hydrogen bonded to a conserved histidine (H267) and the 2'-hydroxyl of the coenzyme ribose - functions as a catalytic proton donor in the protonation of the product alcohol. A hydrogen bond network of water molecules and the side-chains of amino acid residues D360 and H267 links bulk solvent to this proposed catalytic water molecule.

Crystal structures and kinetic studies of a laboratory evolved aldehyde reductase explain the dramatic shift of its new substrate specificity.

The Fe2+-dependent E. coli enzyme FucO catalyzes the reversible interconversion of short-chain (S)-lactaldehyde and (S)-1,2-propanediol, using NADH and NAD+ as cofactors, respectively. Laboratory-directed evolution experiments have been carried out previously using phenylacetaldehyde as the substrate for screening catalytic activity with bulky substrates, which are very poorly reduced by wild-type FucO. These experiments identified the N151G/L259V double mutant (dubbed DA1472) as the most active variant with this substrate via a two-step evolutionary pathway, in which each step consisted of one point mutation. Here the crystal structures of DA1472 and its parent D93 (L259V) are reported, showing that these amino acid substitutions provide more space in the active site, though they do not cause changes in the main-chain conformation. The catalytic activity of DA1472 with the physiological substrate (S)-lactaldehyde and a series of substituted phenylacetaldehyde derivatives were systematically quantified and compared with that of wild-type as well as with the corresponding point-mutation variants (N151G and L259V). There is a 9000-fold increase in activity, when expressed as kcat/KM values, for DA1472 compared with wild-type FucO for the phenylacetaldehyde substrate. The crystal structure of DA1472 complexed with a non-reactive analog of this substrate (3,4-dimethoxyphenylacetamide) suggests the mode of binding of the bulky group of the new substrate. These combined structure-function studies therefore explain the dramatic increase in catalytic activity of the DA1472 variant for bulky aldehyde substrates. The structure comparisons also suggest why the active site in which Fe2+ is replaced by Zn2+ is not able to support catalysis.

IceBear: an intuitive and versatile web application for research-data tracking from crystallization experiment to PDB deposition.

The web-based IceBear software is a versatile tool to monitor the results of crystallization experiments and is designed to facilitate supervisor and student communications. It also records and tracks all relevant information from crystallization setup to PDB deposition in protein crystallography projects. Fully automated data collection is now possible at several synchrotrons, which means that the number of samples tested at the synchrotron is currently increasing rapidly. Therefore, the protein crystallography research communities at the University of Oulu, Weizmann Institute of Science and Diamond Light Source have joined forces to automate the uploading of sample metadata to the synchrotron. In IceBear, each crystal selected for data collection is given a unique sample name and a crystal page is generated. Subsequently, the metadata required for data collection are uploaded directly to the ISPyB synchrotron database by a shipment module, and for each sample a link to the relevant ISPyB page is stored. IceBear allows notes to be made for each sample during cryocooling treatment and during data collection, as well as in later steps of the structure determination. Protocols are also available to aid the recycling of pins, pucks and dewars when the dewar returns from the synchrotron. The IceBear database is organized around projects, and project members can easily access the crystallization and diffraction metadata for each sample, as well as any additional information that has been provided via the notes. The crystal page for each sample connects the crystallization, diffraction and structural information by providing links to the IceBear drop-viewer page and to the ISPyB data-collection page, as well as to the structure deposited in the Protein Data Bank.

Crystallographic binding studies of rat peroxisomal multifunctional enzyme type 1 with 3-ketodecanoyl-CoA: capturing active and inactive states of its hydratase and dehydrogenase catalytic sites.

The peroxisomal multifunctional enzyme type 1 (MFE1) catalyzes two successive reactions in the ß-oxidation cycle: the 2E-enoyl-CoA hydratase (ECH) and NAD+-dependent 3S-hydroxyacyl-CoA dehydrogenase (HAD) reactions. MFE1 is a monomeric enzyme that has five domains. The N-terminal part (domains A and B) adopts the crotonase fold and the C-terminal part (domains C, D and E) adopts the HAD fold. A new crystal form of MFE1 has captured a conformation in which both active sites are noncompetent. This structure, at 1.7 Šresolution, shows the importance of the interactions between Phe272 in domain B (the linker helix; helix H10 of the crotonase fold) and the beginning of loop 2 (of the crotonase fold) in stabilizing the competent ECH active-site geometry. In addition, protein crystallographic binding studies using optimized crystal-treatment protocols have captured a structure with both the 3-ketodecanoyl-CoA product and NAD+ bound in the HAD active site, showing the interactions between 3-ketodecanoyl-CoA and residues of the C, D and E domains. Structural comparisons show the importance of domain movements, in particular of the C domain with respect to the D/E domains and of the A domain with respect to the HAD part. These comparisons suggest that the N-terminal part of the linker helix, which interacts tightly with domains A and E, functions as a hinge region for movement of the A domain with respect to the HAD part.

Crystal structure of human filamin C domain 23 and small angle scattering model for filamin C 23-24 dimer.

Filamin C is a dimeric, actin-binding protein involved in organization of cortical cytoskeleton and of the sarcomere. We performed crystallographic, small-angle X-ray scattering and analytical ultracentrifugation experiments on the constructs containing carboxy-terminal domains of the protein (domains 23-24 and 19-21). The crystal structure of domain 23 of filamin C showed that the protein adopts the expected immunoglobulin (Ig)-like fold. Small-angle X-ray scattering experiments performed on filamin C tandem Ig-like domains 23 and 24 reveal a dimer that is formed by domain 24 and that domain 23 has little interactions with itself or with domain 24, while the analytical ultracentrifugation experiments showed that the filamin C domains 19-21 form elongated monomers in diluted solutions.

Structural basis for vertebrate filamin dimerization.

Filamins are essential in cell motility and many developmental processes. They are large actin cross linking proteins that contain actin binding domains in their N termini and a long rod region constructed from 24 tandem Ig domains. Dimerization is crucial for the actin crosslinking function of filamins and requires the most C-terminal Ig domain. We describe here the crystal structure of this 24th Ig domain (Ig24) of human filamin C and show how it mediates dimerization. The dimer interface is novel and quite different to that seen in the Dictyostelium discoideum filamin analog. The sequence signature of the dimerization interface suggests that the C-terminal domains of all vertebrate filamins share the same dimerization mechanism. Furthermore, we show that point mutations in the dimerization interface disrupt the dimer and that the dissociation constant for recombinant Ig24 is in the micromolar range.

Structural enzymology comparisons of multifunctional enzyme, type-1 (MFE1): the flexibility of its dehydrogenase part.

Multifunctional enzyme, type-1 (MFE1) is a monomeric enzyme with a 2E-enoyl-CoA hydratase and a 3S-hydroxyacyl-CoA dehydrogenase (HAD) active site. Enzyme kinetic data of rat peroxisomal MFE1 show that the catalytic efficiencies for converting the short-chain substrate 2E-butenoyl-CoA into acetoacetyl-CoA are much lower when compared with those of the homologous monofunctional enzymes. The mode of binding of acetoacetyl-CoA (to the hydratase active site) and the very similar mode of binding of NAD + and NADH (to the HAD part) are described and compared with those of their monofunctional counterparts. Structural comparisons suggest that the conformational flexibility of the HAD and hydratase parts of MFE1 are correlated. The possible importance of the conformational flexibility of MFE1 for its biocatalytic properties is discussed. Database: Structural data are available in PDB database under the accession number 5MGB.

Crystallographic substrate binding studies of Leishmania mexicana SCP2-thiolase (type-2): unique features of oxyanion hole-1.

C: Structures of the C123A variant of the dimeric Leishmania mexicana SCP2-thiolase (type-2) (Lm-thiolase), complexed with acetyl-CoA and acetoacetyl-CoA, respectively, are reported. The catalytic site of thiolase contains two oxyanion holes, OAH1 and OAH2, which are important for catalysis. The two structures reveal for the first time the hydrogen bond interactions of the CoA-thioester oxygen atom of the substrate with the hydrogen bond donors of OAH1 of a CHH-thiolase. The amino acid sequence fingerprints ( xS, EAF, G P) of three catalytic loops identify the active site geometry of the well-studied CNH-thiolases, whereas SCP2-thiolases (type-1, type-2) are classified as CHH-thiolases, having as corresponding fingerprints xS, DCF and G P. In all thiolases, OAH2 is formed by the main chain NH groups of two catalytic loops. In the well-studied CNH-thiolases, OAH1 is formed by a water (of the Wat-Asn(NEAF) dyad) and NE2 (of the GHP-histidine). In the two described liganded Lm-thiolase structures, it is seen that in this CHH-thiolase, OAH1 is formed by NE2 of His338 (HDCF) and His388 (GHP). Analysis of the OAH1 hydrogen bond networks suggests that the GHP-histidine is doubly protonated and positively charged in these complexes, whereas the HDCF histidine is neutral and singly protonated.

Crystal structures of two monomeric triosephosphate isomerase variants identified via a directed-evolution protocol selecting for L-arabinose isomerase activity.

The crystal structures are described of two variants of A-TIM: Ma18 (2.7 Šresolution) and Ma21 (1.55 Šresolution). A-TIM is a monomeric loop-deletion variant of triosephosphate isomerase (TIM) which has lost the TIM catalytic properties. Ma18 and Ma21 were identified after extensive directed-evolution selection experiments using an Escherichia coli L-arabinose isomerase knockout strain expressing a randomly mutated A-TIM gene. These variants facilitate better growth of the Escherichia coli selection strain in medium supplemented with 40 mM L-arabinose. Ma18 and Ma21 differ from A-TIM by four and one point mutations, respectively. Ma18 and Ma21 are more stable proteins than A-TIM, as judged from CD melting experiments. Like A-TIM, both proteins are monomeric in solution. In the Ma18 crystal structure loop 6 is open and in the Ma21 crystal structure loop 6 is closed, being stabilized by a bound glycolate molecule. The crystal structures show only small differences in the active site compared with A-TIM. In the case of Ma21 it is observed that the point mutation (Q65L) contributes to small structural rearrangements near Asn11 of loop 1, which correlate with different ligand-binding properties such as a loss of citrate binding in the active site. The Ma21 structure also shows that its Leu65 side chain is involved in van der Waals interactions with neighbouring hydrophobic side-chain moieties, correlating with its increased stability. The experimental data suggest that the increased stability and solubility properties of Ma21 and Ma18 compared with A-TIM cause better growth of the selection strain when coexpressing Ma21 and Ma18 instead of A-TIM.

Organization of the multifunctional enzyme type 1: interaction between N- and C-terminal domains is required for the hydratase-1/isomerase activity.

Rat peroxisomal multifunctional enzyme type 1 (perMFE-1) is a monomeric protein of beta-oxidation. We have defined five functional domains (A, B, C, D and E) in the perMFE-1 based on comparison of the amino acid sequence with homologous proteins from databases and structural data of the hydratase-1/isomerases (H1/I) and (3 S )-hydroxyacyl-CoA dehydrogenases (HAD). Domain A (residues 1-190) comprises the H1/I fold and catalyses both 2-enoyl-CoA hydratase-1 and Delta(3)-Delta(2)-enoyl-CoA isomerase reactions. Domain B (residues 191-280) links domain A to the (3 S )-dehydrogenase region, which includes both domain C (residues 281-474) and domain D (residues 480-583). Domains C and D carry features of the dinucleotide-binding and the dimerization domains of monofunctional HADs respectively. Domain E (residues 584-722) has sequence similarity to domain D of the perMFE-1, which suggests that it has evolved via partial gene duplication. Experiments with engineered perMFE-1 variants demonstrate that the H1/I competence of domain A requires stabilizing interactions with domains D and E. The variant His-perMFE (residues 288-479)Delta, in which the domain C is deleted, is stable and has hydratase-1 activity. It is proposed that the extreme C-terminal domain E in perMFE-1 serves the following three functions: (i) participation in the folding of the N-terminus into a functionally competent H1/I fold, (ii) stabilization of the dehydrogenation domains by interaction with the domain D and (iii) the targeting of the perMFE-1 to peroxisomes via its C-terminal tripeptide.

Crystallization and characterization of the dehydrogenase domain from rat peroxisomal multifunctional enzyme type 1.

Peroxisomal multifunctional enzyme type 1 from rat (perMFE-1) is a monomeric multidomain protein shown to have 2-enoyl-CoA hydratase/Delta(3)-Delta(2)-enoyl-CoA isomerase and (3S)-hydroxyacyl-CoA dehydrogenase domains followed by a C-terminal extension of 130 amino acids with unknown function apart from being a carrier of the peroxisomal targeting signal type 1. The truncated perMFE-1 without the N-terminal hydratase/isomerase domain (perMFE-1DH; residues 260-722) was overexpressed as an enzymatically active recombinant protein, purified and characterized. Using (3S)-hydroxydecanoyl-CoA as a substrate, the specific enzymatic activity of perMFE-1DH was determined to be 2.2 micromol min(-1) mg(-1), comparable with that of perMFE-1 purified from rat liver (2.8 micromol min(-1) mg(-1)). The protein was crystallized in the apo form by the hanging-drop method and a complete data set to 2.45 A resolution was collected using a rotating-anode X-ray source. The crystals have primitive tetragonal symmetry, with unit-cell parameters a = b = 125.9, c = 60.2 A.