Ribose biosynthesis in methanogenic bacteria

NMR spectroscopy was used to determine the labeling patterns of the ribose moieties of ribonucleosides purified from Methanospirillum hungatei, Methanococcus voltae, Methanobrevibacter smithii, Methanosphaera stadtmanae, Methanosarcina barkeri and Methanobacterium bryantii labeled with ¹³C-precursors. In most methanogens tested ribose was labeled in a manner consistent with the operation of the oxidative branch of the pentose phosphate pathway. In contrast, transaldolase and transketolase reactions typical of a partial nonoxidative pentose phosphate pathway are hypothesized to explain the different labeling patterns and enrichments of carbon atoms observed in the ribose moiety of Methanococcus voltae. The source of erythrose 4-phosphate needed for the transaldolase reaction proposed in Methanococcus voltae, and for biosynthesis of aromatic amino acids in methanogenic bacteria in general, was assessed. Phenylalanine carbon atom C-7 was labeled by [1-¹³C]pyruvate in Methanospirillum hungatei, Methanococcus voltae, and Methanococcus jannaschii, the only methanogens which incorporated sufficient label from pyruvate for testing. Reductive carboxylation of a triose precursor (derived from pyruvate) to synthesize erythrose 4-phosphate is consistent with the labeling patterns observed in phenylalanine and ribose.Abbreviation TCA Tricarboxylic acid Issued as NRCC Publication No. 37382     

Activities of formylmethanofuran dehydrogenase,methylenetetrahydromethanopterin dehydrogenase,methylenetetrahydromethanopterin reductase,and heterodisulfide reductase in methanogenic bacteria

The activities of formylmethanofuran dehydrogenase, methylenetetrahydromethanopterin dehydrogenase, methylenetetrahydromethanopterin reductase, and heterodisulfide reductase were tested in cell extracts of 10 different methanogenic bacteria grown on H₂/CO₂ or on other methanogenic substrates. The four activities were found in all the organisms investigated: Methanobacterium thermoautotrophicum,M. wolfei, Methanobrevibacter arboriphilus, Methanosphaera stadtmanae, Methanosarcina barkeri (strains Fusaro and MS), Methanothrix soehngenii, Methanospirillum hungatei, Methanogenium organophilum, and Methanococcus voltae. Cell extracts of H₂/CO₂ grown M. barkeri and of methanol grown M. barkeri showed the same specific activities suggesting that the four enzymes are of equal importance in CO₂ reduction to methane and in methanol disproportionation to CO₂ and CH₄. In contrast, cell extracts of acetate grown M. barkeri exhibited much lower activities of formylmethanofuran dehydrogenase and methylenetetrahydromethanopterin dehydrogenase suggesting that these two enzymes are not involved in methanogenesis from acetate. In M. stadtmanae, which grows on H₂ and methanol, only heterodisulfide reductase was detected in activities sufficient to account for the in vivo methane formation rate. This finding is consistent with the view that the three other oxidoreductases are not required for methanol reduction to methane with H₂.     

Purine and pyrimidine biosynthesis in methanogenic bacteria

Following long-term labeling with [1-¹³C]acetate, [2-¹³C]acetate, ¹³CO₂, H¹³COOH, or ¹³CH₃OH, NMR spectroscopy was used to determine the labeling patterns of the purified ribonucleosides of Methanospirillum hungatei, Methanococcus voltae, Methanobrevibacter smithii, Methanosphaera stadtmanae, Methanosarcina barkeri and Methanobacterium bryantii. Major differences were observed among the methanogens studied, specifically at carbon positions 2 and 8 of the purines, positions at which one-carbon carriers are involved during synthesis. In Methanospirillum hungatei and Methanosarcina barkeri, the labcl at both positions came from carbon atom C-2 of acetate, as predicted from known eubacterial pathways, whereas in Methanococcus voltae and Methanobacterium bryantii both originated from CO₂. In Methanosphaera stadtmanae grown in the presence of formate, the C-2 of purines originated exclusively from formate and the C-8 was labeled by the C-2 of acetate. When grown in media devoid of formate, the C-2 of the purine ring originated mainly from the C-2 of acetate and in part from CH₃OH. In Methanobrevibacter smithii grown in the presence of formate, C-2 and C-8 of purines were derived from CO₂ and/or formate. The labeling patterns obtained for pyrimidines are consistent with the biosynthetic pathways common to eubacteria and eucaryotes.Abbreviations CODH Carbon monoxide dehydrogenase - FH₄ tetrahydrofolate - H₄MPT tetrahydromethanopterin Issued as NRCC Publication No. 37383     

Tetrahydrofolate-specific enzymes in <Emphasis Type="Italic">Methanosarcina barkeri</Emphasis> and growth dependence of this methanogenic archaeon on folic acid or <Emphasis Type="Italic">p</Emphasis>-aminobenzoic acid

Methanogenic archaea are generally thought to use tetrahydromethanopterin or tetrahydrosarcinapterin (H₄SPT) rather than tetrahydrofolate (H₄F) as a pterin C₁ carrier. However, the genome sequence of Methanosarcina species recently revealed a cluster of genes, purN, folD, glyA and metF, that are predicted to encode for H₄F-specific enzymes. We show here for folD and glyA from M. barkeri that this prediction is correct: FolD (bifunctional N⁵,N¹⁰-methylene-H₄F dehydrogenase/N⁵,N¹⁰-methenyl-H₄F cyclohydrolase) and GlyA (serine:H₄F hydroxymethyltransferase) were heterologously overproduced in Escherichia coli, purified and found to be specific for methylene-H₄F and H₄F, respectively (apparent K_m below 5 M). Western blot analyses and enzyme activity measurements revealed that both enzymes were synthesized in M. barkeri. The results thus indicate that M. barkeri should contain H₄F, which was supported by the finding that growth of M. barkeri was dependent on folic acid and that the vitamin could be substituted by p-aminobenzoic acid, a biosynthetic precursor of H₄F. From the p-aminobenzoic acid requirement, an intracellular H₄F concentration of approximately 5 M was estimated. Evidence is presented that the p-aminobenzoic acid taken up by the growing cells was not required for the biosynthesis of H₄SPT, which was found to be present in the cells at a concentration above 3 mM. The presence of both H₄SPT and H₄F in M. barkeri is in agreement with earlier isotope labeling studies indicating that there are two separate C₁ pools in these methanogens.     

Investigation of mercaptans,organic sulfides,and inorganic sulfur compounds as sulfur sources for the growth of methanogenic bacteria

A variety of compounds were investigated for use as sulfur sources for the growth of methanogenic bacteria.Methanococcus (Mc.) deltae, Mc. maripaludis, Methanobacterium (Mb.) speciesGC-2B, GC-3B, andMMY, Methanobrevibacter (Mbr.) ruminantium, andMethanosarcina (Ms.) barkeri strain 227 grew well with sulfide, S^o, thiosulfate, or cysteine as sole sulfur source.Mbr. ruminatium was able to grow on SO ₄ ⁼ or SO ₃ ⁼ , andMs. barkeri strain 227 was able to grow on SO ₃ ⁼ , but not on SO ₄ ⁼ as a sole sulfur source.Mc. jannaschii grew with sulfide, S^o, thiosulfate or SO ₃ ⁼ , but not on cysteine or SO ₄ ⁼ as sole surface source.Mc. thermolithotrophicus, Mc. jannaschii, Mc. deltae, andMb. thermoautotrophicum strains Marburg and H were able to grow with methanethiol, ethanethiol,n-propanethiol,n-butanethiol, methyl sulfide, dimethyl sulfoxide, ethyl sulfide, or CS₂ as a sulfur source, when very low levels (20–30 M) of sulfide were present; no growth occurred on 5–100 M sulfide alone. Methanethiol, ethanethiol, and methyl sulfide-using cultures produced sulfide during growth.     

Methylation of the nucleobases in RNA oligonucleotides mediates duplex-hairpin conversion

We have systematically investigated the duplex to hairpin conversion of oligoribonucleotides under the aspect of nucleobase methylation. The first part of our study refers to the self-complementary sequence rCGCGAAUUCGCGA, which forms a stable Watson–Crick base paired duplex under various buffer conditions. It is shown that this sequence is forced to adopt a hairpin conformation if one of the central 6 nt is replaced by the corresponding methylated nucleotide, such as 1-methylguanosine N²,N²-dimethylguanosine, N⁶,N⁶-dimethyladenosine (m⁶₂A) or 3-methyluridine. On the other hand, the duplex structure is retained and even stabilized by replacement of a central nucleotide with N²-methylguanosine (m²G) or N⁴-methylcytidine. A borderline case is represented by N⁶-methyladenosine (m⁶A). Although generally a duplex-preserving modification, our data indicate that m⁶A in specific strand positions and at low strand concentrations is able to effectuate duplex–hairpin conversion. Our studies also include the ssu ribosomal helix 45 sequence motif, rGACCm²GGm⁶₂Am⁶₂AGGUC. In analogy, it is demonstrated that the tandem m⁶₂A nucleobases of this oligoribonucleotide prevent duplex formation with complementary strands. Therefore, it can be concluded that nucleobase methylations at the Watson–Crick base pairing site provide the potential not only to modulate but to substantially affect RNA structure by formation of different secondary structure motifs.     

Effect of pH on Anaerobic Mild Steel Corrosion by Methanogenic Bacteria

Methanogens can use H₂ produced by cathodic depolarization-mediated oxidation of elemental iron to produce methane. Thermodynamic consideration of the cathodic depolarization mechanism predicts more oxidation of Fe⁰ at lower pH. Methanogenic responses to pH by Methanococcus deltae, Methanococcus thermolithotrophicus, and Methanosarcina barkeri were examined. When grown on H₂-CO₂, these bacteria had pH optima from 6.2 to 7.0, but when all H₂ was supplied from Fe⁰, methanogenic pH optima were lower, 5.4 to 6.5. Corrosion was monitored with and without cultures and at various pHs; more corrosion occurred when cultures were present, biologically induced corrosion was greatest at the pH optima for methanogenesis from Fe⁰, and corrosion without cultures increased with a drop in pH.     

Dynamical insight into <Emphasis Type="Italic">Caenorhabditis elegans</Emphasis> eIF4E recognition specificity for mono-and trimethylated structures of mRNA 5′ cap

Specific recognition and binding of the ribonucleic acid 5′ termini (mRNA 5′ cap) by the eukaryotic translation initiation factor 4E (eIF4E) is a key, rate limiting step in translation initiation. Contrary to mammalian and yeast eIF4Es that discriminate in favor of 7-methylguanosine cap, three out of five eIF4E isoforms from the nematode Caenorhabditis elegans as well as eIF4Es from the parasites Schistosome mansoni and Ascaris suum, exhibit dual binding specificity for both 7-methylguanosine-and N²,N²,7-trimethylguanosine cap. To address the problem of the differences in the mechanism of the cap recognition by those highly homologic proteins, we carried out molecular dynamics simulations in water of three factors, IFE-3 and IFE-5 isoforms from C. elegans and murine eIF4E, in the apo form as well as in the complexes with 7-methyl-GDP and N²,N²,7-trimethyl-GDP. The results clearly pointed to a dynamical mechanism of discrimination between each type of the cap, viz. differences in mobility of the loops located at the entrance into the protein binding pockets during the cap association and dissociation. Additionally, our data showed that the hydrogen bond involving the N²-amino group of 7-methylguanosine and the carboxylate of glutamic acid was not stable. The dynamic mechanism proposed here differs from a typical, static one in that the differences in the protein-ligand binding specificity cannot be ascribed to formation and/or disruption of well defined stabilizing contacts.     

Function of methylcobalamin: coenzyme M methyltransferase isoenzyme II in Methanosarcina barkeri

Methanosarcina barkeri was recently shown to contain two cytoplasmic isoenzymes of methylcobalamin: coenzyme M methyltransferase (methyltransferase 2). Isoenzyme I predominated in methanol-grown cells and isoenzyme II in acetate-grown cells. It was therefore suggested that isoenzyme I functions in methanogenesis from methanol and isoenzyme II in methanogenesis from acetate. We report here that cells of M. barkeri grown on trimethylamine, H₂/CO₂, or acetate contain mainly isoenzyme II. These cells were found to have in common that they can catalyze the formation of methane from trimethylamine and H₂, whereas only acetate-grown cells can mediate the formation of methane from acetate. Methanol-grown cells, which contained only low concentrations of isoenzyme II, were unable to mediate the formation of methane from both trimethylamine and acetate. These and other results suggest that isoenzyme II (i) is employed for methane formation from trimethylamine rather than from acetate, (ii) is constitutively expressed rather than trimethylamine-induced, and (iii) is repressed by methanol. The constitutive expression of isoenzyme II in acetate-grown M. barkeri can explain its presence in these cells. The N-terminal amino acid sequences of isoenzyme I and isoenzyme II were analyzed and found to be only 55% similar.Abbreviations H-S-CoM coenzyme M or 2-mercaptoethane-sulfonate - CH₃-S-CoM methyl-coenzyme M or 2(methylthio)-ethanesulfonate - [Co] cobalamin - CH₃-[Co] methylcobalamin - H₄MPT tetrahydromethanopterin - CH₃-H₄MPT N ⁵-methyltetrahydromethanopterin - MT₁ methyltransferase 1 or methanol: 5-hydroxybenzimidazolyl cobamide methyltransferase - MT₂ methyltransferase 2 or methylcobalamin: coenzyme M methyltransferase - Mops morpholinopropanesulfonate - 1 U = 1 mol/min     

Synthesis of guanosine and its derivatives from AICA-riboside

A novel and convenient method for the synthesis of guanosine is described. The reaction of AICA-riboside with sodium methylxanthate gave 2-mercaptoinosine in almost quantitative yield. The latter was oxidized with hydrogen peroxide to afford inosine-2-sulfonic acids, which was readily animated to give guanosine in excellent yield. Similarly, the preparation of N²-methylguanosine and N²,N²-dimethylguanosine, minor constituents of transfer RNA, was also accomplished. Furthermore, this procedure was extended to the synthesis of 2′,3′-O-isopropylideneguanosine and the isopropylidene derivatives of various N²-substituted guanosines from 2′,3′-O-isopropylidene-AICA-riboside. Guanosine via 2′,3′-O-isopropylideneguanosine was successfully phosphorylated to give 5′-guanylic acid.     

Quantification of corrinoids in methanogenic bacteria

Corrinoids in several diverse species of methanogens were quantified by a bioassay utilizingEscherichia coli 113–3, a corrinoid auxotroph. All five species examined contained >0.65 nmol corrinoid/mg dry cells when grown on H₂/CO₂ as carbon and energy source. The highest corrinoid levels (4.1 nmol/mg cells) were found inMethanosarcina barkeri grown on methanol. The amount of corrinoids found in this species was dependent on growth conditions, but, regardless of energy source, metabolized levels inMethanosarcina barkeri were higher than those found in theMethanobacterium species examined (M. arbophilicum, M. formicium, M. ruminantium, andM. thermoautotrophicum).     

Overproduction and one-step purification of the N 5,N 10-methenyltetrahydro-methanopterin cyclohydrolase (Mch) from the hyperthermophilic Methanopyrus kandleri

N ⁵,N ¹⁰-Methenyltetrahydromethanopterin cyclohydrolase (Mch) is an enzyme involved in methanogenesis from CO₂ and H₂ which represents the energy metabolism of Methanopyrus kandleri, a methanogenic Archaeon growing at a temperature optimum of 98°C. The gene mch from M. kandleri was cloned, sequenced, and expressed in Escherichia coli. The overproduced enzyme could be purified in yields above 90% in one step by chromatography on phenyl Sepharose in 80% ammonium sulfate. From 3.5 g cells (250 mg protein), approximately 18 mg cyclohydrolase was obtained. The purified enzyme showed essentially the same catalytic properties as the enzyme purified from M. kandleri cells. The primary structure and properties of the cyclohydrolase are compared with those of the enzyme from Methanococcus jannaschii (growth temperature optimum 85°C), from Methanobacterium thermoautotrophicum (65°C), and from Methanosarcina barkeri (37°C). Of the four enzymes, that from M. kandleri has the lowest isoelectric point (3.8) and the lowest hydrophobicity of amino acid composition. Besides, it has the highest relative content of glutamate, leucine, and valine and the lowest relative content of isoleucine, serine, and lysine. Some of these properties are unusual for enzymes from hyperthermophilic organisms. They may reflect the observation that the cyclohydrolase from M. kandleri is not only adapted to hyperthermophilic conditions but also to the high intracellular concentrations of lyotrophic salts prevailing in this organism. Received: July 14, 1997 / Accepted: August 28, 1997     

Effect of trace elements and vitamins on the growth of Methanosarcina barkeri

Growth of Methanosarcina barkeri on methanol as energy source was found to be dependent on cobalt and molybdenum. In the presence of 10^?6 M Co and 5 × 10^?7M Mo optimal growth occurred. Furthermore it could be demonstrated that nickel and selenium each in a concentration of 10^?7 M stimulated the growth of this methanogenic bacterium while the following elements tested in the range of 10^?7 M to 10^?3 had no influence: B, Cr, Cu, Mn, Pb. The requirement of Co and Ni for optimal growth are in accordance with the results that the cells contain the Co containing corrinoid Factor III (0.1 – 0.2 mg 5-hydroxylbenzimidazolylcyanocobamide per g wet cells) and Factor F₄₃₀, a nickel component. Studies on the vitamin dependency of M. barkeri showed that this strain needs only the vitamin riboflavin for the growth in a defined medium. Under these conditions a cell density of 2.6 g dry cells/l could be obtained in a fed batch culture.     

Pyruvate oxidation by Methanococcus spp.

In the absence of H₂, Methanococcus spp. utilized pyruvate as an electron donor for methanogenesis. For Methanococcus voltae A3, Methanococcus maripaludis JJ1, and Methanococcus vannielii, typical rates of pyruvate-dependent methanogenesis were 3.4, 2.8, and 3.9 nmol min^-1 mg^-1 cell dry wt, respectively. These rates were 1–4% of the rates of H₂-dependent methanogenesis. For M. voltae, the concentration of pyruvate required for one-half the maximum rate of methanogenesis was 7 mM, and pyruvate-dependent methanogenesis was linear for 3 days. Radiolabeled acetate was formed from [3-¹⁴C]pyruvate, and the stoichiometry of pyruvate consumed per acetate produced was 1.12±0.27. The stoichiometry of pyruvate consumed per CH₄ produced was 3.64±0.34. These values are close to the expected values of 1 acetate and 4 CH₄. Although 10–30% of total cell carbon could be obtained from exogenous pyruvate during growth with H₂, pyruvate did not replace the nutritional requirement for acetate in Methanococcus voltae A3 or two acetate auxotrophs of Methanococcus maripaludis, JJ6 and JJ7. These results suggest that pyruvate was not oxidized in the presence of H₂. The inability to oxidize pyruvate during H₂-dependent methanogenesis would prevent a futile cycle of pyruvate oxidation and biosynthesis during autotrophic growth.     

Function of coenzyme F420-dependent NADP reductase in methanogenic archaea containing an NADP-dependent alcohol dehydrogenase

Methanogenic archaea growing on ethanol or isopropanol as the electron donor for CO₂ reduction to CH₄ contain either an NADP-dependent or a coenzyme F₄₂₀-dependent alcohol dehydrogenase. We report here that in both groups of methanogens, the N ⁵, N ¹⁰-methylenetetrahydromethanopterin dehydrogenase and the N ⁵, N ¹⁰-methylenetetrahydromethanopterin reductase, two enzymes involved in CO₂ reduction to CH₄, are specific for F₄₂₀. This raised the question how F₄₂₀H₂ is regenerated in the methanogens with an NADP-dependent alcohol dehydrogenase. We found that these organisms contain catabolic activities of an enzyme catalyzing the reduction of F₄₂₀ with NADPH. The F₄₂₀-dependent NADP reductase from Methanogenium organophilum was purified and characterized. The N-terminal amino acid sequence showed 42% sequence identity to a putative gene product in Methanococcus jannaschii, the total genome of which has recently been sequenced. Received: 12 May 1997 / Accepted: 1 July 1997     

Purification and characterization of glycerate kinase from the thermoacidophilic archaeonThermoplasma acidophilum: An enzyme belonging to the second glycerate kinase family

Thermoplasma acidophilum is a thermoacidophilic archaeon that grows optimally at 59°C and pH 2. Along with another thermoacidophilic archaeon,Sulfolobus solfataricus, it is known to metabolize glucose by the non-phosphorylated Entner-Doudoroff (nED) pathway. In the course of these studies, the specific activities of glyceraldehyde dehydrogenase and glycerate kinase, two enzymes that are involved in the downstream part of the nED pathway, were found to be much higher inT. acidophilum than inS. solfataricus. To characterize glycerate kinase, the enzyme was purified to homogeneity fromT. acidophilum cell extracts. TheN-terminal sequence of the purified enzyme was in exact agreement with that of Ta0453m in the genome database, with the removal of the initiator methionine. Furthermore, the enzyme was a monomer with a molecular weight of 49 kDa and followed Michaelis-Menten kinetics withK _m values of 0.56 and 0.32 mM forDL-glycerate and ATP, respectively. The enzyme also exhibited excellent thermal stability at 70°C. Of the seven sugars and four phosphate donors tested, onlyDL-glycerate and ATP were utilized by glycerate kinase as substrates. In addition, a coupled enzyme assay indicated that 2-phosphoglycerate was produced as a product. When divalent metal ions, such as Mn²⁺, Co²⁺, Ni²⁺, Zn²⁺, Ca²⁺, and Sr²⁺, were substituted for Mg²⁺, the enzyme activities were less than 10% of that obtained in the presence of Mg²⁺. The amino acid sequence ofT. acidophilum glycerate kinase showed no similarity withE. coli glycerate kinases, which belong to the first glycerate kinase family. This is the first report on the biochemical characterization of an enzyme which belongs to a member of the second glycerate kinase family.     

Hybridization of DNA from methanogenic bacteria with nitrogenase structural genes (nifHDK)

Summary Using the Southern hybridization technique, homologies were examined between restricted DNA of four methanogenic bacteria (Methanobacterium ivanovi, Methanobacterium thermoautotrophicum, Methanococcus voltae, Methanosarcina barkeri) and the nif (nitrogen fixation) genes of Klebsiella pneumoniae and Anabaena strain 7120. With K. pneumoniae probes, no hybridization was observed with nifA, nifNE, and nifJ but positive results were obtained with the nifHDK genes coding for nitrogenase. Homology was detected, in the four strains, with K. pneumoniae and Anabaena nifH probes. In M. voltae and M. ivanovi, the homology found with nifH was estimated to be about 70% and a weaker hybridization was observed also with nifD and nifK. In M. voltae, the sequence homologous to nifH was found on a 3.0 kbp HindIII fragment and sequences homologous to nifD and nifK on a 3.8 kbp HindIII fragment. The 3.0 kbp fragment was cloned and the region homologous to nifH was localized more precisely. When this fragment was used as a probe against other DNAs, it behaved as a K. pneumoniae and Anabaena nifH probe. The results suggest that the structural genes for nitrogenase may be present in archaebacteria and raise interesting questions regarding their evolution.     

Lysine-2,3-Aminomutase and β-Lysine Acetyltransferase Genes of Methanogenic Archaea Are Salt Induced and Are Essential for the Biosynthesis of Nɛ-Acetyl-β-Lysine and Growth at High Salinity

The compatible solute N-acetyl-β-lysine is unique to methanogenic archaea and is produced under salt stress only. However, the molecular basis for the salt-dependent regulation of N-acetyl-β-lysine formation is unknown. Genes potentially encoding lysine-2,3-aminomutase (ablA) and β-lysine acetyltransferase (ablB), which are assumed to catalyze N-acetyl-β-lysine formation from α-lysine, were identified on the chromosomes of the methanogenic archaea Methanosarcina mazei Gö1, Methanosarcina acetivorans, Methanosarcina barkeri, Methanococcus jannaschii, and Methanococcus maripaludis. The order of the two genes was identical in the five organisms, and the deduced proteins were very similar, indicating a high degree of conservation of structure and function. Northern blot analysis revealed that the two genes are organized in an operon (termed the abl operon) in M. mazei Gö1. Expression of the abl operon was strictly salt dependent. The abl operon was deleted in the genetically tractable M. maripaludis. Δabl mutants of M. maripaludis no longer produced N-acetyl-β-lysine and were incapable of growth at high salt concentrations, indicating that the abl operon is essential for N-acetyl-β-lysine synthesis. These experiments revealed the first genes involved in the biosynthesis of compatible solutes in methanogens.

     

Conversion of lactose to methane by defined bacterial cocultures

The conversion of lactose — the main constituent of whey — to methane and carbon dioxide was studied using different defined constructed cultures, imploying strains of Methanosarcina barkeri, Methanobacterium bryantii, Escherichia coli, Acetobacterium woodii, Lactobacillus casei, and Lactobacillus plantarum. The following combinations of strains (food chains) were studied with respect to efficiency and yield of lactose conversion (methane yield in parentheses): E. coli and M. barkeri (4.5–7.6%), E. coli and M. bryantii (13.3%),E. coli, M. barkeri and M. bryantii (54%), L. casei, A. woodii and M. barkeri (93.3%). These conversions were carried out in pH controlled batch fermentations. A very efficient coculture was a combination of L. plantarum with A. woodii and M. barkeri: in chemostat cultures lactose was converted to methane and carbon dioxide with a yield of about 90%, at dilution rates of 0.27 d^-1to 0.37 d^-1.