The structure of methanol dehydrogenase (MDH) at 0.194 nm (1.94 A) has been used to provide a model structure for part of a membrane quinoprotein glucose dehydrogenase (GDH). The basic superbarrel structure is retained, along with the tryptophan-docking motifs. The active-site regions are similar, but there are important differences, the most important being that GDH lacks the novel disulphide ring structure formed from adjacent cysteines in MDH; in GDH the equivalent region is occupied by His-262. Because of the overall similarities in the active-site region, the mechanism of action of GDH is likely to be similar to that of MDH. The differences in co-ordination to the cation and bonding to the pyrrolo-quinoline quinone (PQQ) in the active site may explain the relative ease of dissociation of the prosthetic group from the holo-GDH. There are considerable differences in the external loops, particularly those involved in formation of the shallow funnel leading to the active site, the configuration of which influences substrate specificity. The proposed model is consistent in many respects with previous proposals for the active-site structure based on the effects of chemical modification on binding of PQQ and enzymic activity.

A fluorescence method is described for direct measurement of the interaction between methanol dehydrogenase (MDH) and its electron acceptor cytochrome cL. This has permitted a distinction to be made between factors affecting electron transfer and those affecting the initial binding or docking process. It was confirmed that the initial interaction is electrostatic, but previous conclusions with respect to the mechanism of EDTA inhibition have been modified. It is proposed that the initial ‘docking’ of MDH and cytochrome cL is by way of ionic interactions between lysyl residues on its surface and carboxylate groups on the surface of cytochrome cL. This interaction is not inhibited by EDTA, which we suggest acts by binding to nearby lysyl residues, thus preventing movement of the ‘docked’ cytochrome to its optimal position for electron transfer, which probably involves interaction with the hydrophobic funnel in the surface of MDH.

The 1.94 A structure of methanol dehydrogenase has been used to provide a model structure for part of a membrane quinohaemoprotein alcohol dehydrogenase. The basic superbarrel structure and the active-site region are retained, indicating essentially similar mechanisms of action, but there are considerable differences in the external loops, particularly those involved in formation of the shallow funnel leading to the active site.

All cysteines in methanol dehydrogenase (MDH) from Methylobacterium extorquens are involved in intra-subunit disulphide bridge formation. One of these is between adjacent cysteine residues which form a novel ring structure in the active site. It is readily reduced, the reduced enzyme being inactive in electron transfer to cytochrome cL. The inactivation is not a result of major structural change or to modification of the prosthetic group pyrrolo-quinoline quinone (PQQ). The reduced enzyme appears to remain active with the artificial electron acceptor phenazine ethosulphate but this is because the dye re-oxidizes the adjacent thiols back to the original disulphide bridge. No free thiols were detected during the reaction cycle with cytochrome cL. Carboxymethylation of the thiols produced by reduction of the novel disulphide ring led to formation of active enzyme. Reconstitution of inactive Ca(2+)-free MDH with Ca2+ led to active enzyme containing the oxidized bridge and reduced quinol, PQQH2, consistent with the conclusion that no hydrogen transfer occurs between these groups in the active site. It is concluded that the disulphide ring in the active site of MDH does not function as a redox component of the reaction. The disulphide ring has no special function in the process of Ca2+ incorporation into the active site. It is suggested that this novel structure might function in the stabilization or protection of the free radical semiquinone form of the prosthetic group (PQQH.) from solvent at the entrance to the active site.

Methanol dehydrogenase (MDH) from Methylobacterium extorquens, Methylophilus methylotrophus, Paracoccus denitrificans and Hyphomicrobium X all contained a single atom of Ca2+ per alpha 2 beta 2 tetramer. The role of Ca2+ was investigated using the MDH from Methylobacterium extorquens. This was shown to be similar to the MDH from Hyphomicrobium X in having 2 mol of prosthetic group (pyrroloquinoline quinine; PQQ) per mol of tetramer, the PQQ being predominantly in the semiquinone form. MDH isolated from the methanol oxidation mutants MoxA-, K- and L- contained no Ca2+. They were identical with the enzyme isolated from wild-type bacteria with respect to molecular size, subunit configuration, pI, N-terminal amino acid sequence and stability under denaturing conditions (low pH, high urea and high guanidinium chloride) and in the nature and content of the prosthetic group (2 mol of PQQ per mol of MDH). They differed in their lack of Ca2+, the oxidation state of the extracted PQQ (fully oxidized), absence of the semiquinone form of PQQ in the enzyme, reactivity with the suicide inhibitor cyclopropanol and absorption spectrum, which indicated that PQQ is bound differently from that in normal MDH. Incubation of MDH from the mutants in calcium salts led to irreversible time-dependent reconstitution of full activity concomitant with restoration of a spectrum corresponding to that of fully reduced normal MDH. It is concluded that Ca2+ in MDH is directly or indirectly involved in binding PQQ in the active site. The MoxA, K and L proteins may be involved in maintaining a high Ca2+ concentration in the periplasm. It is more likely, however, that they fill a ‘chaperone’ function, stabilizing a configuration of MDH which permits incorporation of low concentrations of Ca2+ into the protein.

The quinoprotein methanol dehydrogenase (MDH) of Acetobacter methanolicus has an alpha 2 beta 2 structure. By contrast with other MDHs, the beta-subunit (approx. 8.5 kDa) does not contain the five lysine residues previously proposed to be involved in ionic interactions with the electron acceptor cytochrome cL. That electrostatic interactions are involved was confirmed by the demonstration that methanol:cytochrome cL oxidoreductase activity was inhibited by high ionic strength (I), the strength of interaction being inversely related to the square root of I. Specific modifiers of arginine residues on MDH inhibited this reaction but not the dye-linked MDH activity. Modification of lysine residues on MDH that altered its charge had no effect on the dye-linked activity but inhibited reaction with cytochrome cL. When the charge was retained on modification of lysine residues, little effect on either activity was observed. Cross-linking experiments confirmed that lysine residues on the alpha-subunit, but not the beta-subunit, are involved in the ‘docking’ process between the proteins.

The nucleotide and deduced amino acid sequence of a novel small (beta) subunit of methanol dehydrogenase of Methylobacterium extorquens AM1 (previously Pseudomonas AM1) has been determined. Work with the whole protein has shown that is has an alpha 2 beta 2 configuration.

The terminal respiratory oxidase was solubilized from membranes of organism 4025, an obligate methylotroph. The partially purified oxidase is probably a cytochrome co. It does not oxidize amicyanin, but it oxidizes ‘azurin’ and cytochromes cH and cL. By using a complete ‘methylamine oxidase’ system reconstituted from pure methylamine dehydrogenase, purified oxidase and soluble blue copper proteins and cytochromes, it was confirmed that amicyanin is essential for methylamine oxidation; it could not be replaced by ‘azurin’ or cytochrome cH or cL. It was shown that the usual mediator between amicyanin and the oxidase is cytochrome cH, with ‘azurin’ able to replace it during growth at the high copper concentrations required for optimum growth of this unusual methylotroph.

The nucleotide sequence and deduced amino acid sequence of the cytochrome cL of Methylobacterium extorquens (Pseudomonas AM1; Methylobacterium AM1) shows that this cytochrome c is completely different, except for its haem-binding site, from all other cytochromes.

Organism 4025, an obligate methylotroph, when grown on methylamine in the presence of a high concentration of copper, contained high concentrations of methylamine dehydrogenase and two blue copper proteins, amicyanin and an azurin-type protein; these were purified to homogeneity and characterized. The methylamine dehydrogenase is a basic protein (pI 8.8) and consists of light and heavy subunits (Mr 14100 and 43000; total Mr 112000). This dehydrogenase differed slightly from other methylamine dehydrogenases in its absorption spectrum and in its lack of thermal stability. Amicyanin, the more abundant blue copper protein, had an Mr of 11500, a midpoint redox potential of 294mV at pH 7.0, and a much lower isoelectric point (pI5.3) than other amicyanins. Its absorption maximum was 620 nm (7-24 nm higher than those of other amicyanins); its absorption coefficient (at 620 nm) was 3.8 mM-1 X cm-1. The ‘azurin’ (6% of the blue copper protein) had an Mr of 12500, a midpoint redox potential of 323 mV and a high isoelectric point (pI 9.4). Its absorption maximum was 620 nm, the absorption coefficient (16 mM-1 X cm-1) at this wavelength being considerably greater than that of any blue copper protein described previously. The partially-purified soluble cytochromes cH and cL were similar to those of other methylotrophs. The interactions of the purified redox proteins were investigated in order to elucidate their role in methylamine oxidation. Methylamine dehydrogenase was able to donate electrons only to amicyanin, the rate of reaction being 2.04 mmol/min per mumol of methylamine dehydrogenase; this is sufficient to account for the rate of respiration in whole bacteria. The blue copper proteins were able to react rapidly with each other and with both the soluble cytochromes c.

The two types of soluble cytochrome c (cytochrome cH and cytochrome cL) found in methylotrophs are completely distinct proteins; one type is not a dimer or degradation product of the other. Free thiol groups are probably not involved in the unusually rapid autoreduction of the cytochromes at high pH. The axial ligands to the haem iron, histidine and methionine, are the same as in other low-spin cytochromes c. The methionine ligand is displaced at high pH by an alternative strong-field ligand. This displacement does not occur on reduction of cytochrome cL by methanol dehydrogenase, but this does not rule out the possibility that the autoreduction mechanism is involved in the interaction of the dehydrogenase and cytochrome c.

The obligate methylotroph Methylophilus methylotrophus contains three distinct soluble cytochromes c. The major cytochromes, cytochrome cH (about 50% of the total) and cytochrome cL (about 42%), were similar in most respects to the cytochromes cH and cL of the facultative methylotroph Pseudomonas AM1 [O'Keeffe & Anthony (1980) Biochem. J. 192, 411-419]. Cytochrome cH had a high isoelectric point, a midpoint redox potential at pH 7.0 of 373 mV and a low molecular weight (8500). The cytochrome cL had a low isoelectric point, a midpoint potential of 310 mV and a molecular weight of 21,000. The third cytochrome, cytochrome cLM, was clearly distinct from cytochromes cH and cL. Like the cytochrome cL of Pseudomonas AM1, the cytochrome cL of M. methylotrophus had the lowest midpoint potential, it reacted most rapidly with methanol dehydrogenase and it combined to the greatest extent with CO. Cytochromes cH and L of M. methylotrophus differed from those from Pseudomonas AM1 in having unusually high midpoint redox potentials for non-photosynthetic bacteria and in exhibiting a split alpha-band at low temperatures.

Cytochromes cH and cL were autoreduced at high pH (pK greater than 10) and the autoreduced cytochromes reacted with CO. The autoreduction was first-order with respect to oxidized cytochrome c and was reversible by lowering the pH. Pure methanol dehydrogenase reduced cytochrome c (in the absence of methanol) by lowering the pK for autoreduction to less than 8.5. A mechanism is proposed for the autoreduction of cytochrome c and its involvement in the reaction with methanol dehydrogenase.

Two cytochromes b with absorption maxima at 555 and 562 nm and differing in their mid-point redox potentials are synthesized in Pseudomonas AM1 during growth on methanol or succinate in batch culture, or in NH4+-limited or carbon-limited continuous culture. Both cytochromes b were also present in a cytochrome c-deficient mutant in all growth conditions.

By using wild-type and deoxystreptamine-negative mutants of Streptomyces fradiae grown in media containing [6(-3)H]glucose or [U-14C]glucose, and by subsequent hydrolysis of the labelled neomycin produced, neamines labelled with 3H in both rings I and II, but with 14C in ring I only, were prepared. A mixture of these two forms of neamine was converted by deoxystreptamine-negative Streptomyces rimosus forma paromomycinus into neomycin (not paromomycin) with a 30% yield. The3H: 14C ratio in this neomycin was the same as the measured in neamine produced by hydrolysis of the neomycin, and in unused neamine reisolated from the incubation medium. The 3H:14C ratio in the neomycin was not affected by the presence of unlabelled deoxystreptamine during the incubation. The radioactivity in the neomycin was associated with rings I and II only. It is concluded that the added neamine is incorporated into antibiotic intact, without initial hydrolysis, and that the probable first step in the subunit assembly of neomycin is the formation of neamine.

Pseudomonas AM1 contains cytochromes a, b and c and more than one CO-binding pigment (cytochrome a3, cytochrome c and possibly a cytochrome o). The soluble cytochrome c has been purified; its isoelectric point is low and its molecular weight is 20000. This cytochrome is reduced in whole bacteria by all oxidizable substrates at rates determined by the primary dehydrogenases. A mutant lacking cytochrome c oxidizes all substrates except methanol, ethanol and methylamine; these no longer support growth. The role of cytochrome c in electron transport in Pseudomonas AM1 is discussed.

Pseudomonas AM1 grows on β-hydroxybutyrate and methanol at similar rates. β-Hydroxybutyrate is not metabolized by way of the glyoxylate bypass, but is assimilated by the novel route (with acetate as an intermediate) that operates during growth of this organism on ethanol. Evidence from short-term labelling experiments indicates that acetate, which is a possible intermediate in the assimilation of C 1 compounds, is rapidly metabolized to glycine during growth of Pseudomonas AM1 on methanol.

Pseudomonas AM1 grows on ethanol with a mean generation time of about 10h. A single alcohol dehydrogenase is responsible for oxidation of both methanol and ethanol. It is proposed that the glyoxylate bypass does not operate in Pseudomonas AM1 during growth on ethanol. Although malate synthase is present in extracts of ethanol-grown Pseudomonas AM1, the activity of isocitrate lyase is negligible. Short-term incubation experiments with [ 14 C]ethanol and [ 14 C]acetate indicate that a novel pathway operates during growth of Pseudomonas AM1 on ethanol. Glycollate, glyoxylate and malate are probably intermediates in this pathway.

Succinate (or a product of succinate metabolism) is a catabolite repressor of some enzymes of the serine pathway (hydroxypyruvate reductase, serine–glyoxylate aminotransferase and glycerate kinase) but not of methanol dehydrogenase nor methylamine dehydrogenase. A mutant (PCT64) of Pseudomonas AM1, which is unable to grow on C 1 compounds, lacks glycerate kinase, showing that this enzyme is essential for the operation of the serine pathway. Mutant PCT48, unable to convert acetate into glycollate, has lost the ability to grow both on C 1 compounds and on ethanol. The properties of a third mutant (PCT57) show that Pseudomonas AM1 contains enzymes catalysing the conversion of acetate into glyoxylate. Evidence is presented that hydroxypyruvate reductase is involved in the oxidation of glycollate to glyoxylate during growth on ethanol. A scheme is proposed for the conversion of ethanol and of C 1 compounds into glyoxylate in which acetate (or a derivative) and glycollate are intermediates.