Influence of an aggregated multienzyme system on transient time: kinetic evidence for compartmentation by an aromatic-amino-acid synthesizing complex of Neurospora crassa

Structure, function, and mechanism of proline utilization A (PutA)

Proline has important roles in multiple biological processes such as cellular bioenergetics, cell growth, oxidative and osmotic stress response, protein folding and stability, and redox signaling. The proline catabolic pathway, which forms glutamate, enables organisms to utilize proline as a carbon, nitrogen, and energy source. FAD-dependent proline dehydrogenase (PRODH) and NAD⁺-dependent glutamate semialdehyde dehydrogenase (GSALDH) convert proline to glutamate in two sequential oxidative steps. Depletion of PRODH and GSALDH in humans leads to hyperprolinemia, which is associated with mental disorders such as schizophrenia. Also, some pathogens require proline catabolism for virulence. A unique aspect of proline catabolism is the multifunctional proline utilization A (PutA) enzyme found in Gram-negative bacteria. PutA is a large (>1000 residues) bifunctional enzyme that combines PRODH and GSALDH activities into one polypeptide chain. In addition, some PutAs function as a DNA-binding transcriptional repressor of proline utilization genes. This review describes several attributes of PutA that make it a remarkable flavoenzyme: (1) diversity of oligomeric state and quaternary structure; (2) substrate channeling and enzyme hysteresis; (3) DNA-binding activity and transcriptional repressor function; and (4) flavin redox dependent changes in subcellular location and function in response to proline (functional switching).

The shikimate dehydrogenase family: functional diversity within a conserved structural and mechanistic framework

Shikimate dehydrogenase (SDH) catalyzes the NADPH-dependent reduction of 3-deydroshikimate to shikimate, an essential reaction in the biosynthesis of the aromatic amino acids and a large number of other secondary metabolites in plants and microbes. The indispensible nature of this enzyme makes it a potential target for herbicides and antimicrobials. SDH is the archetypal member of a large protein family, which contains at least four additional functional classes with diverse metabolic roles. The different members of the SDH family share a highly similar three-dimensional structure and utilize a conserved catalytic mechanism, but exhibit distinct substrate preferences, making the family a particularly interesting system for studying modes of substrate recognition used by enzymes. Here, we review our current understanding of the biochemical and structural properties of each of the five previously identified SDH family functional classes.

Complex patterns of gene fission in the eukaryotic folate biosynthesis pathway

Shared derived genomic characters can be useful for polarizing phylogenetic relationships, for example, gene fusions have been used to identify deep-branching relationships in the eukaryotes. Here, we report the evolutionary analysis of a three-gene fusion of folB, folK, and folP, which encode enzymes that catalyze consecutive steps in de novo folate biosynthesis. The folK-folP fusion was found across the eukaryotes and a sparse collection of prokaryotes. This suggests an ancient derivation with a number of gene losses in the eukaryotes potentially as a consequence of adaptation to heterotrophic lifestyles. In contrast, the folB-folK-folP gene is specific to a mosaic collection of Amorphea taxa (a group encompassing: Amoebozoa, Apusomonadida, Breviatea, and Opisthokonta). Next, we investigated the stability of this character. We identified numerous gene losses and a total of nine gene fission events, either by break up of an open reading frame (four events identified) or loss of a component domain (five events identified). This indicates that this three gene fusion is highly labile. These data are consistent with a growing body of data indicating gene fission events occur at high relative rates. Accounting for these sources of homoplasy, our data suggest that the folB-folK-folP gene fusion was present in the last common ancestor of Amoebozoa and Opisthokonta but absent in the Metazoa including the human genome. Comparative genomic data of these genes provides an important resource for designing therapeutic strategies targeting the de novo folate biosynthesis pathway of a variety of eukaryotic pathogens such as Acanthamoeba castellanii.

Evolutionary origins of the eukaryotic shikimate pathway: gene fusions, horizontal gene transfer, and endosymbiotic replacements

Currently the shikimate pathway is reported as a metabolic feature of prokaryotes, ascomycete fungi, apicomplexans, and plants. The plant shikimate pathway enzymes have similarities to prokaryote homologues and are largely active in chloroplasts, suggesting ancestry from the plastid progenitor genome. Toxoplasma gondii, which also possesses an alga-derived plastid organelle, encodes a shikimate pathway with similarities to ascomycete genes, including a five-enzyme pentafunctional arom. These data suggests that the shikimate pathway and the pentafunctional arom either had an ancient origin in the eukaryotes or was conveyed by eukaryote-to-eukaryote horizontal gene transfer (HGT). We expand sampling and analyses of the shikimate pathway genes to include the oomycetes, ciliates, diatoms, basidiomycetes, zygomycetes, and the green and red algae. Sequencing of cDNA from Tetrahymena thermophila confirmed the presence of a pentafused arom, as in fungi and T. gondii. Phylogenies and taxon distribution suggest that the arom gene fusion event may be an ancient eukaryotic innovation. Conversely, the Plantae lineage (represented here by both Viridaeplantae and the red algae) acquired different prokaryotic genes for all seven steps of the shikimate pathway. Two of the phylogenies suggest a derivation of the Plantae genes from the cyanobacterial plastid progenitor genome, but if the full Plantae pathway was originally of cyanobacterial origin, then the five other shikimate pathway genes were obtained from a minimum of two other eubacterial genomes. Thus, the phylogenies demonstrate both separate HGTs and shared derived HGTs within the Plantae clade either by primary HGT transfer or secondarily via the plastid progenitor genome. The shared derived characters support the holophyly of the Plantae lineage and a single ancestral primary plastid endosymbiosis. Our analyses also pinpoints a minimum of 50 gene/domain loss events, demonstrating that loss and replacement events have been an important process in eukaryote genome evolution.

The molecular biology of multidomain proteins. Selected examples

The aim of this review is to give an overview of the contribution molecular biology can make to an understanding of the functions and interactions within multidomain proteins. The contemporary advantages ascribed to multidomain proteins include (a) the potential for metabolite channelling and the protection of unstable intermediates; (b) the potential for interactions between domains catalysing sequential steps in a metabolic pathway, thereby giving the potential for allosteric interactions; and (c) the facility to produce enzymic activities in a fixed stoichiometric ratio. The alleged advantages in (a) and (b) however apply equally well to multi-enzyme complexes; therefore, specific examples of these phenomena are examined in multidomain proteins to determine whether the proposed advantages are apparent. Some transcription-regulating proteins active in the control of metabolic pathways are composed of multiple domains and their control is exerted and modulated at the molecular level by protein-DNA, protein-protein and protein-metabolite interactions. These complex recognition events place strong constraints upon the proteins involved, requiring the recognition of and interaction with different classes of cellular metabolites and macromolecules. Specific examples of transcription-regulating proteins are examined to probe how their multidomain nature facilitates a general solution to the problem of multiple recognition events. A general unifying theme that emerges from these case studies is that a basic unitary design of modules provided by enzymes is exploited to produce multidomain proteins by a complex series of gene duplication and fusion events. Successful modules provided by enzymes are co-opted to new function by selection apparently acting upon duplicated copies of the genes encoding the enzymes. In multidomain transcription-regulating proteins, former enzyme modules can be recruited as molecular sensors that facilitate presumed allosteric interactions necessary for the molecular control of transcription.

Differential flux through the quinate and shikimate pathways. Implications for the channelling hypothesis

The qutC gene encoding dehydroshikimate dehydratase has been constitutively overexpressed in Aspergillus nidulans from a range of 1-30-fold over the normal wild-type level. This overexpression leads to impaired growth in minimal medium which can be alleviated by the addition of aromatic amino acids to the medium. Overexpression of the qutC gene in mutant strains lacking protocatechuic acid (PCA) oxygenase leads to the build up of PCA in the medium, which can be measured by a simple assay. Measuring the rate of production of PCA in strains overproducing dehydroshikimate dehydratase and correlating this with the level of overproduction and impaired ability to grow in minimal medium lacking aromatic amino acids leads to the conclusion that (a) the metabolites 3-dehydroquinate and dehydroshikimate leak from the AROM protein at a rate comparable with the extent of flux catalysed by the AROM protein, (b) the AROM protein has a low-level channelling function probably as a result of the close juxtaposition of five active sites and (c) this channelling function is only physiologically significant under non-optimal conditions of nutrient supply and oxygenation, when the organism is in situ in its natural environment.

In vivo overproduction of the pentafunctional arom polypeptide in Aspergillus nidulans affects metabolic flux in the quinate pathway

The shikimate pathway and the quinic acid utilisation (QUT) pathway of Aspergillus nidulans and other fungi share the two common metabolic intermediates, 3-dehydroquinic acid (DHQ) and dehydroshikimic acid (DHS), which are interconverted by two isoenzymes, catabolic 3-dehydroquinase, (cDHQase) and biosynthetic dehydroquinase, (bDHQase). bDHQase is one of five consecutive enzymatic activities associated with the pentafunctional arom protein encoded by the complex AROM locus, whereas cDHQase is encoded by the single-function QUTE gene, one of seven genes comprising the QUT gene cluster in A. nidulans, which is required for the catabolism of quinate to protocatechuate. We addressed the question of how much (if any) leakage there is of the two common substrates between the two pathways, by increasing the concentration of the arom protein in vivo by means of recombinant DNA technology. We demonstrated that constitutive overproduction of the arom protein by 12-fold in the presence of quinate inhibits germination of conidiospores, but showed that 12-fold quinate-inducible overproduction of arom protein does not have this effect. In addition we showed that a qutE mutant (lacking cDHQase) can grow with quinic acid as sole carbon source whtn the arom protein is overproduced fivefold. The data are most simply interpreted as simple competition for common substrates by the enzymes of the two pathways and demonstrate that any channelling function of the arom protein in vivo is relatively leaky.

The Saccharomyces cerevisiae ARO1 gene. An example of the co-ordinate regulation of five enzymes on a single biosynthetic pathway

The ARO1 gene of Saccharomyces cerevisiae encodes the arom multifunctional enzyme. Specific inhibitors of amino acid biosynthesis have been used to obtain evidence that expression of a cloned ARO1 gene is regulated in response to amino acid limitation. Northern blot analysis and sequence studies indicate that ARO1 is regulated by the well characterised S. cerevisiae 'general control' mechanism. This provides a very economical means of simultaneously tailoring the synthesis of five shikimate pathway enzymes to the needs of the cell.

The pentafunctional arom enzyme of Saccharomyces cerevisiae is a mosaic of monofunctional domains

The nucleotide sequence of the Saccharomyces cerevisiae ARO1 gene which encodes the arom multifunctional enzyme has been determined. The protein sequence deduced for the pentafunctional arom polypeptide is 1588 amino acids in length and has a calculated Mr of 174555. Functional regions within the polypeptide chain have been identified by comparison with the sequences of the five monofunctional Escherichia coli enzymes whose activities correspond with those of the arom multifunctional enzyme. The observed homologies demonstrate that the arom polypeptide is a mosaic of functional domains and are consistent with the hypothesis that the ARO1 gene evolved by the linking of ancestral E. coli-like genes.

Is ubiquinone diffusion rate-limiting for electron transfer?

The different possible dispositions of the electron transfer components in electron transfer chains are discussed: random distribution of complexes and ubiquinone with diffusion-controlled collisions of ubiquinone with the complexes, random distribution as above, but with ubiquinone diffusion not rate-limiting, diffusion and collision of protein complexes carrying bound ubiquinone, and solid-state assembly. Discrimination among these possibilities requires knowledge of the mobility of the electron transfer chain components. The collisional frequency of ubiquinone-10 with the fluorescent probe 12-(9-anthroyl)stearate, investigated by fluorescence quenching, is 2.3 X 10(9) M-1 sec-1 corresponding to a diffusion coefficient in the range of 10(-6) cm2/sec (Fato, R., Battino, M., Degli Esposti, M., Parenti Castelli, G., and Lenaz, G., Biochemistry, 25, 3378-3390, 1986); the long-range diffusion of a short-chain polar Q derivative measured by fluorescence photobleaching recovery (FRAP) (Gupte, S., Wu, E. S., Höchli, L., Höchli, M., Jacobson, K., Sowers, A. E., and Hackenbrock, C. R., Proc. Natl. Acad. Sci. USA 81, 2606-2610, 1984) is 3 X 10(-9) cm2/sec. The discrepancy between these results is carefully scrutinized, and is mainly ascribed to the differences in diffusion ranges measured by the two techniques; it is proposed that short-range diffusion, measured by fluorescence quenching, is more meaningful for electron transfer than long-range diffusion measured by FRAP, or microcollisions, which are not sensed by either method. Calculation of the distances traveled by random walk of ubiquinone in the membrane allows a large excess of collisions per turnover of the respiratory chain. Moreover, the second-order rate constants of NADH-ubiquinone reductase and ubiquinol-cytochrome c reductase are at least three orders of magnitude lower than the second-order collisional constant calculated from the diffusion of ubiquinone. The activation energies of either the above activities or integrated electron transfer (NADH-cytochrome c reductase) are well above that for diffusion (found to be ca. 1 kcal/mol). Cholesterol incorporation in liposomes, increasing bilayer viscosity, lowers the diffusion coefficients of ubiquinone but not ubiquinol-cytochrome c reductase or succinate-cytochrome c reductase activities. The decrease of activity by ubiquinone dilution in the membrane is explained by its concentration falling below the Km of the partner enzymes. It is calculated that ubiquinone diffusion is not rate-limiting, favoring a random model of the respiratory chain organization.(ABSTRACT TRUNCATED AT 400 WORDS)

The effect of feedback on pathway transient response

The effect of variation of the rate of input of material on the transient behaviour of metabolic pathways is examined. This reveals the existence of three transient times which make up the overall pathway transient. Two of these have been described previously and represent the times required for the accumulation of the free intermediate pool and the pool of enzyme-bound intermediate. They are state functions and as such are independent of the way in which the steady state was reached. The third is attributable to the variation in the rate of input of material to the pathway. It is dependent on three further factors. These are (a) the time required for the initial enzyme to reach its own steady state, (b) substrate depletion and (c) feedback. The description of the transient is: (Formula: see text) where V0 represents the rate of input and Vss represents the steady-state flux. The transient time associated with the transition between steady-states is shown to be a simple function of the transients for the establishment of each steady state from rest and may be expressed as: tau = tau b-Va/Vb . tau a where Va and Vb refer to the fluxes in the two steady states and tau a and tau b represent the transient times for the establishment of each of the steady-states from rest. The total pathway transient may now be completely defined as: (formula: see text) where summation over all intermediates, I, is implied. The significance of this to the analysis of pathway behaviour is discussed with more general examples of pathway transient analysis.

Rates of membrane-associated reactions: reduction of dimensionality revisited

The hypothesis that reactions associated with intracellular membranes enjoy a kinetic advantage from a reduced dimensionality for diffusion is inconsistent with available data on lateral diffusion rates, membrane-substrate affinities, and endogenous concentrations of enzymes and their aqueous substrates.

Gene organization and regulation in the qa (quinic acid) gene cluster of Neurospora crassa

Images

The 3-dehydroquinate synthase activity of the pentafunctional arom enzyme complex of Neurospora crassa is Zn2+-dependent

We have demonstrated the co-purification in constant ratio of all five activities of the pentafunctional arom enzyme complex from Neurospora crassa. Progressive inactivation of the 3-dehydroquinate synthase component of the purified enzyme complex by chelating agents was blocked by the substrate, 3-deoxy-D-arabino-heptulosonate 7-phosphate, but not by the cofactor NAD+. Full activity was restored at Zn2+ concentrations as low as 0.05 nM. Atomic absorption data indicated that the intact enzyme complex contained 1 atom per subunit of tightly bound zinc. The arom 3-dehydroquinate synthase had a calculated turnover number of 19s-1, this being within the narrow range of values obtained for the other four activities of the intact multifunctional enzyme. The Km for 3-deoxy-D-arabino-heptulosonate 7-phosphate was 1.4 microM in a phosphate-free buffer; inorganic phosphate was a competitive inhibitor with respect to 3-deoxy-D-arabino-heptulosonate 7-phosphate.

Molecular compartmentation by enzyme cluster formation. A view over current investigations

Current investigations in different fields of cellular metabolism focus on the phenomenon of molecular compartmentation as an essential part of metabolic control. This type of compartment without surrounding membranes arises from enzyme cluster formation in the cell. The organization of the enzymes ranges from very loose, non-covalent aggregations, sometimes only transiently associated - dependent on metabolic or developmental state of the cell - to the very fixed, even covalently linked structures. These organized multienzyme systems produce a chemical microheterogeneity concerning the metabolite concentrations in the cell. Molecular compartmentation is the description of this chemical microheterogeneity in a biological term.

Kinetics of coupled reactions catalyzed by aspartate aminotransferase and glutamate dehydrogenase

Liver mitochondrial aspartate aminotransferase and glutamate dehydrogenase catalyze following sequence of reactions: see formula in text. In the presence of a slight excess of dehydrogenase, the time course of NADPH oxidation resulting from the overall reaction goes through a lag phase and reaches a linear phase. The slopes of the linear part of this curve is a linear function of transaminase concentration. At high concentration (approximately or equal to 10 microM) of both enzymes the lag phase, as observed after rapid mixing of the two enzymes in a Durrum stopped-flow spectrophotometer, is shorter, than that predicted from the kinetic parameters determined for the separate reactions catalyzed by each enzyme.

Proteolytically induced changes in the molecular form of the carbamyl phosphate synthetase-uracil-aspartate transcarbamylase complex coded for by the URA2 locus in Saccharomyces cerevisiae

When a uracil-auxotrophic yeast strain is grown under uracil-limiting conditions, the aspartate transcarbamylase activity found in crude extracts shows a variation in sensitivity to feedback inhibition by uridine 5'-triphosphate. In this study we correlated this variation with changes in the molecular form of the carbamyl phosphate synthetase-uracil-aspartate transcarbamylase complex. Carbamyl phosphate synthetase-uracil (molecular weight, 240,000) and uridine 5'-triphosphate-insensitive aspartate transcarbamylase (molecular weight, 140,000) were present separately in extracts from cells collected in the early exponential phase; this was in contrast to the presence of a single high-molecular-weight form (molecular weight, about 900,000) bearing both activities in extracts from stationary-phase cells. The lack of sensitivity to uridine 5'-triphosphate by aspartate transcarbamylase was delayed by adding uridine 5'-triphosphate before cell disruption and was prevented completely by adding phenylmethylsulfonyl fluoride. Thus, this event was attributed to a transient serine protease activity detected only in early exponential-phase cell extracts. However, even in the presence of phenylmethylsulfonyl fluoride, a sucrose density gradient analysis in the absence of uridine 5'-triphosphate revealed a change in the aggregation state of the complex which might have occurred in vivo. None of these events was observed in extracts from cells that lacked protease B activity (strain HP232-2B).

Dependence of the catalytic activities on the aggregation and conformation states of uridine 5'-phosphate synthase

Uridine 5'-phosphate (UMP) synthase is a multifunctional protein that contains the last two enzyme activities for the de novo biosynthesis of UMP, orotate phosphoribosyltransferase (EC 2.4.2.10) and orotidine-5'-phosphate (OMP) decarboxylase (EC 4.1.1.23). The native enzyme from mouse Ehrlich ascites cells exists in at least three distinct aggregation/conformation states as measured by sedimentation in sucrose gradients: a 3.6S monomer, a 5.1S dimer, and a conformationally altered 5.6S dimer. It has previously been reported that a variety of ligands (of which the most effective is OMP) mediate the conversion of the 3.6S monomer to the two types of dimers. Initial velocity studies with the enzyme in the different native states show that all three forms of UMP synthase have phosphoribosyltransferase activity but that the OMP decarboxylase is either uniquely or at least predominantly associated with the 5.6S form. Activation of this enzyme activity by the substrate appears to be the result of both a dimerization and a conformation step.

Multienzyme complex for metabolic channeling in mammalian DNA replication

In the DNA-synthesizing phase (S phase) of CHEF/18 Chinese hamster embryo fibroblast cells, six enzymes associated with DNA metabolism, including DNA polymerase (deoxynucleoside triphosphate:DNA deoxynucleotidyl-transferase, EC 2.7.7.7), were largely localized in the nuclear region (karyoplasts). By contrast, in quiescent and G1 phase cells these enzymatic activites were mainly absent from the nucleus and were recovered in the cytoplasmic portion (cytoplasts). These nuclear (but not cytoplasmic) enzymatic activities cosedimented rapidly on sucrose density gradients. Further, the rapidly sedimenting enzyme activities were unique to cells in S phase. An organized supramolecular structure that allows channeling of metabolites into DNA was demonstrated by kinetics of nucleotide incorporation. "Permeabilized" cells selectively channeled incorporation of ribonucleoside diphosphates into DNA in preference to deoxyribonucleoside triphosphates. Deoxyribonucleoside triphosphate incorporation occurred when ribonucleoside-diphosphate reductase (2'-deoxyribonucleoside-diphosphate: oxidized-thioredoxin 2'-oxidoreductase, EC 1.17.4.1) activity was abolished by hydroxyurea. Our interpretation is that during DNA replication, the nucleus contains a complex of DNA precursor-synthesizing enzymes juxtaposed with the "replication apparatus" comprising DNA polymerase, other enzymes, and structural proteins. Functional integrity of this structure is impaired when one of its essential components is inactivated. We propose the name "replitase" for this multienzyme complex for DNA replication and suggest that it incorporates precursors rapidly and efficiently. Possibly its assembly signals the initiation of the S phase of the cell cycle.

Biochemical genetic analysis of pyrimidine biosynthesis in mammalian cells: III. Association of carbamyl phosphate synthetase, aspartate transcarbamylase, and dihydroorotase in mutants of cultured Chinese hamster cells

Carbamyl phosphate synthetase (EC 2.7.2.9), aspartate transcarbamylase (EC 2.1.3.2), and dihydroorotase (EC 3.5.2.3), the first three enzymes in de novo pyrimidine synthesis in Chinese hamster ovary cell strain Kl (CHO-Kl), cose diment through a glycerol gradient. When an extract from Urd- A, a pyrimidine-requiring auxotroph reduced in all three activities, is run on a glycerol gradient, the enzyme activities appear in two peaks higher in the gradient, a peak of aspartate transcarbamylase separated from a peak of carbamyl phosphate synthetase and dihydroorotase. Revertants of Urd- A have increased activity of all three enzymes and give glycerol gradient patterns similar to either CHO-Kl or Urd- A. The gradient pattern for Urd- A and some of its revertants can be mimicked by treating the CHO-Kl cell extract with trypsin. Hybrids made between a CHO-Kl purine-requiring auxotroph (Ade- C) and a Urd- A revertant gave a glycerol gradient pattern which is a composite of the CHO-Kl and revertant patterns. A model is presented for the structure of this multifunctional protein.

Enzyme associations in T4 phage DNA precursor synthesis

A DIRECT APPROACH IS DESCRIBED TO THE QUESTION: Are enzymes of DNA precursor synthesis organized into a supramolecular structure? This approach involved sedimentation analysis of several T4 phage-coded early enzyme activities in crude lysates of infected Escherichia coli. One-third to one-half of several activities tested-dCMP hydroxymethylase, dTMP synthetase, deoxynucleoside 5'-monophosphate kinase, deoxyuridine triphosphatase, and probably dCMP deaminase, but not dihydrofolate reductase or DNA polymerase-sedimented much more rapidly than expected from molecular weight. About 5% of the host cell nucleoside diphosphate kinase, known to participate in T4 DNA precursor synthesis, cosedimented with these activities. To show that this rapidly sedimenting material represents an organized enzyme complex rather than a nonspecific aggregate, we studied the kinetics of formation of dTTP with dUMP as the initial substrate. This three-step reaction sequence reached its maximal rate within a few seconds when catalyzed by enzymes in the aggregate, whereas an equivalent mixture of uncomplexed enzymes required nearly 20 min before dTTP synthesis reached its maximal rate. The effect of aggregation is evidently to decrease the volume into which intermediates are free to diffuse. Because there is reason to believe that intracellular concentration gradients of DNA precursors exist, the properties of this enzyme aggregate in vitro may help to explain how such gradients are maintained.

The subunit structure of the arom multienzyme complex of Neurospora crassa. A possible pentafunctional polypeptide chain

A new procedure for the purification of the arom multienzyme complex from Neurospora crassa is presented. Important factors are the inactivation of proteinases by phenylmethanesulphonyl fluoride and the use of cellulose phosphate as an affinity adsorbent. A homogeneous enzyme, with a specific shikimate dehydrogenase activity of 70 units/mg of protein, is obtained in 25% yield. Polyacrylamide-gel electrophoresis in the presence of sodium dodecyl sulphate, combined with cross-linking studies using dimethyl suberimidate, suggest that the complex is composed of two subunits of molecular weight 165000. Glycerol-density-gradient centrifugation indicates a molecular weight for the intact complex of about 270000. Evidence for the effects of proteolysis, both during the preparation and on storage of the purified complex, is presented, and previous reports in the literature of the occurrence of multiple subunits are discussed in this light.