Amines are a carbon source for the growth of a number of bacterial species and they also play key roles in neurotransmission, cell growth and differentiation, and neoplastic cell proliferation. Enzymes have evolved to catalyse these reactions and these oxidoreductases can be grouped into the flavoprotein and quinoprotein families. The mechanism of amine oxidation catalysed by the quinoprotein amine oxidases is understood reasonably well and occurs through the formation of enzyme–substrate covalent adducts with TPQ (topaquinone), TTQ (tryptophan tryptophylquinone), CTQ (cysteine tryptophylquinone) and LTQ (lysine tyrosyl quinone) redox centres. Oxidation of amines by flavoenzymes is less well understood. The role of protein-based radicals and flavin semiquinone radicals in the oxidation of amines is discussed.
- electron transfer
Our group has studied the mechanisms of amine oxidation by flavoproteins and the associated electron transfer reactions in a number of enzyme systems. The mechanisms of substrate oxidation by flavoproteins remain controversial, and this is particularly true for the oxidation of amine substrates. Mechanisms involving (i) proton abstraction by an active-site base to generate a carbanion species , (ii) an amminium radical cation species , (iii) H-atom abstraction by an active-site radical species  and (iv) nucleophilic attack by the substrate nitrogen on the flavin C4a atom, followed by proton abstraction by an active-site base  or the flavin N5 atom  (analogous to a similar mechanism proposed for D-amino acid oxidase ) have been considered over the years. A consensus view, however, has not emerged and this paper aims to provide an overview of recent developments arising from our own studies of the flavoenzymes TMADH (trimethylamine dehydrogenase) and MAO (monoamine oxidase).
One of the most remarkable developments in recent years has been the elucidation of crystallographic structures for MAO A  and B , and this has reinvigorated the longstanding debate concerning the mechanisms of these enzymes. Much of the debate has centred on the existence of radical species, direct evidence for which, until very recently , was not forthcoming. Several mechanisms have been proposed for the MAO catalysed oxidation of amines. An early proposal based on flavin model reactions invoked a polar nucleophilic mechanism involving attack of the deprotonated amine substrate at the flavin C4a to form a substrate–flavin C4a adduct. This is followed by proton abstraction from the α-carbon of the adduct by an active-site base . Reactivity at the flavin C4a atom was originally seen as an attractive mechanism, not only for the oxidation of amines, but also for dehydrogenation of alcohols and nicotinamide coenzymes, and for dehydrogenation of α, β to a carbonyl group . Much later, support for the polar nucleophilic mechanism also came from chemical model studies in reactions of amines with lumiflavins [4,11]. The proposed mechanism involves reversible protonation at the N5 atom of 3MLF (3-methyllumiflavin) and nucleophilic addition to form a lumiflavin 4a adduct. QSARs (quantitative structure–activity relationships) with MAO B were originally used to support an alternative mechanism in which substrate α-C–H bond cleavage is by direct hydrogen atom transfer to a protein-based non-flavin radical, followed by electron transfer to the flavin [3,12]. This mechanism requires a relatively stable protein-based radical in the active site of the enzyme. An organic radical species was originally reported in EPR spectra of resting (i.e. flavin in the oxidized state) bovine liver MAO B [13,14], but more recent studies have confirmed the lack of a protein-based radical in highly purified forms of recombinant MAO A and MAO B in the resting state [15,16]. The lack of an observed radical species in the resting states of MAO A and MAO B therefore was seen to seriously question the validity of this proposed mechanism for the MAO enzymes. Miller and Edmondson  have provided support for a concerted polar nucleophilic mechanism for MAO A involving a substrate–flavin C4a adduct and proton abstraction by the highly basic N5 atom of the flavin (Scheme 1). This mechanism is consistent with (i) observed electronic effects in QSAR studies with para- and meta-substituted benzylamines, (ii) kinetic isotope effects and (iii) the apparent lack of an organic protein-based radical in EPR spectra of the resting form of the enzyme. These results support a proton abstraction mechanism for the substrate C–H bond, but the results are also consistent with the amminium cation radical mechanism proposed by Silverman et al. , which is the most widely quoted mechanism for MAO catalysed oxidation of amines. The latter mechanism involves the transfer of a single electron from the substrate to the enzyme flavin to generate the aminyl radical cation; subsequent abstraction of H• (route d, Scheme 2 ) or deprotonation followed by radical transfer (route a, Scheme 2 ) yields the reduced flavin and product iminium ion; the latter route is consistent only with the strong correlation between reaction rate and electronic effects in QSAR studies with substituted benzylamine substrates  and potentially involves a protein-based radical (route c). The amminium radical cation mechanism is based on the susceptibility of amines to undergo single-electron transfer chemistry during electrochemical and chemical oxidations  and is consistent with the results of mechanism-based inhibitor studies with a series of cyclopropyl inhibitors, which undergo rapid ring opening on formation of the cyclopropylaminyl radical . Of course, the key evidence required to discriminate between the aminyl radical cation and polar nucleophilic mechanisms rests on detecting a radical species. Our recent EPR, ENDOR (electron-nuclear double resonance) and absorption studies have revealed the presence of a stable tyrosyl radical in the enzyme in partially reduced MAO A . The EPR spectrum of partially reduced MAO A (Figure 1) has features not observed for typical anionic flavosemiquinones  and subtraction of typical anionic flavosemiquinone EPR spectra produces a residual spectrum that exhibits a gav of 2.0042 . This value is too high to arise from the flavosemiquinone radical (2.0032) , but is consistent with assignment as a neutral tyrosine (tyrosyl) radical . The radical accounts for 18% of the total number of unpaired electrons observed. The hyperfine splitting observed in the EPR spectrum is typical of a neutral tyrosine radical . The EPR spectrum for partially reduced enzyme indicates a smaller linewidth than previously reported for a tyrosyl radical, and we have attributed this to electron exchange between two sites or species on a timescale that is rapid compared with the spectrometer frequency , perhaps reflecting rapid exchange between spatially close tyrosine residues in the active site of MAO. We have also used ENDOR spectroscopy to measure the hyperfine couplings to the protons of the tyrosyl radical . The ENDOR spectrum of the YD• tyrosyl radical of Photosystem II from Phormidium laminosum is similar to that of MAO A . The rotameric angle of the tyrosyl radical ring relative to the β-CH2 group has been determined from the β-CH2 hyperfine coupling constants. The angles θ calculated for the two β-CH2 protons were found to be 54.6 and 61.1° and the structure of MAO B shows that tyrosine residues Tyr60, Tyr398 and Tyr435 (equivalent toTyr69, Tyr407 and Tyr444 in MAO A) are in the vicinity of the active site . Of these three, Tyr398 (Tyr407 in MAO A) has θ values closest to those measured for the tyrosyl radical (∼50° and 70°), and the orientation is similar in the MAO A structure. Mutation of Tyr69 to Ala, Ser or Phe has no effect on activity. Mutation of Tyr407 and Tyr444 to Ser [Y407S (Tyr407→Ser), Y444S] in MAO A leads to loss of activity ; however, mutation to Phe has a much larger effect at position 444 (Y444F) than 407 (Y407F). The presence of a tyrosyl radical in partially reduced MAO A is consistent with radical chemistry for substrate oxidation. The proposed mechanism requires that a redox equilibrium exists between the flavosemiquinone and a tyrosine radical in the active site to facilitate oxidation of the substrate. On establishing the tyrosyl radical, the reaction could occur either by direct H transfer or, alternatively, by proton transfer and formation of a radical centred on the α-carbon route (route a to c in Scheme 2). The latter route is consistent with the electronic effects seen in QSAR analysis of MAO A  and with studies of mechanism-based inhibitors [24,25]. There is no obvious amino acid proton acceptor in the active site of MAO A  and MAO B , except for two tyrosine residues, Tyr398 and Tyr435 in MAO B. However, mutagenesis of the homologous residues in MAO A does not abolish catalytic activity . During catalytic turnover, reduction of the enzyme by single-electron transfer from the substrate to flavin generates a redox equilibrium with appearance of the tyrosine radical and the aminyl radical cation intermediate. The radical must be short-lived in reactions with substrate because in rapid scan stopped-flow experiments an absorption band for the tyrosyl radical species is not observed . This is consistent with (i) reversible electron transfer from substrate to flavin, with the reverse reaction being fast, and (ii) rapid establishment of an equilibrium with FADH2 and the tyrosyl radical species.
We have also studied the mechanism of amine oxidation and associated electron transfer reactions in the bacterial enzyme TMADH. This enzyme is a complex iron–sulphur flavoprotein that catalyses the oxidative demethylation of trimethylamine to form dimethylamine and formaldehyde . Substrate oxidation involves the transfer of reducing equivalents from substrate to a 6-S-cysteinyl FMN and is accompanied by a large KIE (kinetic isotope effect) in native and mutant forms of the enzyme . Temperature-dependent studies of the KIE indicate that H transfer from substrate occurs by quantum mechanical tunnelling . Following reduction of the flavin, electrons are transferred from the dihydroflavin to a 4Fe-4S centre [30–33]. Subsequent electron transfer is from the 4Fe-4S centre of TMADH to an ETF (electron transferring flavoprotein), and this involves the assembly of a highly conformationally dynamic TMADH–2ETF electron transfer complex [34–36]. The structure of the TMADH–2ETF indicates that domain motion in ETF is pivotal to the electron transfer between TMADH and ETF, and our work on the complex formed between medium chain acyl-CoA dehydrogenase and ETF indicates this mechanism as a general feature of ETF–protein complexes . In stopped-flow kinetic studies, the reductive half-reaction of TMADH (i.e. rapid mixing of oxidized enzyme and substrate) occurs in three kinetic phases . A fast phase represents two-electron reduction of the 6-S-cysteinyl FMN, and an intermediate and slow phase report on electron transfer from the dihydroflavin to 4Fe-4S and the formation of a spin interacting state between the flavin semiquinone and reduced 4Fe-4S centre respectively. Ionization of Tyr169 on binding substrate in two-electron reduced enzyme, or on elevating the solution pH to values >9.5 effects an electron redistribution in the flavin semiquinone so that the unpaired spin becomes coupled with that located in the reduced 4Fe-4S centre, thus generating an unusual spin interaction signal in EPR spectra . Similar spin interacting species between a flavin semiquinone and iron–sulphur centre have been seen in our more recent studies of electron distribution during titrations of dihydro-orotate dehydrogenase from Lactococcus casei . Over the years there have been a number of mechanistic proposals for substrate oxidation by TMADH. An early proposal considered a carbanion mechanism in which an activesite base deprotonates a substrate methyl group to form a substrate carbanion ; reduction of the flavin was then achieved by the formation of a carbanion–flavin N5 adduct, with subsequent formation of the product imine and dihydroflavin. Active-site residues were identified as potential bases to support such a reaction mechanism, but mutagenesis and stopped-flow kinetic studies have eliminated the participation of these residues in a carbanion-type mechanism [39,41–43], indicating that a proton abstraction mechanism initiated by an active-site residue does not occur. Earlier mechanistic proposals also invoked the trimethylammonium cation as the reactive species in the enzyme–substrate complex, owing to the high pKa (9.81) of free trimethylamine, and this was used to argue against mechanisms requiring trimethylamine base. Recent stopped-flow studies with trimethylamine and perdeuterated trimethylamine have now established that trimethylamine base is in fact the reactive species in the enzyme–substrate complex  and this has led to a mechanism involving addition of trimethylamine base at the C4a position of the flavin and abstraction of a substrate proton by the N5 atom of the flavin . This mechanism is analogous to that proposed previously by Miller and Edmondson for MAO A based on QSAR analysis with para-substituted benzylamines  and is consistent with computational studies of TMADH that have indicated that the C4a position is an electrophilic centre . Studies of inactivation of TMADH by phenylhydrazine where modification of the flavin occurs at the C4a position also support this mechanism . Other mechanistic possibilities, however, exist. For example, the equivalent of the aminyl radical cation mechanism proposed for MAO remains a possibility, but evidence for a protein based in TMADH is still lacking.
How does TMADH lower the pKa of trimethylamine (from 9.81 to ∼6.5) to facilitate catalysis with the free base form? We have investigated the pH dependence of flavin reduction by trimethylamine and perdeuterated trimethylamine in detail, and we have identified two kinetically influential ionizations . The first ionization is perturbed by approx. 0.5 pH units to higher pH when trimethylamine is replaced by perdeuterated trimethylamine, consistent with deprotonation of substrate. The second ionization is attributed to His172 in the active site of TMADH as this ionization is lost in the H172Q mutant enzyme , and is perturbed in a Y169F mutant enzyme (Tyr169 forms a hydrogen bond to His172 in native enzyme). The substrate pKa in the H172Q enzyme–substrate complex is perturbed and elevated, indicating that His172 is (partially) responsible for the lowering of the substrate pKa (by ∼1.5 pH units) when bound to the enzyme . Residue Tyr60 also plays a major role since replacement of Tyr60 by Phe elevates the substrate pKa in the enzyme–substrate complex by approx. 1.3 pH units to a value of approx. 8.8. In the double mutant (H172Q, Y60F), the substrate pKa is raised even further to approx. 9.3, which is close to that of the free trimethylamine base . These results indicate key roles for His172 and Tyr60 in stabilizing the basic form of the substrate in the enzyme active site, thus facilitating catalysis at physiological pH values where the lone pair on the substrate nitrogen is required to initiate substrate oxidation.
We conclude with the following remarks. Our kinetic, structural, spectroscopic and mutagenesis studies indicate that protein-based and flavin semiquinone radicals play a major role in the mechanism of oxidation of amine substrates and subsequent electron transfer reactions. Evidence in favour of the aminyl radical cation mechanism for MAO is now available, but much is still to be done. By no means can we conclude that similar mechanisms operate in other flavoprotein amine oxidases/dehydrogenases, and progress will require further detailed study using the tools of fast reaction kinetics, spectroscopy and mutagenesis.
Coenzymology: the biochemistry of vitamin biogenesis and cofactor-containing enzymes: Independent Meeting held at King's College, Cambridge, U.K., 4–7 April 2005. Organized and Edited by A.G. Smith (Cambridge, U.K.) and A.W. Munro (Leicester, U.K.).
Abbreviations: ENDOR, electron-nuclear double resonance; ETF, electron transferring flavoprotein; KIE, kinetic isotope effect; MAO, monoamine oxidase; TMADH, trimethylamine dehydrogenase; QSAR, quantitative structure–activity relationship
- © 2005 The Biochemical Society