Formic Hydrogenlyase and the Photoassimilation of Formate by a Strain of Rhodopseudomonas palustris

A photosynthetic bacterium isolated by enrichment on media containing formate as major source of cell carbon was identified as a strain of Rhodopseudomonas palustris. It grew on a wide range of simple organic compounds including alcohols, fatty acids, and hydroxyacids, on a chemically defined medium with biotin and p-aminobenzoic acid as essential growth factors. The organism grew on formate or photoautotrophically with molecular hydrogen or thiosulfate only in the presence of yeast extract. Ability to photoassimilate formate could be shown only in organisms grown in the presence of formate. The organism contained an inducible formic hydrogenlyase consisting of a soluble formic dehydrogenase, a particulate hydrogenase, and one or more intermediate, but as yet unidentified, electron carriers. The formic hydrogenlyase could be reconstituted from a particulate hydrogenase and a partially purified soluble formic dehydrogenase. Some properties of the formic dehydrogenase and hydrogenase have been compared with that of the formic hydrogenlyase system.     

A physiological study of formate dehydrogenase,formate oxidase and hydrogenlyase fromEscherichia coli K-12

Escherichia coli was grown under various culture conditions. Variations in the levels of formate dehydrogenase which reacts with methylene blue (MB) or phenazine methosulfate (PMS) (N enzyme), formate dehydrogenase which reacts with benzyl viologen (BV) (H enzyme), formate oxidase and hydrogenlyase were analyzed. It was observed that formate dehydrogenase N and formate oxidase were induced by nitrate and repressed by oxygen. Synthesis of formate dehydrogenase H and hydrogenlyase was induced by formate and repressed by nitrate and oxygen. Selenite was required for the biosynthesis of formate dehydrogenase H and hydrogenlyase. Activity of both formate oxidase and hydrogenlyase was inhibited by azide and KCN but not by N-heptyl hydroxyquinoline-N-oxide (HOQNO); on the other hand, formate oxidase was extremely sensitive to HOQNO. Data were obtained which suggest that cytochromes are not involved in hydrogen formation from formate. Part of this work was carried out when the senior author was visiting Research Biologist in the Laboratory of Dr. J. A. de Mosss at the University of California, San Diego. Thanks are given to Dr. De Moss for his hospitality and advise and to Dr. Warren Butler of the University of California, San Diego for making available his spectrophotometer to carry out cytochrome analyses. Most of this work was sustained by a grant from the Research Corporation, Brown Hazen Fund and the financial help of the C.O.F.A.A. from the Instituto Politécnico Nacional.     

Reductive Pentose Cycle and Formate Assimilation in Rhodopseudomonas palustris

Rhodopseudomonas palustris assimilated formate autotrophically as carbon dioxide and hydrogen arising from the activity of the formic hydrogenlyase system. Kinetic analyses of cell suspensions pulse-labeled with (14)C-formate or (14)C-bicarbonate showed similar distributions of incorporated radioactivity. In both cases phosphate esters were the first assimilation products. Ribulose diphosphate carboxylase, phosphoribose isomerase, and phosphoribulokinase, characteristic enzymes of the reductive pentose cycle, were present in extracts of cells grown on formate.     

Thiosulphate improves yield of hydrogen production from glucose by the immobilized formate hydrogenlyase system of Escherichia coli.

Escherichia coli cells with formate hydrogenlyase activity that were immobilized in agar beads produced hydrogen from glucose at the approximate yield of 0.6 mole of hydrogen per mole. Succinate or thiosulphate added to glucose increased hydrogen production two-fold. Thiosulphate did not increase the rate of hydrogen production but reduced the consumption of glucose for the same amount of hydrogen produced compared to control. Oxaloacetate and traces of pyruvate instead of succinate accumulated at the end when thiosulphate was present.     

Genetic and physiological characterization of new Escherichia coli mutants impaired in hydrogenase activity

The Mu dl (ApR lac) bacteriophage was used to generate mutants of Escherichia coli which were defective in formate hydrogenlyase. Three mutants were chosen for further analysis: they lacked hydrogenase (hydrogen: benzyl viologen oxidoreductase) activity, but produced normal levels of fumarate reductase activity and two- to three-fold reduced levels of benzyl viologen (BV)-dependent formate dehydrogenase activity. Two of them (hydC) were shown to contain about 4-fold reduced amounts of formate hydrogenlyase and fumarate-dependent H2 uptake activities. The third one (hydD) was totally devoid of both activities. Their insertion sites were located at 77 min on the E. coli map. Subdivision of these mutants into two classes was subsequently based on the restoration capacity of hydrogenase activity with high concentration of nickel in the growth media. Addition of 500 microM NiCl2 led to a complete recovery of hydrogenase activity, and to the concomitant restoration of normal BV-linked formate dehydrogenase, formate hydrogenlyase and fumarate-dependent H2 uptake activities in the hydC mutants. The hydD mutant was insensitive to the effect of nickel. Expression of the lac operon in hydC and hydD mutants was induced by anaerobiosis. It was not increased by the addition of formate under anaerobic conditions. The presence of nitrate resulted in slightly reduced beta-galactosidase activities in the hydC mutants, whereas those found in the hydD mutant reached only one third of the level obtained in its absence. Fumarate had no effect on both classes. Moreover, in contrast to the hydD locus, the hydC::Mu dl fusions were found to be dependent upon the positive control exerted by the nirR gene product and were totally repressed by an excess of nickel. In addition, the low levels of overall hydrogenase-dependent activities found in a nirR strain were also relieved by the presence of nickel. Our results strongly suggest that the pleiotropic regulatory gene nirR is essential for the expression of a gene (hydC) involved in either transport or processing of nickel in the cell, whose alteration leads to a loss of hydrogenase activity.     

Involvement of outer membrane proteins in enterochelin-mediated iron uptake in Escherichia coli.

Escherichia coli K-12 grown in iron-deficient media contained a large amount of outer membrane proteins O-2a, O-2b, and O-3, while cells grown in iron-supplemented media contained far smaller amounts of these proteins. The iron uptake by the iron-deficient cells was significantly stimulated in the presence of enterochelin, while that by the iron-rich cells was not. The outer membrane isolated from cells grown in the iron-deficient media showed enterochelin-stimulated binding of iron, while the outer membrane from iron-rich cells and cytoplasmic membranes from both types of cells did not show such binding activity. The amount of iron bound by the outer membrane was almost equivalent to the amount of O-2a, O2b, or O-3, irrespective of the amount of these proteins in the outer membrane, which is controlled by the amount of iron in the medium. Small particles rich in these proteins were prepared from cells by EDTA extraction. The particles were active in enterochelin-mediated iron binding and the amount of iron bound was equivalent to the amount of each of these proteins in the particles. Although the outer membrane of E. coli B was as active in iron binding as that of E. coli K-12, it did not possess an appreciable amount of O-2a. Gel electrophoretic analysis revealed that 9-2b and 9-3 were identical with the proteins missing mutants feuB and feuA, respectively.     

Effects of iron deficiency on heme biosynthesis in Rhizobium japonicum.

The effects of iron deficiency on heme biosynthesis in Rhizobium japonicum were examined. Iron-deficient cells had a decreased maximum cell yield and a decreased cytochrome content and excreted protoporphyrin into the growth medium. The activities of the first two enzymes of heme biosynthesis, delta-aminolevulinic acid synthase (EC 2.3.1.37) and delta-aminolevulinic acid dehydrase (EC 4.2.1.24), were diminished in iron-deficient cells, but were returned to normal levels upon addition of iron to the cultures. The addition of iron salts, iron chelators, hemin, or protoporphyrin to cell-free extracts did not affect the activity of these enzymes. The addition of levulinic acid to iron-deficient cultures blocked protoporphyrin excretion and also resulted in high delta-aminolevulinic acid synthase and delta-aminolevulinic acid dehydrase activities. These results suggest the possibility that rhizobial heme biosynthesis in the legume root nodule may be affected by the release of iron from the host plant to the bacteroids.     

Synthesis and lysis of formate by immobilized cells of Escherichia coli

Formate hydrogenlyase (FHL) activity was induced in a strain of Escherichia coli S13 during anaerobic growth in yeast extract-tryptone medium containing 100 mM formate. The cells obtained at the optimum growth phase were immobilized in 2.5% (w/v) agar gel when 50-60% of the whole cell FHL activity was retained. The immobilized FHL system had good storage stability and recycling efficiency. In the lysis of formate, an increase of formate concentration to 1.18M increased QH(2) (initial) value of the immobilized cell, and subsequently cells, hydrogen evolution, in general, ceased after 6 to 8 of incubation, resulting in incomplete lysis of formate. Presence of small amount of glucose (28 mM) was more or less quantitatively lysed with concomitant disappearence of glucose from the medium. Synthesis of formate from hydrogen and bicarbonate solution by the immobilized cells was also characterized. Presence of glucose (10 mM) in 50 mM bicarbonate solution stimulated formate synthesis by immobilized cells. The pH optimum range, K(m), and specific activity of the immobilized cells for the lysis of formate were 6.8-7.2 0.4M, and 66 mL/g cell-h, respectively. The cells could fix hydrogen to the extent of 24.4% (w/w) of its own wet cell mass in a 72-h reaction cycle. Potentiality of the immobilized FHL system for biotechnological exploitation was discussed.     

Nitrate reductase complex of Escherichia coli K-12: participation of specific formate dehydrogenase and cytochrome b1 components in nitrate reduction

The participation of distinct formate dehydrogenases and cytochrome components in nitrate reduction by Escherichia coli was studied. The formate dehydrogenase activity present in extracts prepared from nitrate-induced cells of strain HfrH was active with various electron acceptors, including methylene blue, phenazine methosulfate, and benzyl viologen. Certain mutants which are unable to reduce nitrate had low or undetectable levels of formate dehydrogenase activity assayed with methylene blue or phenazine methosulfate as electron acceptor. Of nine such mutants, five produced gas when grown anaerobically without nitrate and possessed a benzyl viologen-linked formate dehydrogenase activity, suggesting that distinct formate dehydrogenases participate in the nitrate reductase and formic hydrogenlyase systems. The other four mutants formed little gas when grown anaerobically in the absence of nitrate and lacked the benzyl viologen-linked formate dehydrogenase as well as the methylene blue or phenazine methosulfate-linked activity. The cytochrome b(1) present in nitrate-induced cells was distinguished by its spectral properties and its genetic control from the major cytochrome b(1) components of aerobic cells and of cells grown anaerobically in the absence of nitrate. The nitrate-specific cytochrome b(1) was completely and rapidly reduced by 1 mm formate but was not reduced by 1 mm reduced nicotinamide adenine dinucleotide; ascorbate reduced only part of the cytochrome b(1) which was reduced by formate. When nitrate was added, the formate-reduced cytochrome b(1) was oxidized with biphasic kinetics, but the ascorbate-reduced cytochrome b(1) was oxidized with monophasic kinetics. The inhibitory effects of n-heptyl hydroxyquinoline-N-oxide on the oxidation of cytochrome b(1) by nitrate provided evidence that the nitrate-specific cytochrome is composed of two components which have different redox potentials but identical spectral properties. We conclude from these studies that nitrate reduction in E. coli is mediated by the sequential operation of a specific formate dehydrogenase, two specific cytochrome b(1) components, and nitrate reductase.     

On culturing Escherichia coli on a mineral salts medium during anaerobic conditions

The substrate and products of the hydrogenlyase complex, formic acid, carbon dioxide, and molecular hydrogen, are co-operatively implicated in maintaining growth of E. coli under anaerobic conditions. Growth is observed in the presence of a combination of carbon dioxide + molecular hydrogen, or carbon dioxide + formic acid in the medium. The study shows that it is possible to culture E. coli under anaerobic conditions while sparging with nitrogen, without supplementing exogenous carbon dioxide, formic acid or molecular hydrogen. This condition occurs when the strain is allowed an appropriate induction period and is present at a sufficiently high cell density, since the cell density affects the rate of e.g. CO₂ production. In a system sparged with nitrogen gas, the removal of CO₂ due to this sparging must be balanced with a cell density dependent production rate of CO₂. It is concluded that the formic hydrogenlyase complex should be considered as an integral part of the general maintenance of the anabolism of E. coli during anaerobic conditions on a mineral salts medium, as well as being a net producer of end products in E. coli metabolism.This work was supported by the Swedish National Board for Industrial and Technical Development. A. Askendahl is acknowledged for valuable assistance in the preparation of figures and P. Warkentin for language editing.     

Nutritional and biochemical characterization of Methanospirillum hungatii.

To ascertain its physiological similarity to other methanogenic bacteria, Methanospirillum hungatii, the type species of the genus, was characterized nutritionally and biochemically. Good growth occurred in a medium consisting of mineral salts, cysteine sulfide reducing buffer, and an H2-CO2 (80:20) atmosphere. Addition of amino acids and B vitamins stimulated growth. Cell-free extracts contained methylcobalamin-coenzyme M methyltransferase, methylreductase, and formate hydrogenlyase. Cells contained coenzyme M and coenzyme F420. Coenzyme F420 was required for formate hydrogenlyase activity. Coenzyme F420 purified from M. hungatii had identical properties to that purified from species of Methanobacterium. The physiological basis of the family Methanobacteriaceae is strengthened by these findings.     

Involvement of the AtoS-AtoC signal transduction system in poly-(R)-3-hydroxybutyrate biosynthesis in Escherichia coli

The AtoS-AtoC signal transduction system in E. coli, which induces the atoDAEB operon for the growth of E. coli in short-chain fatty acids, can positively modulate the levels of poly-(R)-3-hydroxybutyrate (cPHB) biosynthesis, a biopolymer with many physiological roles in E. coli. Increased amounts of cPHB were synthesized in E. coli upon exposure of the cells to acetoacetate, the inducer of the AtoS-AtoC two-component system. While E. coli that overproduce both components of the signal transduction system synthesize higher quantities of cPHB (1.5-4.5 fold), those that overproduce either AtoS or AtoC alone do not display such a phenotype. Lack of enhanced cPHB production was also observed in cells overexpressing AtoS and phosphorylation-impaired AtoC mutants. The results were not affected by the nature of the carbon source used, i.e., glucose, acetate or acetoacetate. An E. coli strain with a deletion in the atoS-atoC locus (delta atoSC) synthesized lower amounts of cPHB compared to wild-type cells. When the delta atoSC strain was transformed with a plasmid carrying a 6.4-kb fragment encoding the AtoS-AtoC system, cPHB biosynthesis was restored to the level of the atoSC+ cells. Introduction of a multicopy plasmid carrying a functional atoDAEB operon, but not one with a promoterless operon, resulted in increased cPHB synthesis only in atoSC+ cells in the presence of acetoacetate. These results indicate that the presence of both a functional AtoS-AtoC two-component signal transduction system and a functional atoDAEB operon is critical for the enhanced cPHB biosynthesis in E. coli.     

Nutritional and biochemical characterization of Methanospirillum hungatii.

To ascertain its physiological similarity to other methanogenic bacteria, Methanospirillum hungatii, the type species of the genus, was characterized nutritionally and biochemically. Good growth occurred in a medium consisting of mineral salts, cysteine sulfide reducing buffer, and an H2-CO2 (80:20) atmosphere. Addition of amino acids and B vitamins stimulated growth. Cell-free extracts contained methylcobalamin-coenzyme M methyltransferase, methylreductase, and formate hydrogenlyase. Cells contained coenzyme M and coenzyme F420. Coenzyme F420 was required for formate hydrogenlyase activity. Coenzyme F420 purified from M. hungatii had identical properties to that purified from species of Methanobacterium. The physiological basis of the family Methanobacteriaceae is strengthened by these findings.     

Identification of a new gene, molR, essential for utilization of molybdate by Escherichia coli.

A mutation in a new gene, molR, prevented the synthesis in Escherichia coli of molybdoenzymes, including the two formate dehydrogenase isoenzymes, nitrate reductase and trimethylamine-N-oxide reductase. This phenotype was suppressed by supplementing the media with molybdate. Thus, the molR mutant was phenotypically similar to previously described chlD mutants, thought to be defective in molybdate transport. The molR gene is located at 65.3 min in the E. coli chromosome, in contrast to the chlD gene, which maps at 17 min and thus can be readily distinguished. The molR gene is also cotransducible with a hitherto unidentified gene essential for the production of 2-oxoglutarate from isocitrate, designated icdB (located at 66 min). The molR mutant strain SE1100 also failed to produce the hydrogenase component of formate hydrogenlyase (HYD3) in molybdate-unsupplemented media. The amount of molybdate required by strain SE1100 for the production of parental levels of formate hydrogenlyase activity was dependent on the growth medium. In Luria-Bertani medium, this value was about 100 microM, and in glucose-minimal medium, 1.0 microM was sufficient. In low-sulfur medium, this value decreased to about 50 nM. The addition of sulfate or selenite increased the amount of molybdate needed for the production of formate hydrogenlyase activity. These data suggest that in the absence of the high-affinity molybdate transport system, E. coli utilizes sulfate and selenite transport systems for transporting molybdate, preferring sulfate transport over the selenite transport system.     

Purification of the membrane-bound hydrogenase of Escherichia coli.

The membrane-bound hydrogenase (EC class 1.12) of aerobically grown Escherichia coli cells was solubilized by treatment with deoxycholate and pancreatin. The enzyme was further purified to electrophoretic homogeneity by chromoatographic methods, including hydrophobic-interaction chromatography, with a yield of 10% as judged by activity and an overall purification of 2140-fold. The hydrogenase was a dimer of identical subunits with a mol.wt. of 113,000 and contained 12 iron and 12 acid-labile sulphur atoms per molecule. The epsilon 400 was 49,000M-1 . cm-1. The hydrogenase catalysed both H2 evolution and H2 uptake with a variety of artificial electron carriers, but would not interact with flavodoxin, ferredoxin or nicotinamide and flavin nucleotides. We were unable to identify any physiological electron carrier for the hydrogenase. With Methyl Viologen as the electron carrier, the pH optimum for H2 evolution and H2 uptake was 6.5 and 8.5 respectively. The enzyme was stable for long periods at neutral pH, low temperatures and under anaerobic conditions. The half-life of the hydrogenase under air at room temperature was about 12 h, but it could be stabilized by Methyl Viologen and Benzyl Viologen, both of which are electron carriers for the enzyme, and by bovine serum albumin. The hydrogenase was strongly inhibited by carbon monoxide (Ki = 1870Pa), heavy-metal salts and high concentrations of buffers, but was resistant to inhibition by thiol-blocking and metal-complexing reagents. These aerobically grown E. coli cells lacked formate hydrogenlyase activity and cytochrome c552.     

Formic Hydrogenlyase Induction as a Basis for the Eijkman Fecal Coliform Concept

Many Enterobacter species recovered from a relatively unpolluted freshwater stream had the capacity to produce gas from glucose, lactose, and formate at 44.5 C. Composition of the evolved gas was shown to be CO₂ and H₂, which suggested that the enzyme system formic hydrogenlyase functions as the basis for the Eijkman fecal coliform concept. It is postulated that there are at least two different and distinct biochemical types of Enterobacter in the natural aquatic environment: one that lacks an active formic hydrogenlyase at 44.5 C and is associated with the intestinal tract of man and certain animals, and another which possesses the active enzyme at the elevated temperature and is found in soil or on vegetation free from fecal pollution.     

Inhibition of deoxycholate-induced apoptosis in iron-depleted HCT-116 cells

The bile acid, deoxycholate (DOC), can induce apoptosis in cells containing adequate amounts of all key nutrients, but it is unknown whether DOC-induced apoptosis can occur in cells lacking a single key nutrient. The aim of this study was to determine if DOC is able to induce apoptosis in HCT-116 colon epithelial cells depleted of iron. The cells were made iron-deficient by pre-treating them with the iron chelator, deferoxamine (DFO), before subsequent exposure to DOC. Mitochondrial dysfunction was detected in control cells exposed to DOC, but not in iron-depleted cells exposed to DOC. Moreover, characteristic features of apoptosis, namely, membrane blebbing, formation of apoptotic bodies, cytochrome c release into cytosol, generation of the activated form of caspase-3, chromatin condensation and fragmentation, and also plasma membrane phospholipid translocation, were all induced by DOC in control cells but not in iron-depleted cells. Treating DFO-pretreated cells with ferrous sulfate to replenish cellular iron restored the ability of DOC to induce apoptosis. In relating these findings to oxidative stress, it was found that DOC also induced the formation of reactive oxygen species and caused DNA damage in control cells, but not in iron-depleted cells. Collectively, the results suggest that in order for HCT-116 cells to undergo apoptosis when exposed to DOC, adequate amounts of intracellular iron must be present.     

Cloning of hydrogenase genes and fine structure analysis of an operon essential for H2 metabolism in Escherichia coli.

Escherichia coli has two unlinked genes that code for hydrogenase synthesis and activity. The DNA fragments containing the two genes (hydA and hydB) were cloned into a plasmid vector, pBR322. The plasmids containing the hyd genes (pSE-290 and pSE-111 carrying the hydA and hydB genes, respectively) were used to genetically map a total of 51 mutant strains with defects in hydrogenase activity. A total of 37 mutants carried a mutation in the hydB gene, whereas the remaining 14 hyd were hydA. This complementation analysis also established the presence of two new genes, so far unidentified, one coding for formate dehydrogenase-2 (fdv) and another producing an electron transport protein (fhl) coupling formate dehydrogenase-2 to hydrogenase. Three of the four genes, hydB, fhl, and fdv, may constitute a single operon, and all three genes are carried by a 5.6-kilobase-pair chromosomal DNA insert in plasmid pSE-128. Plasmids carrying a part of this 5.6-kilobase-pair DNA (pSE-130) or fragments derived from this DNA in different orientations (pSE-126 and pSE-129) inhibited the production of active formate hydrogenlyase. This inhibition occurred even in a prototrophic E. coli, strain K-10, but only during an early induction period. These results, based on complementation analysis with cloned DNA fragments, show that both hydA and hydB genes are essential for the production of active hydrogenase. For the expression of active formate hydrogenlyase, two other gene products, fhl and fdv are also needed. All four genes map between 58 and 59 min in the E. coli chromosome.