The porphobilinogen synthase catalyzed reaction mechanism

Porphobilinogen synthase (PBGS) catalyzes the first common reaction in the biosynthesis of the tetrapyrroles, the asymmetric condensation of two molecules of delta-aminolevulinic acid to form porphobilinogen. There is a variable requirement for an essential active site zinc that necessitates consideration of PBGS as an enzyme that may exhibit phylogenetic diversity in its chemical reaction mechanism. Recent crystal structures suggest reaction mechanisms that involve two covalent Schiff base linkages between adjacent active site lysine residues and each of the two substrate molecules. The reaction appears to stall at a covalently bound almost-product intermediate that is poised for breakdown to product upon binding of a substrate molecule to an adjacent active site and a subsequent conformational change.     

Mechanistic implications of mutations to the active site lysine of porphobilinogen synthase

Porphobilinogen synthase (PBGS) is a homo-octameric protein that catalyzes the complex asymmetric condensation of two molecules of 5-aminolevulinic acid (ALA). The only characterized intermediate in the PBGS-catalyzed reaction is a Schiff base that forms between the first ALA that binds and a conserved lysine, which in Escherichia coli PBGS is Lys-246 and in human PBGS is Lys-252. In this study, E. coli PBGS mutants K246H, K246M, K246W, K246N, and K246G and human PBGS mutant K252G were characterized. Alterations to this lysine result in a disabled but not totally inactive protein suggesting an alternate mechanism in which proximity and orientation are major catalytic devices. (13)C NMR studies of [3,5-(13)C]porphobilinogen bound at the active sites of the E. coli PBGS and the mutants show only minor chemical shift differences, i.e. environmental alterations. Mammalian PBGS is established to have four functional active sites, whereas the crystal structure of E. coli PBGS shows eight spatially distinct and structurally equivalent subunits. Biochemical data for E. coli PBGS have been interpreted to support both four and eight active sites. A unifying hypothesis is that formation of the Schiff base between this lysine and ALA triggers a conformational change that results in asymmetry. Product binding studies with wild-type E. coli PBGS and K246G demonstrate that both bind porphobilinogen at four per octamer although the latter cannot form the Schiff base from substrate. Thus, formation of the lysine to ALA Schiff base is not required to initiate the asymmetry that results in half-site reactivity.     

UDP-apiose/UDP-xylose synthase. Subunit composition and binding studies.

The UDP-apiose/UDP-xylose synthase from cell suspension cultures of parsley has been purified 1400-fold by an improved method. The ratio of apiose to xylose formed from UDP-D-glucuronic acid (UDP-GlcUA) remained constant throughout the purification procedure. Dodecylsulfate-gel electrophoresis and sedimentation equilibrium measurements showed that this enzyme preparation is composed of two proteins with molecular weights of 65000 and 86000. The two proteins which are present in a molar ratio of about 1:0.7 to 1:0.9 could not be separated by ammonium sulfate fractionation, chromatography on DEAE-cellulose at different pH-values, and on omega-aminoalkyl-Sepharose, and by gel filtration on Acrylex P-100. Each protein is composed of two apparently identical subunits. The presence of only two different subunits was confirmed by end group analysis in which glycine was found as N-terminal amino acid for the larger and lysine for the smaller protein. Crosslinking with dimethylsuberimidate gave dimers of the identical subunits but no hybrids. Separation of the two proteins was achieved on DEAE-cellulose in the presence of urea. After dialysis only the 86000-Mr protein showed enzyme activity with no significant change in the apiose/xylose ratio. However, in the absence of the 65000-Mr protein enzyme stability was decreased drastically. By equilibrium dialysis it was found that 0.5 mol UDP-GlcUA are bound per mole of 86000-Mr protein. NAD+ alone was not bound, but in the presence of UDP it was also bound in a ratio of 0.5 mol/mol catalytic protein. Experiments in which sodium borohydride was added to the enzyme incubation gave no indication that the 4-keto intermediate is bound as a Schiff base to the enzyme. Also no evidence for epimerization at C-3 of the 4-ulose intermediate prior to ring contraction to apiose was found.     

Pig liver monoamine oxidase: Studies on the subunit structure

The molecular weight of pig liver MAO has previously been shown to be about 115,000 with 1 mole of covalently bound FAD per mole of enzyme. Gel filtration of purified enzyme on Sepharose 4B in 6 m guanidine and 0.1 m mercaptoethanol (MCE) and analytical ultracentrifugation in 0.1% sodium dodecyl sulfate (SDS) and 0.1% MCE yielded molecular weights of 55,000 and 63,000, respectively. By polyacrylamide electrophoresis in 0.1% SDS + MCE one band of 60,000 MW appeared. These results seem to imply that the enzyme is composed of two subunits of which one carries the active site. If MCE was omitted during the gel electrophoresis two equally large bands of about 60,000 MW were formed. By using enzyme inhibited by [¹⁴C]pargyline, a MAO-inhibitor blocking the active site of the enzyme in a 1:1 molar ratio, it was found, however, that both bands contained pargyline. Furthermore, amino acid analyses yielded the same amino acid composition of the two bands. The results are interpreted that the enzyme is composed of two subunits of identical molecular size (about 60,000) of which only one contains the active site and that the enzyme preparation contained two forms of the enzyme presumably differing in the number of disulfide bonds.     

Ornithine carbamoyltransferase from Escherichia coli W. Purification, structure and steady-state kinetic analysis.

Ornithine carbamoyltransferase from Escherichia coli W was purified to homogeneity. The enzyme has a molecular weight of 105000. It is composed of three apparently identical subunits with molecular weights of 35000. The mechanism of the ornithine carbamoyltransferase enzyme system from E. coli W was investigated kinetically by using the approach of product inhibition and dead-end inhibition of both forward and reverse reactions. On the basis of the kinetic data and binding studies it appears that the mechanism of the reaction involves a compulsory sequence of substrate binding to the enzyme, in which carbamoylphosphate is the first substrate to bind to the enzyme and phosphate the last product to be released. The same studies also indicate that the mechanism involves dead-end complexes. The reaction mechanism appears consistent with that proposed by Theorell and Chance. Values have been determined for the Michaelis and dissociation constants involved in the combination of each reactant with the enzyme. Comparison of the values for the kinetic constants which are common to both forward and reverse reaction have shown that they are always of a comparable magnitude.     

Mechanism of O-acetylserine sulfhydrylase fromSalmonella typhimurium LT-2

Summary O-Acetylserine sulfhydrylase is a pyridoxal 5-phosphate (PLP) dependent enzyme that catalyzes the final step of L-cysteine biosynthesis inSalmonella, viz. the conversion of O-acetyl-L-serine (OAS) and sulfide to L-cysteine and acetate. A spectrophotometric assay is available using 5-thio(2-nitrobenzoate) (TNB) as an analog of sulfide and monitoring the disappearance of absorbance at 412 nm. The enzyme catalyzes a ping pong mechanism with-aminoacrylate in Schiff base with the active site PLP as a covalent intermediate. Using data obtained from the pH dependence of kinetic parameters, the acid-base chemical mechanism and the optimum protonation state of enzyme and substrate functional groups necessary for binding has been determined. The Schiff base and the-amine of the substrate OAS are unprotonated for binding. There also appears to be a requirement for one active site general base to accept a proton from the-amine and to donate a proton to form cysteine. The enzyme also catalyzes an OAS hydrolase activity, and the pH dependence of this reaction suggests that the active site lysine that participated in the Schiff base linkage is protonated to start the second half reaction, and has a pK of about 8.2. The stereochemistry of³H-borohydride reduction of the Schiff base in free enzyme has been determined by degradation of the resulting pyridoxyllysine to pyridoxamine and measuring³H-release with apo-aspartate aminotransferase. The sequence around the active site lysine is AsnProSerPheSerValLysCysArg.     

Inactivation of Salmonella phosphoribosylpyrophosphate synthetase by specific chemical modification of a lysine residue.

Phosphoribosylpyrophosphate synthetase from Salmonella typhimurium contains nine lysine residues per subunit and can be inactivated by reagents specific for this amino acid. Pyridoxal-P reversibly inhibited the enzyme by about 70% by forming a Schiff base derivative with lysine. Reduction with NaBH₄ made this inactivation irreversible. Kinetic experiments indicated that the failure to inactivate the enzyme completely in a single treatment with pyridoxal-P reflects a reversible equilibrium between inactive Schiff base and a noncovalent complex. Modification of one lysine residue per subunit correlated with apparently total loss of activity. The rate of inactivation of the enzyme was decreased fourfold by saturating concentrations of ATP and was decreased at least 20-fold by formation of a quaternary complex of the enzyme with Mg²⁺, α,β-methylene ATP, and ribose-5-P. Trinitrobenzenesulfonate also irreversibly inactivated the enzyme, but this reagent was less specific in that the loss of activity corresponded to the modification of four to five lysine residues. These results suggest that an essential lysine is near the active site of Phosphoribosylpyrophosphate synthetase.     

Active-site-directed inactivation of wheat-germ aspartate transcarbamoylase by pyridoxal 5''-phosphate.

Treatment of 1 microM wheat-germ aspartate transcarbamoylase with 1 mM-pyridoxal 5'-phosphate caused a rapid loss of activity, concomitant with the formation of a Schiff base. Complete loss of activity occurred within 10 min when the Schiff base was reduced with a 100-fold excess of NaBH4. Concomitantly, one amino group per chain was modified. No further residues were modified in the ensuing 30 min. The kinetics of inactivation were examined under conditions where the Schiff base was reduced before assay. Inactivation was apparently first-order. The pseudo-first-order rate constant, kapp., showed a hyperbolic dependence upon the concentration of pyridoxal 5'-phosphate, suggesting that the enzyme first formed a non-covalent complex with the reagent, modification of a lysine then proceeding within this complex. Inactivation of the enzyme by pyridoxal was 20 times slower than that by pyridoxal 5'-phosphate, indicating that the phosphate group was important in forming the initial complex. Partial protection against pyridoxal phosphate was provided by the leading substrate, carbamoyl phosphate, and nearly complete protection was provided by the bisubstrate analogue, N-phosphonoacetyl-L-aspartate, and the ligand-pair carbamoyl phosphate plus succinate. Steady-state kinetic studies, under conditions that minimized inactivation, showed that pyridoxal 5'-phosphate was also a competitive inhibitor with respect to the leading substrate, carbamoyl phosphate. Pyridoxal 5'-phosphate therefore appears to be an active-site-directed reagent. A sample of the enzyme containing one reduced pyridoxyl group per chain was digested with trypsin, and the labelled peptide was isolated and shown to contain a single pyridoxyl-lysine residue. Partial sequencing around the labelled lysine showed little homology with the sequence surrounding lysine-84, an active-centre residue of the catalytic subunit of aspartate transcarbamoylase from Escherichia coli, whose reaction with pyridoxal 5'-phosphate shows many similarities to the results described in the present paper. Arguably the reactive lysine is conserved between the two enzymes whereas the residues immediately surrounding the lysine are not. The same conclusion has been drawn in a comparison of reactive histidine residues in the two enzymes [Cole & Yon (1986) Biochemistry 25, 7168-7174].     

The crystal structure of human muscle aldolase at 3.0 A resolution

The three-dimensional structure of fructose-1,6-bisphosphate aldolase from human muscle has been determined at 3.0 A resolution by X-ray crystallography. The active protein is a tetramer of 4 identical subunits each of which is composed of an eight-stranded alpha/beta-barrel structure. The lysine residue responsible for Schiff base formation with the substrate is located near the centre of the barrel in the middle of the sixth beta-strand. While the overall topology of the alpha/beta-barrel is very similar to those found in several other enzymes, the distribution of charged residues inside the core of the barrel seems distinct. The quaternary fold of human muscle aldolase uses interfacial regions also involved in the subunit association of other alpha/beta-barrel proteins found in glycolysis, but exploits these regions in a manner not seen previously.     

Human erythrocytes glucose-6-phosphate dehydrogenase: Labelling of a reactive lysyl residue by pyridoxal-5′-phosphate

Reaction of glucose-6-phosphate dehydrogenase from human erythrocytes with pyridoxal-5′-phosphate causes 80% loss of activity. The substrate glucose-6-phosphate fully protects the enzyme against this inhibition, which is reversible upon dilution, but becomes irreversible after treatment with NaBH₄. We presume that pyridoxal-5′-phosphate forms with the enzyme a Schiff base which is reduced by NaBH₄. One mole of N-?-pyridoxyl-lysine is formed per mole of enzyme subunit when the remaining activity reaches its minimal level of 20%.     

Cloning, expression, and purification of cytidine deaminase from Arabidopsis thaliana

The complementary DNA (cDNA) coding for Arabidopsis thaliana cytidine deaminase 1 (AT-CDA1) was obtained from the amplified A. thaliana cDNA expression library, provided by R. W. Davis (Stanford University, CA). AT-CDA1 cDNA was subcloned into the expression vector pTrc99-A and the protein, expressed in Escherichia coli following induction with isopropyl 1-thio-beta-d-galactopyranoside, showed high cytidine deaminase activity. The nucleotide sequence showed a 903-bp open reading frame encoding a polypeptide of 301 amino acids with a calculated molecular mass of 32,582. The deduced amino acid sequence of AT-CDA1 showed no transit peptide for targeting to the chloroplast or mitochondria indicating that this form of cytidine deaminase is probably expressed in the cytosol. The recombinant AT-CDA1 was purified to homogeneity by a heat treatment followed by an ion-exchange chromatography. The final enzyme preparation was >98% pure as judged by SDS-PAGE and showed a specific activity of 74 U/mg. The molecular mass of AT-CDA1 estimated by gel filtration was 63 kDa, indicating, in contrast to the other eukaryotic CDAs, that the enzyme is a dimer composed of two identical subunits. Inductively coupled plasma-optical emission spectroscopy analysis indicated that the enzyme contains 1 mol of zinc atom per mole of subunit. The kinetic properties of AT-CDA1 both toward the natural substrates and with analogs indicated that the catalytic mechanism of the plant enzyme is probably very similar to that of the human the E. coli enzymes.     

Role of cysteine-82 in the catalytic mechanism of human S-adenosylmethionine decarboxylase

S-Adenosylmethionine decarboxylase is a pyruvate-dependent enzyme. The enzyme forms a Schiff base with substrate, S-adenosylmethionine, through the pyruvoyl moiety. This facilitates the release of CO2 from the substrate, which must then be protonated on the alpha carbon in order to permit hydrolysis of the Schiff base to release the product. The catalytic mechanism of human S-adenosylmethionine decarboxylase was investigated via mutagenic and kinetic approaches. The results of enzyme kinetic studies indicated that Cys-82 is a crucial residue for activity and this residue has a basic pKa. Iodoacetic acid inhibited wild-type enzyme activity in a time- and pH-dependent manner but did not affect the already reduced activity of mutant C82A. Reaction of this mutant with iodoacetic acid led to approximately one less mole of reagent being incorporated per mole of enzyme alphabeta dimer than with wild-type S-adenosylmethionine decarboxylase. Both wild-type and C82A mutant S-adenosylmethionine decarboxylases were inactivated by substrate-mediated transamination, but this reaction occurred much more frequently with C82A than with wild-type enzyme. A major proportion of the recombinant C82A mutant protein was in the transaminated form in which the pyruvoyl cofactor is converted into alanine. This suggests that incorrect protonation of the pyruvate, rather than the substrate, occurs much more readily when Cys-82 is altered. On the basis of these results, it was postulated that residue Cys-82 may be the proton donor of the decarboxylation reaction catalyzed by S-adenosylmethionine decarboxylase.     

Effect of dicyclohexylcarbodiimide on unisite and multisite catalytic activities of the adenosinetriphosphatase of Escherichia coli

The inhibitory effect of dicyclohexylcarbodiimide (DCCD) on the activity of the adenosine-triphosphatase of Escherichia coli (ECF1) has been examined in detail. DCCD reacted with ECF1 predominantly in beta subunits with a maximum of 2 mol of reagent per mole of ECF1 being incorporated in these subunits. Ninety-five percent inhibition of steady-state or multistate ATPase activity required incorporation of 1 mol of DCCD per mole of enzyme into beta subunits. Seventy-five percent inhibition of the initial rate of unisite catalysis was only obtained after incorporation of 2 mol of DCCD per mole of ECF1 into beta subunits. Analyses of the kinetics of unisite catalysis and nucleotide binding experiments both indicate that DCCD binds outside the substrate ATP binding site. Inhibition by this reagent appears to be due in part to an effect on the catalytic sites but mainly to the blocking of cooperativity between these sites.     

Mechanism of the Schiff base forming fructose-1,6-bisphosphate aldolase: structural analysis of reaction intermediates

The glycolytic enzyme fructose-1,6-bisphosphate aldolase (FBPA) catalyzes the reversible cleavage of fructose 1,6-bisphosphate to glyceraldehyde 3-phosphate and dihydroxyacetone phosphate. Catalysis of Schiff base forming class I FBPA relies on a number of intermediates covalently bound to the catalytic lysine. Using active site mutants of FBPA I from Thermoproteus tenax, we have solved the crystal structures of the enzyme covalently bound to the carbinolamine of the substrate fructose 1,6-bisphosphate and noncovalently bound to the cyclic form of the substrate. The structures, determined at a resolution of 1.9 A and refined to crystallographic R factors of 0.148 and 0.149, respectively, represent the first view of any FBPA I in these two stages of the reaction pathway and allow detailed analysis of the roles of active site residues in catalysis. The active site geometry of the Tyr146Phe FBPA variant with the carbinolamine intermediate supports the notion that in the archaeal FBPA I Tyr146 is the proton donor catalyzing the conversion between the carbinolamine and Schiff base. Our structural analysis furthermore indicates that Glu187 is the proton donor in the eukaryotic FBPA I, whereas an aspartic acid, conserved in all FBPA I enzymes, is in a perfect position to be the general base facilitating carbon-carbon cleavage. The crystal structure of the Trp144Glu, Tyr146Phe double-mutant substrate complex represents the first example where the cyclic form of beta-fructose 1,6-bisphosphate is noncovalently bound to FBPA I. The structure thus allows for the first time the catalytic mechanism of ring opening to be unraveled.     

5-Chlorolevulinate modification of porphobilinogen synthase identifies a potential role for the catalytic zinc.

Porphobilinogen synthase (PBGS) is a Zn(II) metalloenzyme which catalyzes the asymmetric condensation of two molecules of 5-aminolevulinate (ALA). The nitrogen of the first substrate ends up in the pyrrole ring of product (P-side ALA); by contrast, the nitrogen of the second substrate molecule remains an amino group (A-side ALA). A reactive mimic of the substrate molecules, 5-chlorolevulinate (5-CLA), has been prepared and used as an active site directed irreversible inhibitor of PBGS. Native octameric PBGS binds eight substrate molecules and eight Zn(II) ions, with two types of sites for each ligand. As originally demonstrated by Seehra and Jordan [(1981) Eur. J. Biochem. 113, 435-446], 5-CLA inactivates the enzyme at the site where one of the two substrate molecules binds, and modification at four sites per octamer (one per active site) affords near-total inactivation. Here we report that 5-CLA-modified PBGS (5-CLA-PBGS) can bind up to four substrate molecules and four Zn(II) ions. Contrary to the conclusion of Seehra and Jordan, we find that the preferential site of 5-CLA inactivation is the A-side ALA binding site. On the basis of the dissociation constants, the metal ion binding sites lost upon 5-CLA modification are assigned to the four catalytic Zn(II) sites. 5-CLA-PBGS is shown to be modified at cysteine-223 on half of the subunits. We conclude that cysteine-223 is near the amino group of A-side ALA and propose that this cysteine is a ligand to the catalytic Zn(II). The vacant substrate binding site on 5-CLA-PBGS is that of P-side ALA. We have used 13C and 15N NMR to view [4-13C]ALA and [15N]ALA bound to 5-CLA-PBGS. The NMR results are nearly identical to those obtained previously for the enzyme-bound P-side Schiff base intermediate [Jaffe et al. (1990) Biochemistry 29, 8345-8350]. It appears that, in the absence of the catalytic Zn(II), 5-CLA-PBGS does not catalyze the condensation of the amino group of the P-side Schiff base intermediate with the C4 carbonyl derived from 5-CLA. On this basis we propose that Zn(II) plays an essential role in formation of the first bond between the two substrate molecules.     

Twisted Schiff base intermediates and substrate locale revise transaldolase mechanism

We examined the catalytic cycle of transaldolase (TAL) from Thermoplasma acidophilum by cryocrystallography and were able to structurally characterize--for the first time, to our knowledge--different genuine TAL reaction intermediates. These include the Schiff base adducts formed between the catalytic lysine and the donor ketose substrates fructose-6-phosphate and sedoheptulose-7-phosphate as well as the Michaelis complex with acceptor aldose erythrose-4-phosphate. These structural snapshots necessitate a revision of the accepted reaction mechanism with respect to functional roles of active site residues, and they further reveal fundamental insights into the general structural features of enzymatic Schiff base intermediates and the role of conformational dynamics in enzyme catalysis, substrate binding and discrimination. A nonplanar arrangement of the substituents around the Schiff base double bond was observed, suggesting that a structurally encoded reactant-state destabilization is a driving force of catalysis. Protein dynamics and the intrinsic hydrogen-bonding pattern appear to be crucial for selective recognition and binding of ketose as first substrate.     

NgoII, a restriction endonuclease from Neisseria gonorrhoeae.

EndoR . NgoII, a class II restriction endonuclease isolated from Neisseria gonorrhoeae, was purified to electrophoretic homogeneity. We were able to separate it from another restriction endonuclease of N. gonorrhoeae, NgoI, by phosphocellulose chromatography. NgoII is an isoschizomer of HaeIII, a restriction endonuclease of Haemophilus aegyptius, and was found to recognize the deoxyribonucleic acid nucleotide base sequence GGCC. NgoII was able to digest phage lambda deoxyribonucleic acid over a wide pH range, with optimal activity at pH 8.5. The enzyme has an absolute requirement for Mg2+; maximal enzyme activity was observed at 1 mM Mg2+. The active enzyme has a molecular weight of 65,000 and appears to be composed of six subunits of identical molecular weight (11,000). No methylase activity could be detected in the purified enzyme preparation.     

The Schiff base complex of yeast 5-aminolaevulinic acid dehydratase with laevulinic acid.

The X-ray structure of the complex formed between yeast 5-aminolaevulinic acid dehydratase (ALAD) and the inhibitor laevulinic acid has been determined at 2.15 A resolution. The inhibitor binds by forming a Schiff base link with one of the two invariant lysines at the catalytic center: Lys263. It is known that this lysine forms a Schiff base link with substrate bound at the enzyme's so-called P-site. The carboxyl group of laevulinic acid makes hydrogen bonds with the side-chain-OH groups of Tyr329 and Ser290, as well as with the main-chain >NH group of Ser290. The aliphatic moiety of the inhibitor makes hydrophobic interactions with surrounding aromatic residues in the protein including Phe219, which resides in the flap covering the active site. Our analysis strongly suggests that the same interactions will be made by P-side substrate and also indicates that the substrate that binds at the enzyme's A-site will interact with the enzyme's zinc ion bound by three cysteines (133, 135, and 143). Inhibitor binding caused a substantial ordering of the active site flap (residues 217-235), which was largely invisible in the native electron density map and indicates that this highly conserved yet flexible region has a specific role in substrate binding during catalysis.     

Basis for half-of-the-site reactivity and the dominance of the K487 oriental subunit over the E487 subunit in heterotetrameric human liver mitochondrial aldehyde dehydrogenase

Human liver mitochondrial aldehyde dehydrogenase is a tetrameric enzyme composed of 4 identical 500 amino acid containing subunits arranged such that the protein is a dimer of dimers. No kinetic evidence for subunit interactions has been reported. However, the enzyme exhibits half-of-the-site reactivity in that there is a pre-steady-state burst of 2 mol of NADH per mole of enzyme. A variant of the enzyme, found in Asian people, contains a lysine rather than a glutamate at position 487. This enzyme has a high K(M) for NAD(+) and a low specific activity. In heterotetramers composed of both subunit types, it appeared that the lysine-containing subunit was dominant over the glutamate-containing subunits. To allow for the separation of various heterotetrameric forms of the enzyme, surface residues were changed. Each of the five possible tetrameric forms of the modified enzyme was isolated and characterized with respect to steady-state kinetics and pre-steady-state burst magnitudes. The data best fit a model where in each dimer pair there is one functioning and one nonfunctioning subunit. Further, the lysine subunit affects the properties only of its dimer partner. Residue 487 is located at the dimer interface, and the glutamate forms salt bonds with two arginine residues. One is to Arg(264) in the same subunit; the other is to Arg(475) located in the other subunit. Most likely the presence of a lysine affects these salt bonds so the lysine subunit can cause the other subunit to become essentially nonfunctional.     

The X-ray structure of yeast 5-aminolaevulinic acid dehydratase complexed with two diacid inhibitors

The structures of 5-aminolaevulinic acid dehydratase complexed with two irreversible inhibitors (4-oxosebacic acid and 4,7-dioxosebacic acid) have been solved at high resolution. Both inhibitors bind by forming a Schiff base link with Lys 263 at the active site. Previous inhibitor binding studies have defined the interactions made by only one of the two substrate moieties (P-side substrate) which bind to the enzyme during catalysis. The structures reported here provide an improved definition of the interactions made by both of the substrate molecules (A- and P-side substrates). The most intriguing result is the novel finding that 4,7-dioxosebacic acid forms a second Schiff base with the enzyme involving Lys 210. It has been known for many years that P-side substrate forms a Schiff base (with Lys 263) but until now there has been no evidence that binding of A-side substrate involves formation of a Schiff base with the enzyme. A catalytic mechanism involving substrate linked to the enzyme through Schiff bases at both the A- and P-sites is proposed.