Transcarboxylase: role of biotin, metals, and subunits in the reaction and its quaternary structure

Chiral one- and two-dimensional silver(I)-biotin coordination polymers

Reaction of biotin {C(10)H(16)N(2)O(3)S, HL; systematic name: 5-[(3aS,4S,6aR)-2-oxohexahydro-1H-thieno[3,4-d]imidazol-4-yl]pentanoic acid} with silver acetate and a few drops of aqueous ammonia leads to the deprotonation of the carboxylic acid group and the formation of a neutral chiral two-dimensional polymer network, poly[[{μ(3)-5-[(3aS,4S,6aR)-2-oxohexahydro-1H-thieno[3,4-d]imidazol-4-yl]pentanoato}silver(I)] trihydrate], {[Ag(C(10)H(15)N(2)O(3)S)]·3H(2)O}(n) or {[Ag(L)]·3H(2)O}(n), (I). Here, the Ag(I) cations are pentacoordinate, coordinated by four biotin anions via two S atoms and a ureido O atom, and by two carboxylate O atoms of the same molecule. The reaction of biotin with silver salts of potentially coordinating anions, viz. nitrate and perchlorate, leads to the formation of the chiral one-dimensional coordination polymers catena-poly[[bis[nitratosilver(I)]-bis{μ(3)-5-[(3aS,4S,6aR)-2-oxohexahydro-1H-thieno[3,4-d]imidazol-4-yl]pentanoato}] monohydrate], {[Ag(2)(NO(3))(2)(C(10)H(16)N(2)O(3)S)(2)]·H(2)O}(n) or {[Ag(2)(NO(3))(2)(HL)(2)]·H(2)O}(n), (II), and catena-poly[bis[perchloratosilver(I)]-bis{μ(3)-5-[(3aS,4S,6aR)-2-oxohexahydro-1H-thieno[3,4-d]imidazol-4-yl]pentanoato}], [Ag(2)(ClO(4))(2)(C(10)H(16)N(2)O(3)S)(2)](n) or [Ag(2)(ClO(4))(2)(HL)(2)](n), (III), respectively. In (II), the Ag(I) cations are again pentacoordinated by three biotin molecules via two S atoms and a ureido O atom, and by two O atoms of a nitrate anion. In (I), (II) and (III), the Ag(I) cations are bridged by an S atom and are coordinated by the ureido O atom and the O atoms of the anions. The reaction of biotin with silver salts of noncoordinating anions, viz. hexafluoridophosphate (PF(6)(-)) and hexafluoridoantimonate (SbF(6)(-)), gave the chiral double-stranded helical structures catena-poly[[silver(I)-bis{μ(2)-5-[(3aS,4S,6aR)-2-oxohexahydro-1H-thieno[3,4-d]imidazol-4-yl]pentanoato}] hexafluoridophosphate], {[Ag(C(10)H(16)N(2)O(3)S)(2)](PF(6))}(n) or {[Ag(HL)(2)](PF(6))}(n), (IV), and catena-poly[[[{5-[(3aS,4S,6aR)-2-oxohexahydro-1H-thieno[3,4-d]imidazol-4-yl]pentanoato}silver(I)]-μ(2)-{5-[(3aS,4S,6aR)-2-oxohexahydro-1H-thieno[3,4-d]imidazol-4-yl]pentanoato}] hexafluoridoantimonate], {[Ag(C(10)H(16)N(2)O(3)S)(2)](SbF(6))}(n) or {[Ag(HL)(2)](SbF(6))}(n), (V), respectively. In (IV), the Ag(I) cations have a tetrahedral coordination environment, coordinated by four biotin molecules via two S atoms, and by two carboxy O atoms of two different molecules. In (V), however, the Ag(I) cations have a trigonal coordination environment, coordinated by three biotin molecules via two S atoms and one carboxy O atom. In (IV) and (V), neither the ureido O atom nor the F atoms of the anion are involved in coordination. Hence, the coordination environment of the Ag(I) cations varies from AgS(2)O trigonal to AgS(2)O(2) tetrahedral to AgS(2)O(3) square-pyramidal. The conformation of the valeric acid side chain varies from extended to twisted and this, together with the various anions present, has an influence on the solid-state structures of the resulting compounds. The various O-H···O and N-H···O hydrogen bonds present result in the formation of chiral two- and three-dimensional networks, which are further stabilized by C-H···X (X = O, F, S) interactions, and by N-H···F interactions for (IV) and (V). Biotin itself has a twisted valeric acid side chain which is involved in an intramolecular C-H···S hydrogen bond. The tetrahydrothiophene ring has an envelope conformation with the S atom as the flap. It is displaced from the mean plane of the four C atoms (plane B) by 0.8789 (6) Å, towards the ureido ring (plane A). Planes A and B are inclined to one another by 58.89 (14)°. In the crystal, molecules are linked via O-H···O and N-H···O hydrogen bonds, enclosing R(2)(2)(8) loops, forming zigzag chains propagating along [001]. These chains are linked via N-H···O hydrogen bonds, and C-H···S and C-H···O interactions forming a three-dimensional network. The absolute configurations of biotin and complexes (I), (II), (IV) and (V) were confirmed crystallographically by resonant scattering.

Mechanism to combat cobalt toxicity in cobalt resistant mutants of Aspergillus nidulans

Characteristics of fungal species tolerant to high levels of metals in natural environment can be amplified by isolation and selection of resistant mutants. Step-by-step culturing led to identification of highly stable Co-resistant (Co(R)) mutants of A. nidulans. Based on two distinct morphological features, Co-resistant mutants were categorized as Co(R)I and Co(R)II. The two mutants varied in their growth behavior and colony morphology that were reflected in supplemented as well as unsupplemented growth media over the generations. As compared to the Co(R)I, Co(R)II mutant exhibited sparse mycelia and conidiation but secreted higher amount of melanin. Co(R) mutants could tolerate up to 2.5mM Co in the medium, however, required a threshold concentration of 0.25mM Co for optimal growth and germination. Absence of Co in the medium caused a stressful situation for the Co(R) mutants and led to the secretion of a white extracellular precipitate found to be a glycoprotein. In response to interactions with Co-ions, Co(R) mutants produced oxalic acid and bioprecipitated Co as Co-oxalate providing scope for metal reclamation as well as oxalic acid extraction. The mutants could help to recover the insoluble Co-oxalate salt from aqueous solutions by entrapping it in their growing mycelial meshwork.

New and easy strategy for cloning, expression, purification, and characterization of the 5S subunit of transcarboxylase from Propionibacterium f. shermanii

Methylmalonyl CoA-oxalacetate transcarboxylase (EC 2. 1. 3. 1) from Propionibacterium f. shermanii is a biotin dependent enzyme which transfers CO2 from methylmalonyl-CoA (MMCoA) to pyruvate via a carboxylated biotin group to form oxalacetate. It is composed of three subunits, the central cylindrical hexameric 12S subunit, the outer six dimeric 5S subunit, and the twelve 1.3S linkers. We here report the cloning, sequencing, expression, and purification of the 5S subunit. The gene was identified by matching the amino acid sequence with that of deposited in the NCBI database. For cloned 5S subunit sequence shows regions of high homology with that of pyruvate carboxylase and oxaloacetate decarboxylase. The gene encoding the 5S subunit was cloned into the pTXB1 vector. The expressed 5S subunit was purified to apparent homogeneity by a single step process by using Intein mediated protein ligation (IPL) method. The cloned 5S gene encodes a protein of 505 amino acids and of M(r) 55,700.

RAMAN CRYSTALLOGRAPHY AND OTHER BIOCHEMICAL APPLICATIONS OF RAMAN MICROSCOPY

Recent studies using a Raman microscope have shown that single protein crystals provide an ideal platform to undertake Raman difference spectroscopic analyses under nonresonance conditions. This approach, termed Raman crystallography, provides a means of characterizing chemical events within the crystal such as ligand binding and enzyme reactions. In many cases Raman crystallography goes hand in hand with X-ray crystallographic studies because the Raman results can inform the X-ray crystallographer about the status of chemical events in the crystal prior to flash freezing and X-ray analysis. In turn, the combined data from the Raman and X-ray analyses are highly synergistic and offer novel perspectives on structure and dynamics in enzyme active sites. In a related area, protein misfolding, Raman microscopy can provide detailed insights into the chemistry of the amyloid plaques associated with Alzheimer's disease and into the intermediates on the alpha-synuclein protein misfolding pathway implicated in Parkinson's disease.

Proteins can convert to beta-sheet in single crystals

Raman microscopy was used to follow conformational changes in single protein crystals. Crystals of native insulin and of the 5S and 12S subunits of the enzyme transcarboxylase showed a mixture of Raman marker bands signifying alpha-helix, beta-sheet and nonordered secondary structure. However, by reducing the S-S bonds in the insulin crystal, or by lowering the pH for the 5S crystal, or by soaking substrates into the 12S crystal, the secondary structure in each crystal became predominantly beta-sheet. The beta-form crystals could be dissolved only with difficulty and yielded high-molecular weight protein aggregates, indicating that the beta-sheet formation involves intermolecular contacts. Although their morphology appeared unchanged, the crystals no longer diffracted X-rays. Using crystals that had not been exposed to laser light, the dye thioflavin T formed highly fluorescent complexes with the "beta-transformed" crystals but not with the native crystals.

Transcarboxylase 5S structures: assembly and catalytic mechanism of a multienzyme complex subunit

Transcarboxylase is a 1.2 million Dalton (Da) multienzyme complex from Propionibacterium shermanii that couples two carboxylation reactions, transferring CO(2)(-) from methylmalonyl-CoA to pyruvate to yield propionyl-CoA and oxaloacetate. Crystal structures of the 5S metalloenzyme subunit, which catalyzes the second carboxylation reaction, have been solved in free form and bound to its substrate pyruvate, product oxaloacetate, or inhibitor 2-ketobutyrate. The structure reveals a dimer of beta(8)alpha(8) barrels with an active site cobalt ion coordinated by a carbamylated lysine, except in the oxaloacetate complex in which the product's carboxylate group serves as a ligand instead. 5S and human pyruvate carboxylase (PC), an enzyme crucial to gluconeogenesis, catalyze similar reactions. A 5S-based homology model of the PC carboxyltransferase domain indicates a conserved mechanism and explains the molecular basis of mutations in lactic acidemia. PC disease mutations reproduced in 5S result in a similar decrease in carboxyltransferase activity and crystal structures with altered active sites.

Transcarboxylase 12S crystal structure: hexamer assembly and substrate binding to a multienzyme core

Transcarboxylase from Propionibacterium shermanii is a 1.2 MDa multienzyme complex that couples two carboxylation reactions, transferring CO(2)(-) from methylmalonyl-CoA to pyruvate, yielding propionyl-CoA and oxaloacetate. The 1.9 A resolution crystal structure of the central 12S hexameric core, which catalyzes the first carboxylation reaction, has been solved bound to its substrate methylmalonyl-CoA. Overall, the structure reveals two stacked trimers related by 2-fold symmetry, and a domain duplication in the monomer. In the active site, the labile carboxylate group of methylmalonyl-CoA is stabilized by interaction with the N-termini of two alpha-helices. The 12S domains are structurally similar to the crotonase/isomerase superfamily, although only domain 1 of each 12S monomer binds ligand. The 12S reaction is similar to that of human propionyl-CoA carboxylase, whose beta-subunit has 50% sequence identity with 12S. A homology model of the propionyl-CoA carboxylase beta-subunit, based on this 12S crystal structure, provides new insight into the propionyl-CoA carboxylase mechanism, its oligomeric structure and the molecular basis of mutations responsible for enzyme deficiency in propionic acidemia.

Purification and characterization of two components of epoxypropane isomerase/carboxylase from Xanthobacter Py2

Epoxypropane isomerase from Xanthobacter Py2 has been resolved into at least two components (A and B) by ion-exchange chromatography. Both components were required for the degradation of epoxypropane and were purified further. Component A was apparently homohexameric with a subunit M(r) of about 44,000, and possessed NAD(+)-dependent dihydrolipoamide dehydrogenase activity and lipoamide reductase activity. It was sensitive to inhibition by o-phenanthroline and the thiol-specific reagents N-ethylmaleimide(NEM)and p-chloromercuribenzoate. Component B was homodimeric with a subunit M(r), of 62,170 and contained 2 mol.mol-1 FAD. It had an NADPH-dependent lipoamide reductase activity which was sensitive to NEM and p-chloromercuribenzoate. The N-terminal amino acid sequences and monomer sizes of components A and B correspond to those of ORF1 and ORF3 respectively (ORF = open reading frame) of a recently published sequence of a clone which complements mutants unable to degrade epoxypropane. NADPH was found to replace the need for a low-M(r), fraction in epoxypropane degradation assays containing components A and B and NAD+. The predicted amino acid sequence of component A (ORF1) has been analysed and shown to contain a potential ADP binding site near the N-terminus and putative cofactor binding domain near the C-terminus, with sequence similarity to the biotinyl and lipoyl binding domains of biotin-dependent carboxylases and 2-oxoacid dehydrogenases respectively. A reaction mechanism for epoxypropane degradation, incorporating recent evidence for combined isomerization and carboxylation to acetoacetate, is discussed.

The structure and the mechanism of action of pyruvate carboxylase

Pyruvate carboxylase plays an important role in intermediary metabolism, catalysing the formation of oxaloacetate from pyruvate and HCO3-, with concomitant ATP cleavage. It thus provides oxaloacetate for gluconeogenesis and replenishing tricarboxylic acid cycle intermediates for fatty acid, amino acid and neurotransmitter synthesis. The enzyme is highly conserved and is found in a great variety of organisms including fungi, bacteria and plants as well as higher organisms. It is a member of a group of biotin-dependent enzymes and the biotin prosthetic group is covalently bound to the polypeptide chain of the enzyme, there normally being four such chains in the native, tetrameric enzyme. The overall reaction catalysed by pyruvate carboxylase involves two partial reactions that occur at spatially separate subsites within the active site, with the covalently bound biotin acting as a mobile carboxyl group carrier. In the first partial reaction, biotin is carboxylated using ATP and HCO3- as substrates whilst in the second partial reaction, the carboxyl group from carboxybiotin is transferred to pyruvate. The chemical mechanisms of the partial reactions and some of the roles played by amino acid residues of the enzyme in catalysing the reaction have been elucidated. The domain structure of the yeast enzyme has been deduced by comparing its amino acid sequence with those of enzymes that have similar catalytic functions. The quaternary structures of the pyruvate carboxylases studied so far, all involve a tetrahedron-like arrangement of the subunits. The major regulator of enzyme activity, acetyl CoA, stimulates the cleavage of ATP in the first partial reaction and in addition it has been shown to induce a conformational change in the tetrameric structure of the enzyme. In the past, the lack of any detailed structural information on the enzyme has hampered efforts to fully understand how this and other biotin-dependent enzymes function and are regulated. With the recent cloning of the enzyme from a variety of sources and the performance of three-dimensional structural studies, the next few years should see much progress in our understanding the mechanism of action of this enzyme.

Enzymes: chemistry and biochemistry

Basic principles underlying enzyme action are considered. Catalytic antibodies (abzymes), catalytic RNA (ribozymes), and non-biological counterparts of enzyme-catalyzed reactions are mentioned. Enzyme evolution is considered in terms of divergence, convergence, and lateral gene transfer.

Primary structure of the 5 S subunit of transcarboxylase as deduced from the genomic DNA sequence

Transcarboxylase from Propionibacterium shermanii is a complex biotin-containing enzyme composed of 30 polypeptides of three different types. It is composed of six dimeric outer subunits associated with a central cylindrical hexameric subunit through 12 biotinyl subunits; three outer subunits on each face of the central hexamer. Each outer dimer is termed a 5 S subunit which associates with two biotinyl subunits. The enzyme catalyzes a two-step reaction in which methylmalonyl-CoA and pyruvate form propionyl-CoA and oxalacetate, the 5 S subunit specifically catalyzing one of these reactions. We report here the cloning, sequencing and expression of the monomer of the 5 S subunit. The gene was identified by matching amino acid sequences derived from isolated authentic 5 S peptides with the deduced sequence of an open reading frame present on a cloned P. shermanii genomic fragment known to contain the gene encoding the 1.3 S biotinyl subunit. The cloned 5 S gene encodes a protein of 519 amino acids, M(r) 57,793. The deduced sequence shows regions of extensive homology with that of pyruvate carboxylase and oxalacetate decarboxylase, two enzymes which catalyze the same or reverse reaction. A fragment was subcloned into pUC19 in an orientation such that the 5 S open reading frame could be expressed from the lac promoter of the vector. Crude extracts prepared from these cells contained an immunoreactive band on Western blots which co-migrated with authentic 5 S and were fully active in catalyzing the 5 S partial reaction. We conclude that we have cloned, sequenced and expressed the monomer of the 5 S subunit and that the expressed product is catalytically active.

Primary structure of the monomer of the 12S subunit of transcarboxylase as deduced from DNA and characterization of the product expressed in Escherichia coli

Transcarboxylase from Propionibacterium shermanii is a complex biotin-containing enzyme composed of 30 polypeptides of three different types: a hexameric central 12S subunit to which 6 outer 5S subunits are attached through 12 1.3S biotinyl subunits. The enzyme catalyzes a two-step reaction in which methylmalonyl coenzyme A and pyruvate serve as substrates to form propionyl coenzyme A (propionyl-CoA) and oxalacetate, the 12S subunit specifically catalyzing one of the two reactions. We report here the cloning, sequencing, and expression of the 12S subunit. The gene was identified by matching amino acid sequences derived from isolated authentic 12S peptides with the deduced sequence of an open reading frame present in a cloned P. shermanii genomic fragment known to contain the gene encoding the 1.3S biotinyl subunit. The cloned 12S gene encodes a protein of 604 amino acids and of M(r) 65,545. The deduced sequence shows regions of extensive homology with the beta subunit of mammalian propionyl-CoA carboxylase as well as regions of homology with acetyl-CoA carboxylase from several species. Two genomic fragments were subcloned into pUC19 in an orientation such that the 12S open reading frame could be expressed from the lac promoter of the vector. Crude extracts prepared from these cells contained an immunoreactive band on Western blots (immunoblots) which comigrated with authentic 12S. The Escherichia coli-expressed 12S was purified to apparent homogeneity by a three-step procedure and compared with authentic 12S from P. shermanii. Their quaternary structures were identical by electron microscopy, and the E. coli 12S preparation was fully active in the reactions catalyzed by this subunit. We conclude that we have cloned, sequenced, and expressed the 12S subunit which exists in a hexameric active form in E.coli.

Cloning and structural characterization of the genes coding for adenosylcobalamin-dependent methylmalonyl-CoA mutase from Propionibacterium shermanii

The structural genes coding for both subunits of adenosylcobalamin-dependent methylmalonyl-CoA mutase from the Gram-positive bacterium Propionibacterium shermanii have been cloned, with the use of synthetic oligonucleotides as primary hybridization probes. The genes are closely linked and are transcribed in the same direction. Nucleotide sequence analysis of 4.5 kb of DNA encompassing both genes allowed us to infer the complete amino acid sequence of the two subunits: the beta-subunit is the product of the upstream gene, and consists of 638 amino acid residues (Mr 69465) and the alpha-subunit consists of 728 amino acid residues (Mr 80,147). There is a very close structural homology between the two subunits, reflecting the probable duplication of a common ancestral gene. A sequence present only in the alpha-subunit is significantly homologous to a portion of the sequence of the methylmalonyl-CoA-binding subunit of transcarboxylase from P. shermanii [Samols, Thornton, Murtif, Kumar, Haase & Wood (1988) J. Biol. Chem. 263, 6461-6464], and this homologous region may form part of the CoA ester-binding site in both enzymes.

A rapid purification method for rat liver pyruvate carboxylase and amino acid sequence analyses of NH2-terminal and biotin peptide

A rapid method for the purification of pyruvate carboxylase from rat liver has been developed. The method involves extraction of the enzyme from frozen liver powder followed by polyethylene glycol fractionation and avidin-affinity chromatography. The purified enzyme has a specific activity of 9-10 mumol/min/mg protein when assayed at 22 degrees C in the presence of acetyl-CoA. Polyacrylamide gel electrophoresis of the preparation in the presence of sodium dodecyl sulfate showed the presence of one protein band with an estimated Mr 125,000 and no significant contamination by other biotin-containing enzymes. In addition to being rapid, the method is advantageous because prior isolation of mitochondria is not necessary. Using these preparations we have determined the sequence of the first 15 amino acids from the NH2-terminal end of the molecule to be Ser-Gly-Pro-Val-Ala-Pro-Leu-Asn-Val-Leu-Leu-Leu-Glu-Tyr-Pro. The sequence of the 24 amino acid residues around the biotin site was determined to be Gly-Ala-Pro-Leu-Val-Leu-Ser-Ala-Met-biocytin-Met-Glu-Thr-Val-Val-Thr-Ser -Pro- Thr-Glu-Gly-Thr-Ile-Arg.

Sequence homology around the biotin-binding site of human propionyl-CoA carboxylase and pyruvate carboxylase

Biotin-dependent carboxylases require covalently bound biotin for enzymatic activity. The biotin is attached through a lysine residue, which in a number of bacterial, avian, and mammalian carboxylases, is found within the conserved sequence Ala-Met-Lys-Met. We have determined the partial nucleotide sequence of cDNA clones for human propionyl-CoA carboxylase and pyruvate carboxylase. The predicted amino acid sequence of both these proteins contains the conserved tetrapeptide 35 residues from the carboxy terminus. In addition, both proteins contain the tripeptide, Pro-Met-Pro, 26 residues toward the amino terminus from the biotin attachment site. The overall amino acid homology through this region is 43%. Similar findings have been made for the biotin-containing polypeptides of transcarboxylase of Propionibacterium shermanii and acetyl-CoA carboxylase of Escherichia coli (W. L. Maloy, B. U. Bowien, G. K. Zwolinski, K. G. Kumar, and H. G. Wood (1979) J. Biol. Chem. 254, 11615-11622). The implications of this sequence conservation with regard to the function and evolution of biotin-dependent carboxylases is discussed. We propose that the 60 amino acids surrounding the biotin site are bounded by a proline "hinge" and the carboxy terminus has remained conserved as a result of constraints imposed by biotinylation of the enzyme.

Preliminary crystallographic data and quaternary structural implications of the central subunit of the multi-subunit complex transcarboxylase

The hexameric central subunit (Mr = 360,000) of the multi-subunit complex transcarboxylase has been crystallized by bulk dialysis against 250 mM-sodium acetate (pH 5.5). The crystals are cubic, a = 193.1 A, space group P4(1)32 or enantiomorph. The number of molecules per unit cell is four and was deduced from the density of the crystals (1.10 g cm-3) and the mother liquor (1.01 g cm-3) and the specific volume of the protein calculated from molecular dimensions obtained from electron microscopy studies. Four molecules per cell requires the central subunits to lie on 3-fold axes, which are perpendicular to 2-fold rotation axes, so that the molecules satisfy 32 symmetry giving one subunit as the asymmetric unit. Of the four possible models that have been considered for the quaternary structure of transcarboxylase, only that with antiparallel subunits, two sets of isologous binding sites and D3 symmetry is in agreement with the symmetry requirements of the cubic crystals.

Isolation and characterization of DL-methylmalonyl-coenzyme A racemase from rat liver

Certain amino acids and other compounds are metabolized via propionyl-CoA----D-methylmalonyl-CoA----L-methylmalonyl- CoA----succinyl-CoA----tricarboxylic acid cycle. D-Methylmalonyl-CoA can also be converted to methylmalonic acid and coenzyme A by a specific hydrolase that does not act on L-methylmalonyl-CoA [R.J. Kovachy, S.D. Copley, and R.H. Allen (1983) J. Biol. Chem. 258, 11415-11421]. Because little is known about mammalian DL-methylmalonyl-CoA racemase and because it is involved in the flow of D-methylmalonyl-CoA to L-methylmalonyl-CoA----tricarboxylic acid cycle (versus to methylmalonic acid), we developed a new assay and purified rat liver racemase 23,000-fold to homogeneity. The molecular weight of the racemase is 32,000 and it contains two subunits of Mr 16,000 that are not connected by disulfide bonds. The rat liver and the rat and human white blood cell racemase are immunologically related. They are completely inactivated by EDTA and can be activated by the addition of Co+2, with 50% activation occurring at a concentration of 0.2 microM. Lower levels for maximal activation were obtained with higher concentrations of Co+3, Fe+2, and Mn+2. Other metals such as Zn+2, Cu+2, Cu+1, and Cd+2 completely inhibited racemase even in the presence of equal concentrations of Co+2. The purified racemase appears to bind 1 mol Co/mol subunit.

Interactions of biotin with metal ions. X-ray crystal structure of the polymeric biotin--silver(I) nitrate complex: metal bonding to thioether and ureido carbonyl groups

X-ray structure analysis of the silver(I) complex of d-(+)-biotin, [Ag(biotin)(NO3)] X 0.5H2O, shows the complex to be polymeric with the silver ion coordinated tetrahedrally to nitrate and to three different biotin molecules. Binding to the latter involves two thioether sulfur atoms in the cis and trans direction with respect to the ureido ring, and one ureido carbonyl oxygen atom. The biotin carboxylate group is probably protonated and not coordinated with the silver ion.

The anatomy of transcarboxylase and the role of its subunits

Biotin enzymes in general catalyze the fixation of CO2 and in a few instances decarboxylations yielding CO2. Transcarboxylase is an exception; it catalyzes the transfer of a carboxyl group from one compound to another and CO2 is not involved. This enzyme plays an essential role in the formation of propionic acid by propionibacteria and its structure and catalytic mechanism have been extensively investigated including studies of the quaternary structure by electron microscopy. The structure is complex, consisting of three types of subunits: (1) a central hexameric subunit, (2) six dimeric outside subunits, and (3) twelve biotinyl subunits which bind the outside subunits to the central subunit. There are 12 substrate sites on the central subunit (2 per polypeptide) and 2 substrate sites on each of the dimeric outside subunits. The carboxyl is transferred between these sites via the biotin of the biotinyl subunit. The biotinyl subunit (approximately 123 residues) has been completely sequenced and it has been shown that the first 42 residues serve in binding the outside subunits to the central subunit and the remainder of the sequence is involved in placing the biotin between the subunits so that it may serve as the carboxyl carrier between the substrate sites on the central and outside subunits. It is proposed that the dual sites on the polypeptides of the central subunit have arisen as a consequence of gene duplication and fusion. An intriguing question is why such a complicated structure is required for catalysis of a rather simple reaction.