Advertisement

Contrasting roles for two conserved arginines: Stabilizing flavin semiquinone or quaternary structure, in bifurcating electron transfer flavoproteins

Open AccessPublished:February 14, 2022DOI:https://doi.org/10.1016/j.jbc.2022.101733
      Bifurcating electron transfer flavoproteins (Bf ETFs) are important redox enzymes that contain two flavin adenine dinucleotide (FAD) cofactors, with contrasting reactivities and complementary roles in electron bifurcation. However, for both the “electron transfer” (ET) and the “bifurcating” (Bf) FADs, the only charged amino acid within 5 Å of the flavin is a conserved arginine (Arg) residue. To understand how the two sites produce different reactivities utilizing the same residue, we investigated the consequences of replacing each of the Arg residues with lysine, glutamine, histidine, or alanine. We show that absence of a positive charge in the ET site diminishes accumulation of the anionic semiquinone (ASQ) that enables the ET flavin to act as a single electron carrier, due to depression of the oxidized versus. ASQ reduction midpoint potential, E°OX/ASQ. Perturbation of the ET site also affected the remote Bf site, whereas abrogation of Bf FAD binding accelerated chemical modification of the ET flavin. In the Bf site, removal of the positive charge impaired binding of FAD or AMP, resulting in unstable protein. Based on pH dependence, we propose that the Bf site Arg interacts with the phosphate(s) of Bf FAD or AMP, bridging the domain interface via a conserved peptide loop (“zipper”) and favoring nucleotide binding. We further propose a model that rationalizes conservation of the Bf site Arg even in non-Bf ETFs, as well as AMP's stabilizing role in the latter, and provides a mechanism for coupling Bf flavin redox changes to domain-scale motion.

      Keywords

      Abbreviations:

      AfeETF (ETF from Acidaminococcus fermentans), Arg (arginine), ASQ (anionic semiquinone), Bf (bifurcating), Bf ETF (bifurcating electron transfer flavoprotein), CT (charge transfer), ET (electron transfer), FAD (flavin adenine dinucleotide), HQ (hydroquinone), MelETF (ETF from Megasphaera elsdenii), RpaETF (ETF of Rhodopseudomonas palustris)
      Electron transfer flavoproteins (ETFs) are heterodimeric proteins whose fold is described by three domains (Fig. 1). The EtfA monomer includes domains I and II while domain III derives from the EtfB monomer. Domain II has been shown to reorient by some 80° relative to the base comprised of domains I and III (
      • Toogood H.S.
      • Leys D.
      • Scrutton N.S.
      Dynamics driving function : New insights from electron transferring flavoproteins and partner complexes.
      ,
      • Toogood H.S.
      • van Thiel A.
      • Scrutton N.S.
      • Leys D.
      Stabilization of non-productive conformations underpins rapid electron transfer to electron transferring flavoprotein.
      ), and this conformational change has been proposed to gate electron transfer between the flavins of Bf ETF in the course of turnover (Fig. S1 and (
      • Demmer J.K.
      • Bertsch J.
      • Oppinger C.
      • Wohlers H.
      • Kayastha K.
      • Demmer U.
      • Ermler U.
      • Muller V.
      Molecular basis of the flavin-based electron-bifurcating caffeyl-CoA reductase reaction.
      ,
      • Demmer J.K.
      • Chowdhury N.P.
      • Selmer T.
      • Ermler U.
      • Buckel W.
      The semiquinone swing in the bifurcating electron transferring flavoprotein/butyryl-coA dehydrogenase complex from Clostridium difficile.
      )).
      Figure thumbnail gr1
      Figure 1Structure of Bf ETF in the B-like (closed) conformation, interactions with two target Args, and essence of flavin-based bifurcation reaction. A, model of RpaETF based on coordinates 4KPU.pdb generated using Swiss Model (
      • Waterhouse A.
      • Bertoni M.
      • Bienert S.
      • Studer G.
      • Tauriello G.
      • Gumienny R.
      • Heer F.T.
      • de Beer T.A.P.
      • Rempfer C.
      • Bordoli L.
      • Lepore R.
      • Schwede T.
      SWISS-MODEL: Homology modelling of protein structures and complexes.
      ) and then subjected to geometry optimization (
      • Huang J.
      • Rauscher S.
      • Nawrocki G.
      • Ran T.
      • Feig M.
      • de Groot B.L.
      • Grubmüller H.
      • MacKerell Jr., A.D.
      CHARMM36m: An improved force field for folded and intrinsically disordered proteins.
      ) and validated by computation of electronic spectra for each of the two flavins (
      • Mohamed-Raseek N.
      • Duan H.D.
      • Mroginski M.A.
      • Miller A.F.
      Spectroscopic, thermodynamic and computational evidence of the locations of the FADs in the nitrogen fixation-associated electron transfer flavoprotein.
      ). The “A” subunit is in blue and the “B” subunit is in gray, domains are also indicated (
      • Mohamed-Raseek N.
      • Duan H.D.
      • Mroginski M.A.
      • Miller A.F.
      Spectroscopic, thermodynamic and computational evidence of the locations of the FADs in the nitrogen fixation-associated electron transfer flavoprotein.
      ). B, disposition of Arg-273 adjacent to the ET flavin bound in domain II of EtfA, note that Arg-273 derives from EtfA. C, disposition of R165 perpendicular to the Bf flavin, interacting with both it and a stretch of peptide backbone from the other subunit. Arg-165 derives from EtfA (blue Cα), whereas the stretch of peptide derives from EtfB (gray Cα), and the Bf flavin resides in the interface between domains I and III. All molecular graphics images were generated using Chimera (
      • Pettersen E.F.
      • Goddard T.D.
      • Huang C.C.
      • Couch G.S.
      • Greenblatt D.M.
      • Meng E.C.
      • Ferrin T.E.
      UCSF Chimera - a visualization system for exploratory research and analysis.
      ). D, depiction of flavin's role in electron bifurcation, accepting a hydride equivalent to two electrons (2e) from NADH, transmitting one electron (1e) to a higher-energy electron carrier (here ferredoxin, Fd) as paid for by exergonic transfer of the other electron to quinone via other carriers (here ET flavin and FixCX). The Bf flavin is indicated by the fused green hexagons. Note that two NADH-consuming turnovers are required to supply the two reducing equivalents needed to convert one quinone (Q) to a quinol (QH2). Bf, bifurcating; RpaETF, ETF of Rhodopseudomonas palustris.
      The canonical ETFs were discovered first, in mitochondria, and employ a single FAD to mediate single-electron transfer from client CoA dehydrogenases to the quinone pool via an ETF quinone oxidoreductase, at relatively high reduction midpoint potentials (s) (
      • Thorpe C.
      Electron-Transferring flavoproteins.
      ,
      • Jones M.
      • Basran J.
      • Sutcliffe M.J.
      • Grossmann J.G.
      • Scrutton S., N.
      X-ray scattering studies of Methylophilus methylotrophus (sp W(3)A(1)) electron-transferring flavoprotein - evidence for multiple conformational states and an induced fit mechanism for assembly with trimethylamine dehydrogenase.
      ,
      • Roberts D.L.
      • Frerman F.E.
      • Kim J.J.
      Three-dimensional structure of human electron transfer flavoprotein to 2.1-A resolution.
      ,
      • Roberts D.L.
      • Salazar D.
      • Fulmer J.P.
      • Frerman F.E.
      • Kim J.J.
      Crystal structure of paracoccus denitrificans electron transfer flavoprotein: Structural and electrostatic analysis of a conserved flavin binding domain.
      ,
      • Burgess S.G.
      • Messiha H.L.
      • Katona G.
      • Rigby S.E.J.
      • Leys D.
      • Scrutton N.S.
      Probing the dynamic interface between trimethylamine dehydrogenase (TMADH) and electron transferring flavoprotein (ETF) in the TMADH-2ETF complex: Role of the Arg-α237 (ETF) and Tyr-442 (TMADH) residue pair.
      ,
      • Dwyer T.M.
      • Zhang L.
      • Muller M.
      • Marrugo F.
      • Frerman F.E.
      The functions of the flavin contact residues αArg249 and βTyr16, in human electron transfer flavoprotein.
      ,
      • Watmough N.J.
      • Frerman F.E.
      The electron transfer flavoprotein: Ubiquinone oxidoreductases.
      ). More recently, research has focused on a second clade of ETFs containing two FADs (
      • Sato K.
      • Nishina Y.
      • Shiga K.
      Purification of electron-transferring flavoprotein from Megasphaera elsdenii and binding of additional FAD with an unusual absorption spectrum.
      ) and mediating electron bifurcation (
      • Garcia Costas A.M.
      • Poudel S.
      • Miller A.-F.
      • J S.G.
      • Ledbetter R.N.
      • Fixen K.
      • Seefeldt L.C.
      • Adams M.W.
      • Harwood C.S.
      • Boyd E.S.
      • Peters J.W.
      Defining electron bifurcation in the electron transferring flavoprotein family.
      ,
      • Sato K.
      • Nishina Y.
      • Shiga K.
      Interaction between NADH and electron-transferring flavoprotein from Megasphaera elsdenii.
      ,
      • Herrmann G.
      • Jayamani E.
      • Mai G.
      • Buckel W.
      Energy conservation via electron-transferring flavoprotein in anaerobic bacteria.
      ). In this, NADH donates a pair of electrons to the ETF-mediated bifurcation reaction, and in this case, a client CoA dehydrogenase may be a high-potential acceptor, of one electron. This exothermic electron transfer (ET) reaction is used to drive endothermic transfer of the other electron to a lower potential (more reducing) acceptor: ferredoxin or flavodoxin semiquinone. The Bf ETFs that support nitrogen fixation were named FixAB on this basis (
      • Roberts D.L.
      • Salazar D.
      • Fulmer J.P.
      • Frerman F.E.
      • Kim J.J.
      Crystal structure of paracoccus denitrificans electron transfer flavoprotein: Structural and electrostatic analysis of a conserved flavin binding domain.
      ,
      • Garcia Costas A.M.
      • Poudel S.
      • Miller A.-F.
      • J S.G.
      • Ledbetter R.N.
      • Fixen K.
      • Seefeldt L.C.
      • Adams M.W.
      • Harwood C.S.
      • Boyd E.S.
      • Peters J.W.
      Defining electron bifurcation in the electron transferring flavoprotein family.
      ,
      • Weidenhaupt M.
      • Rossi P.
      • Beck C.
      • Fischer H.-M.
      • Hennecke H.
      Bradyrhizobium japonicum possesses two discrete sets of electron transfer flavoprotein genes:fixA, fixB and etfS, etfL.
      ,
      • Ledbetter R.N.
      • Garcia Costas A.M.
      • Lubner C.E.
      • Mulder D.E.
      • Tokmina-Lukaszewska M.
      • Artz J.H.
      • Patterson A.
      • Magnuson T.S.
      • Jay Z.J.
      • Duan H.D.
      • Miller J.
      • Plunkett M.H.
      • Hoben J.P.
      • Barney B.M.
      • Carlson R.P.
      • et al.
      The electron bifurcating FixABCX protein complex from Azotobacter vinelandii: Generation of low-potential reducing equivalents for nitrogenase catalysis.
      ) and employ a quinone reductase comprised of subunits FixC and FixX as an exergonic acceptor. The FAD that accepts a pair of electrons (2e) from NADH and dispenses them to separate paths is called the bifurcating FAD (Bf FAD, green in Fig. 1), whereas the flavin analogous to that of canonical ETFs is called the ET FAD (yellow in Fig. 1). It is critical that the ET FAD mediates single electron (1e) transfer only, as exergonic transfer of both electrons would dissipate the reducing energy inherent in NADH. The flavin moiety of the Bf FAD is bound between domains I and III while its AMP portion replaces the AMP of canonical ETFs (
      • Chowdhury N.P.
      • Mowafy A.M.
      • Demmer J.K.
      • Upadhyay V.
      • Koelzer S.
      • Jayamani E.
      • Kahnt J.
      • Hornung M.
      • Demmer U.
      • Ermler U.
      • Buckel W.
      Studies on the mechanism of electron bifurcation catalyzed by electron transferring flavoprotein (Etf) and butyryl-CoA dehydrogenase (Bcd) of Acidaminococcus fermentans.
      ).
      Working with the Bf ETF of Rhodopseudomonas palustris (RpaETF), Duan et al. (
      • Duan H.D.
      • Lubner C.E.
      • Tokmina-Lukaszewska M.
      • Gauss G.H.
      • Bothner B.
      • King P.W.
      • Peters J.W.
      • Miller A.F.
      Distinct flavin properties underlie flavin-based electron bifurcation within a novel electron-transferring flavoprotein FixAB from Rhodopseudomonas palustris.
      ) demonstrated that the two FADs satisfy the thermodynamic requirements for electron bifurcation. Specifically, the ET FAD should have two high-potential 1e s suiting it to transfer electrons one at a time between the Bf FAD and the FixCX (Fig. 1). Thus, the ET FAD should be able to adopt a semiquinone (SQ) state and cycle between it and either the oxidized (OX) state or the fully reduced hydroquinone (HQ) state. Indeed, the ET FAD was observed to accumulate in the anionic SQ state (ASQ) part way through reductive titrations (
      • Dwyer T.M.
      • Zhang L.
      • Muller M.
      • Marrugo F.
      • Frerman F.E.
      The functions of the flavin contact residues αArg249 and βTyr16, in human electron transfer flavoprotein.
      ,
      • Sato K.
      • Nishina Y.
      • Shiga K.
      Interaction between NADH and electron-transferring flavoprotein from Megasphaera elsdenii.
      ,
      • Yang K.Y.
      • Swenson R.P.
      Modulation of the redox properties of the flavin cofactor through hydrogen-bonding interactions with the N(5) atom: Role of alpha Ser254 in the electron-transfer flavoprotein from the methylotrophic bacterium W3A1.
      ).
      In contrast, the Bf FAD should have a single 2e couple with a lower , between those of NADH and the ET FAD (
      • Nitschke W.
      • Russell M.J.
      Redox bifurcations: Mechanisms and importance to life now, and at its origin.
      ). This too was documented in Bf ETFs from R. palustris (
      • Duan H.D.
      • Lubner C.E.
      • Tokmina-Lukaszewska M.
      • Gauss G.H.
      • Bothner B.
      • King P.W.
      • Peters J.W.
      • Miller A.F.
      Distinct flavin properties underlie flavin-based electron bifurcation within a novel electron-transferring flavoprotein FixAB from Rhodopseudomonas palustris.
      ), Megasphaera elsdenii (MelETF) (
      • Sato K.
      • Nishina Y.
      • Shiga K.
      Interaction between NADH and electron-transferring flavoprotein from Megasphaera elsdenii.
      ), Pyrobaculum aerophilum (PaeETF) (
      • Schut G.J.
      • Mohamed-Raseek N.R.
      • Tokmina-Lukaszewska M.
      • Mulder D.E.
      • Nguyen D.M.N.
      • Lipscomb G.L.
      • Hoben J.P.
      • Patterson A.
      • Lubner C.E.
      • King P.W.
      • Peters J.W.
      • Bothner B.
      • Miller A.F.
      • Adams M.W.W.
      The catalytic mechanism of electron bifurcating electron transfer flavoproteins (ETFs) involves an intermediary complex with NAD+.
      ) and Acidaminococcus fermentans (AfeETF) (
      • Sucharitakul J.
      • Buttranon S.
      • Wongnate T.
      • Chowdhury N.P.
      • Prongjit M.
      • Buckel W.
      • Chaiyen P.
      Modulations of the reduction potentials of flavin-based electron bifurcation complexes and semiquinone stabilities are key to control directional electron flow.
      ). The structural identities of the two FADs were deduced based on the analogy of one with canonical ETFs' FAD, in conjunction with binding of NADH near the other (
      • Chowdhury N.P.
      • Mowafy A.M.
      • Demmer J.K.
      • Upadhyay V.
      • Koelzer S.
      • Jayamani E.
      • Kahnt J.
      • Hornung M.
      • Demmer U.
      • Ermler U.
      • Buckel W.
      Studies on the mechanism of electron bifurcation catalyzed by electron transferring flavoprotein (Etf) and butyryl-CoA dehydrogenase (Bcd) of Acidaminococcus fermentans.
      ). Experimental tests characterized flavin known to reside in domain II, based on mutagenesis to replace threonine 94 and 97 that hydrogen bond (H-bond) to ribose of the FAD between domains I and III (T94,97A-RpaETF), as well as computation to model each flavin's visible spectrum based on its environment in the protein (
      • Mohamed-Raseek N.
      • Duan H.D.
      • Mroginski M.A.
      • Miller A.F.
      Spectroscopic, thermodynamic and computational evidence of the locations of the FADs in the nitrogen fixation-associated electron transfer flavoprotein.
      ). These site-specific approaches confirmed the structural models in assigning functions to each of the flavins in RpaETF (Fig. 1).
      Given the contrasting reactivities of the two flavins, it was expected that their two binding sites would be different too, as noncovalent interactions with their protein surroundings are believed to modulate the reactivity of bound flavins (
      • Yang K.Y.
      • Swenson R.P.
      Modulation of the redox properties of the flavin cofactor through hydrogen-bonding interactions with the N(5) atom: Role of alpha Ser254 in the electron-transfer flavoprotein from the methylotrophic bacterium W3A1.
      ,
      • Massey V.
      • Hemmerich P.
      Active site probes of flavoproteins.
      ,
      • Lostao A.
      • Gomez-Moreno C.
      • Mayhew S.G.
      • Sancho J.
      Differential stabilization of the three FMN redox forms by tyrosine 94 and tryptophan 57 in flavodoxin from Anabaena and its influence on the redox potentials.
      ,
      • Bradley L.H.
      • Swenson R.P.
      Role of hydrogen bonding interactions to N(3)H of the flavin mononucleotide cofactor in the modulation of the redox potentials of the Clostridium beijerinckii flavodoxin.
      ). There are, indeed, differences. However, from an electrostatic standpoint, both sites are dominated by an Arg residue that is highly conserved. Since the Arg residues of interest both derive from the A subunit of ETF, we omit the chain specifier and name them by their residue numbers alone. Residue 273 in the ET site is conserved as Arg in 202 of 216 sequences of Bf ETFs, and as lysine (Lys) in ten of them, whereas residue 165 in the Bf site is Arg in 214 of 216 Bf ETFs (and Lys in 1) (
      • Garcia Costas A.M.
      • Poudel S.
      • Miller A.-F.
      • J S.G.
      • Ledbetter R.N.
      • Fixen K.
      • Seefeldt L.C.
      • Adams M.W.
      • Harwood C.S.
      • Boyd E.S.
      • Peters J.W.
      Defining electron bifurcation in the electron transferring flavoprotein family.
      ). In each case, the Arg is the only charged residue close enough to interact directly with the flavin.
      In the case of the ET site, one can rationalize the high of the OX/ASQ couple (OX/ASQ) in terms of a favorable interaction between the ASQ formed and the nearby Arg-273 side chain (
      • Dwyer T.M.
      • Zhang L.
      • Muller M.
      • Marrugo F.
      • Frerman F.E.
      The functions of the flavin contact residues αArg249 and βTyr16, in human electron transfer flavoprotein.
      ,
      • Talfournier F.
      • Munro A.W.
      • Basran J.
      • Sutcliffe M.J.
      • Daff S.
      • Chapman S.K.
      • Scrutton N.S.
      Alpha Arg-237 in Methylophilus methylotrophus (sp. W3A1) electron-transferring flavoprotein affords approximately 200-millivolt stabilization of the FAD anionic semiquinone and a kinetic block on full reduction to the dihydroquinone.
      ). However, it is difficult to rationalize the unstable ASQ in Bf site with the nearby side chain of Arg-165. Not only is the ASQ thermodynamically suppressed, in that it is not seen as an intermediate between OX and HQ in reductive titrations, but its instability (high energy) is understood to be crucial to the Bf flavin's ability to reduce a low- acceptor (
      • Nitschke W.
      • Russell M.J.
      Redox bifurcations: Mechanisms and importance to life now, and at its origin.
      ,
      • Peters J.W.
      • Miller A.F.
      • Jones A.K.
      • King P.W.
      • Adams M.W.
      Electron bifurcation.
      ). However, we would expect it to be stabilized by a nearby Arg.
      To elucidate the paradox of contrasting reactivities stemming from two flavins presumed to be chemically identical and both in contact with a conserved Arg, we replaced each Arg and characterized the resulting variants' stability, nucleotide content, and redox reactivity. Ours are the first biophysical characterizations to our knowledge of Bf ETFs incorporating amino acid substitutions in the Bf site. Our strategy rests on the firm foundation provided by variants of canonical ETFs containing substitutions in the ET site (
      • Burgess S.G.
      • Messiha H.L.
      • Katona G.
      • Rigby S.E.J.
      • Leys D.
      • Scrutton N.S.
      Probing the dynamic interface between trimethylamine dehydrogenase (TMADH) and electron transferring flavoprotein (ETF) in the TMADH-2ETF complex: Role of the Arg-α237 (ETF) and Tyr-442 (TMADH) residue pair.
      ,
      • Dwyer T.M.
      • Zhang L.
      • Muller M.
      • Marrugo F.
      • Frerman F.E.
      The functions of the flavin contact residues αArg249 and βTyr16, in human electron transfer flavoprotein.
      ,
      • Talfournier F.
      • Munro A.W.
      • Basran J.
      • Sutcliffe M.J.
      • Daff S.
      • Chapman S.K.
      • Scrutton N.S.
      Alpha Arg-237 in Methylophilus methylotrophus (sp. W3A1) electron-transferring flavoprotein affords approximately 200-millivolt stabilization of the FAD anionic semiquinone and a kinetic block on full reduction to the dihydroquinone.
      ). However, our studies extend the approach to the Arg of the Bf site and demonstrate that the status of each flavin site affects the other. These long-range effects raise the possibility of conformational coupling. Indeed, a remarkable 80° rotation of domain II relative to the domain I⋅III base has been documented crystallographically and is proposed to gate electron transfer in Bf ETFs (
      • Demmer J.K.
      • Bertsch J.
      • Oppinger C.
      • Wohlers H.
      • Kayastha K.
      • Demmer U.
      • Ermler U.
      • Muller V.
      Molecular basis of the flavin-based electron-bifurcating caffeyl-CoA reductase reaction.
      ,
      • Demmer J.K.
      • Chowdhury N.P.
      • Selmer T.
      • Ermler U.
      • Buckel W.
      The semiquinone swing in the bifurcating electron transferring flavoprotein/butyryl-coA dehydrogenase complex from Clostridium difficile.
      ) and engage partner proteins (
      • Toogood H.S.
      • Leys D.
      • Scrutton N.S.
      Dynamics driving function : New insights from electron transferring flavoproteins and partner complexes.
      ). With this in mind, we discuss a mechanism suggested by our finding that mutation of Arg-165 appears to abrogate binding not only of FAD, but even AMP in the Bf site. We propose that Arg-165 establishes a network of H-bonds reaching between domains I and III via a conserved peptide loop that bridges between Arg-165 and the phosphate(s) of AMP/FAD. Such a network stabilizing the dimer interface would explain the conservation of Arg-165 even among non-Bf ETFs, as well as retention of AMP as a stabilizer of the quaternary structure (
      • Sato K.
      • Nishima Y.
      • Shiga K.
      In vitro assembly of FAD, AMP, and the two subunits of electron-transferring flavoprotein: An important role of AMP related with the conformational change of the apoprotein.
      ). It also provides an elegant mechanism by which the oxidation state of the Bf flavin could exert conformational consequences coupling the domain rotation to electron movements.

      Results

      Flavin content and stability of variants

      Considering the ET FAD’s unusually stable ASQ state and precedent in canonical ETFs (
      • Dwyer T.M.
      • Zhang L.
      • Muller M.
      • Marrugo F.
      • Frerman F.E.
      The functions of the flavin contact residues αArg249 and βTyr16, in human electron transfer flavoprotein.
      ,
      • Talfournier F.
      • Munro A.W.
      • Basran J.
      • Sutcliffe M.J.
      • Daff S.
      • Chapman S.K.
      • Scrutton N.S.
      Alpha Arg-237 in Methylophilus methylotrophus (sp. W3A1) electron-transferring flavoprotein affords approximately 200-millivolt stabilization of the FAD anionic semiquinone and a kinetic block on full reduction to the dihydroquinone.
      ), it was inferred that a crucial feature of the ET site would be the conserved Arg-273 that should be positively charged under physiological conditions. However, the Bf ETFs demonstrate that conservation of a nearby Arg does not guarantee a stable ASQ state, as the bifurcating site also contains a conserved Arg, yet suppresses the Bf flavin's SQ states. Thus, the effect appears more nuanced than the identity of the residue, and we expect that the nature of the interaction between Arg and flavin is different in the Bf site than in the ET site. Indeed, Arg-273 is seen in crystal structures to form a π-π stacking interaction with ET FAD (
      • Demmer J.K.
      • Chowdhury N.P.
      • Selmer T.
      • Ermler U.
      • Buckel W.
      The semiquinone swing in the bifurcating electron transferring flavoprotein/butyryl-coA dehydrogenase complex from Clostridium difficile.
      ,
      • Chowdhury N.P.
      • Mowafy A.M.
      • Demmer J.K.
      • Upadhyay V.
      • Koelzer S.
      • Jayamani E.
      • Kahnt J.
      • Hornung M.
      • Demmer U.
      • Ermler U.
      • Buckel W.
      Studies on the mechanism of electron bifurcation catalyzed by electron transferring flavoprotein (Etf) and butyryl-CoA dehydrogenase (Bcd) of Acidaminococcus fermentans.
      ) whereas Arg-165 appears to form bidentate H-bonding interactions with the N5 of Bf FAD and a peptide loop that flanks it (Fig. 1C) (
      • Kayastha K.
      • Vitt S.
      • Buckel W.
      • Ermler U.
      Flavins in the electron bifurcation process.
      ). To test the natures of the interactions, we generated variants, which remove the positive charge (replacement with alanine or glutamine, Ala or Gln) or remove bidentate H-bonding (replacement with Lys). We also sought to make the charge tunable via modulation of pH, by replacing Arg with histidine (His). Amino acid substitutions were made in each of the two flavin-binding sites separately.
      The variants affecting Arg-273 in the ET site were expressed in soluble form and purified. The yields of 20 to 25 mg per liter of culture were comparable to that of the WT indicating similarly stable and soluble proteins. Consistent with this, these variants all contained two FADs per heterodimer after purification including exposure to 1 mM FAD at the time of cell lysis, to repopulate any FAD-binding sites having good affinity (Table 1). Of the variants affecting the Bf site, R165K-RpaETF behaved like the WT. (Throughout, the R165K notation is used to indicate the variant in which Lys replaces Arg at position 165 of the EtfA.) However, variants lacking a cation in the Bf site displayed lowered affinity for FAD. The R165H-RpaETF could be purified with close to one bound FAD, whereas R165Q-RpaETF could only be purified in low yield with less than one FAD bound at pH 8 (below). R165A-RpaETF was expressed but partitioned into the insoluble fraction and produced colorless, sparingly stable protein in minute yield.
      Table 1Flavin contents and thermal stabilities
      RpaETF variantFAD content/dimerTm (°C)
      Tm values were measured on 17 to 30 μM ETF in 20 mM working buffer with 200 mM KCl and 10% (w/v) glycerol at pH 8.
      WT2.00 ± 0.00
      Standard deviations of 0.00 indicate that the two independent trials gave the same value.
      46 ± 2
      R273A2.04 ± 0.0644 ± 0
      R273H2.03 ± 0.0444 ± 0
      R273K2.09 ± 0.0744 ± 0
      R165K2.00 ± 0.0044 ± 0
      R165H0.9 ± 0.140 ± 2
      a Tm values were measured on 17 to 30 μM ETF in 20 mM working buffer with 200 mM KCl and 10% (w/v) glycerol at pH 8.
      b Standard deviations of 0.00 indicate that the two independent trials gave the same value.
      The Tm values marking the midpoints of thermal denaturation at pH 8 displayed the same trends. Protein secondary structure was monitored via far-UV CD as the temperature was raised. All variants studied displayed similar secondary structure content at 25 °C. The variants that were soluble and readily purified all had Tm values only slightly lower than that of WT. The exception was the R165H-RpaETF for which the Tm was lower.

      Absorbance spectral features

      To learn about the environment sensed by the flavins in the modified sites, we compared visible spectra of the variants with those of WT. The absorption bands of the two oxidized flavins overlap almost completely, so for better insight, we exploited the fact that in all variants the ET flavin underwent full reduction before the Bf flavin began to reduce, so a spectrum collected halfway through a reductive titration retained the signature of OX Bf flavin and ET flavin HQ. The difference spectrum from the first half of a titration therefore reflects ET FAD OX minus HQ, and the corresponding difference spectrum for Bf FAD was obtained from the second half of the titration (see Experimental procedures and Fig. S2). We assumed that the two HQ spectra present in the endpoint spectrum are very similar, and in any case they are weak compared with the OX spectra, especially at longer wavelengths. Therefore, we used the end-point spectrum divided by two as a proxy for either flavin's HQ. Addition of this spectrum to the above difference spectra yielded the deduced OX spectra for the ET flavin and Bf flavin. These are compared in Figure 2 and Table 2. The R165H-RpaETF and R165Q-RpaETF variants displayed several distinct properties, so they are discussed together in later sections.
      Figure thumbnail gr2
      Figure 2Comparison of deconvoluted absorbance spectra of ET FAD and Bf FAD in the different variants. Signals deriving from the ET flavin are compared in (A) and those of the Bf flavin in (B). The signals of the Bf flavins display larger separation between band I near 450 nm and band II near 380 nm, as well as more prominent shoulders on band I that are attributed to transitions to and from different vibrational substates of the π and π∗ electronic states involved in the transition (
      • Vdovin A.
      • Slenczka A.
      • Dick B.
      Electronic spectroscopy of lumiflavin in superfluid helium nanodroplets.
      ). Comparisons of spectra in nonpolar solvents ± hydrogen bonding partners demonstrate that stronger longer-lived hydrogen bonds correlate with increased strength for such features (
      • Cerda J.F.
      • Koder Jr., R.L.
      • Lichtenstein B.R.
      • Moser C.M.
      • Miller A.-F.
      • Dutton P.L.
      Hydrogen bond-free flavin redox properties: Managing flavins in extreme aprotic solvents.
      ). Reductive titrations of WT and variants were used to deconvolute the ET and Bf FAD spectra. The deconvolution process is described in the section and . Vertical offsets were applied to facilitate comparisons between spectra. Samples were either 30 μM, or their spectra were scaled to produce amplitude corresponding to 30 μΜ. ET FAD spectra were offset by 0.0, 0.15, 0.25, 0.35, 0.45, 0.0.55 for R165H-, R165K-, R273H-, R273A-, R273K- and WT-RpaETF, respectively. For Bf FAD spectra, vertical offsets were 0.15, 0.3, 0.4, 0.55, for R273H-, R273A-, R273K-, and WT-RpaETF, respectively. Bf, bifurcating; FAD, flavin adenine dinucleotide; RpaETF, ETF of Rhodopseudomonas palustris.
      Table 2Comparison of absorption maxima of individual flavins in each variant, at pH 8
      RpaETF variantET FADBf FAD
      Band I (near 450 nm)

      (±2 nm)
      Band II (near 380 nm)

      (±2 nm)
      Band height ratio II/IBand I (near 450 nm)

      (±2 nm)
      Band II (near 380 nm)

      (±2 nm)
      Band height ratio II/I
      WT4483941.044543740.98
      R273K4503961.064543740.99
      R273A4463921.054543740.99
      R273H4463941.054543740.96
      R165K4483961.074603781.09
      R165H4523861.01N/AN/AN/A
      In all our variants except R165K-RpaETF, the Bf FAD displayed vibrational features in band I, and in all cases, the two bands were well separated, absorbing around 450 and 370 nm, respectively, with comparable band heights, usually favoring band I. On the other hand, for ET FAD, the separation between bands I and II was smaller, producing a shallower dip between them (band II is shifted to 386–392 nm), and the amplitude of band II was higher than or equal to that of band I. Thus the red-shifted band II position of ET FAD in WT-RpaETF (394 versus. free FAD at 374 nm) resembles that documented for T94,97A (394 nm) (
      • Mohamed-Raseek N.
      • Duan H.D.
      • Mroginski M.A.
      • Miller A.F.
      Spectroscopic, thermodynamic and computational evidence of the locations of the FADs in the nitrogen fixation-associated electron transfer flavoprotein.
      ), AfeETF (∼410 nm) (
      • Chowdhury N.P.
      • Mowafy A.M.
      • Demmer J.K.
      • Upadhyay V.
      • Koelzer S.
      • Jayamani E.
      • Kahnt J.
      • Hornung M.
      • Demmer U.
      • Ermler U.
      • Buckel W.
      Studies on the mechanism of electron bifurcation catalyzed by electron transferring flavoprotein (Etf) and butyryl-CoA dehydrogenase (Bcd) of Acidaminococcus fermentans.
      ) and MelETF 410 nm) (
      • Sato K.
      • Nishina Y.
      • Shiga K.
      Purification of electron-transferring flavoprotein from Megasphaera elsdenii and binding of additional FAD with an unusual absorption spectrum.
      ,
      • Vigil J.,W.
      • Niks D.
      • Franz-Badur S.
      • Chowdhury N.
      • Buckel W.
      • Hille R.
      Spectral deconvolution of redox species in the crotonyl-CoA-dependent NADH:ferredoxin oxidoreductase from Megasphaera elsdenii. A flavin-dependent bifurcating enzyme.
      ).

      Effects on the ET flavin

      In response to substitution of Arg-273, the ET flavin's band II was generally more responsive than band I, consistent with the larger change in molecular dipole associated with band II than band I (
      • Stanley R.J.
      Advances in flavin and flavoprotein optical spectroscopy.
      ,
      • Kar R.K.
      • Chasen S.
      • Mroginski M.A.
      • Miller A.F.
      Tuning the quantum chemical properties of flavins via modification at C8.
      ). Replacements of Arg-165 in the Bf site also had negligible effects on ET flavin band I’s absorption maximum, but absence of Bf FAD and replacement of Arg-165 with His caused the ET flavin's band II to shift significantly in response to this remote change.
      The ET flavin's responsiveness to substitutions in either site, including the published T94,97A double substitution, can be rationalized as a consequence of the ET flavin being bound to a mobile head domain that brings the ET flavin close to the base domain in one conformation (“closed” or B-like), but exposes it more in another (“open” or D, Fig. S1) (
      • Toogood H.S.
      • Leys D.
      • Scrutton N.S.
      Dynamics driving function : New insights from electron transferring flavoproteins and partner complexes.
      ,
      • Demmer J.K.
      • Chowdhury N.P.
      • Selmer T.
      • Ermler U.
      • Buckel W.
      The semiquinone swing in the bifurcating electron transferring flavoprotein/butyryl-coA dehydrogenase complex from Clostridium difficile.
      ). Thus, the environment of the ET flavin is expected to be affected not only by perturbations that alter the local environment, but also by perturbations of the conformational equilibrium. Our observation of ET flavin spectral changes in response to perturbation of the Bf site is thus consistent with events in the Bf site being coupled to the conformational equilibrium.

      Effects on the Bf flavin

      The absorption maxima of the Bf flavin both shifted to slightly longer wavelengths in response to replacement of Arg-165 by Lys, despite the retention of positive charge. The 6 nm and 4 nm shifts (band I and band II, respectively) suggest that the interaction between the Arg side chain and the flavin includes mechanisms additional to electrostatics. Indeed, the vibronic structure characterizing the Bf flavin’s band I was lost in the R165K-RpaETF variant, but not variants perturbed in the ET site (see Fig. 2 and caption). Because the vibrational structure is a signature of H-bonding (
      • Yagi K.
      • Ohishi N.
      • Nishimoto K.
      • Choi J.D.
      • Song P.-S.
      Effect of hydrogen bonding on electronic spectra and reactivity of flavins.
      ,
      • Cerda J.F.
      • Koder Jr., R.L.
      • Lichtenstein B.R.
      • Moser C.M.
      • Miller A.-F.
      • Dutton P.L.
      Hydrogen bond-free flavin redox properties: Managing flavins in extreme aprotic solvents.
      ), its loss in R165K-RpaETF, specifically, suggests that in WT-RpaETF the flavin H-bonds with Arg-165. Bf FAD's greater sensitivity to local perturbation than displayed by ET flavin is consistent with the lower solvent exposure of the Bf flavin (see Table 2).

      Accumulation of ET FAD ASQ in the different ETF variants

      Whereas free FMN in water forms the SQ state in only 1% of the population when the populations of OX and HQ are equal (system is half reduced) (
      • Mayhew S.G.
      The effects of pH and semiquinone formation on the oxidation-reduction potentials of flavin mononucleotide: A reappraisal.
      ) the ET site of ETF is remarkable in that almost the entire population assumes the ASQ state when this flavin is half reduced (Fig. 3) (
      • Yang K.Y.
      • Swenson R.P.
      Modulation of the redox properties of the flavin cofactor through hydrogen-bonding interactions with the N(5) atom: Role of alpha Ser254 in the electron-transfer flavoprotein from the methylotrophic bacterium W3A1.
      ,
      • Talfournier F.
      • Munro A.W.
      • Basran J.
      • Sutcliffe M.J.
      • Daff S.
      • Chapman S.K.
      • Scrutton N.S.
      Alpha Arg-237 in Methylophilus methylotrophus (sp. W3A1) electron-transferring flavoprotein affords approximately 200-millivolt stabilization of the FAD anionic semiquinone and a kinetic block on full reduction to the dihydroquinone.
      ). This is critical to the bifurcation mechanism, as it disfavors pairwise transfer of both electrons from NADH to the ET flavin. To test for retention of this property in the Arg variants, we reduced them, stepwise, with sodium dithionite (NaDT), Ti(III)-citrate, or the xanthine/xanthine oxidase system (
      • Massey V.
      A simple method for determination of redox potentials.
      ) and measured the maximum extent of ASQ formation between phases 1 and 2 of the reduction.
      Figure thumbnail gr3
      Figure 3Three phases of stepwise reduction of R165K-RpaETF with Ti(III) citrate. Visible spectra observed at each step of reduction, with the first spectrum of fully oxidized R165K in purple, the spectrum at the end of phase 1 in bright green (maximum population of ASQ), the spectrum at the end of phase 2 in light orange and the fully reduced endpoint after phase 3 in dark brown. Inset shows a plot of the A374 versus. A456, revealing the three different slopes that distinguish the three phases of reduction. The first spectrum gives rise to the point furthest to the right and the point corresponding to the last spectrum is in the lower left. In all phases A456 decreases, but in phase 1, growth of ASQ more than compensates for loss of OX at A374. In phase 2 the loss of this stronger absorbance makes A374 drop steeply, whereas in phase 3 A374 reflects loss of the weaker OX signal and its replacement by weaker still HQ. Data for a 22 μM sample are shown. ASQ, anionic semiquinone; HQ, hydroquinone; RpaETF, ETF of Rhodopseudomonas palustris.
      Reduction of all R273-RpaETF variants and R165K-RpaETF showed a triphasic reduction (Fig. 3). As in previous work on RpaETF (
      • Duan H.D.
      • Lubner C.E.
      • Tokmina-Lukaszewska M.
      • Gauss G.H.
      • Bothner B.
      • King P.W.
      • Peters J.W.
      • Miller A.F.
      Distinct flavin properties underlie flavin-based electron bifurcation within a novel electron-transferring flavoprotein FixAB from Rhodopseudomonas palustris.
      ), MelETF (
      • Sato K.
      • Nishina Y.
      • Shiga K.
      Interaction between NADH and electron-transferring flavoprotein from Megasphaera elsdenii.
      ) and AfeETF (
      • Sucharitakul J.
      • Buttranon S.
      • Wongnate T.
      • Chowdhury N.P.
      • Prongjit M.
      • Buckel W.
      • Chaiyen P.
      Modulations of the reduction potentials of flavin-based electron bifurcation complexes and semiquinone stabilities are key to control directional electron flow.
      ), reduction first generated an ASQ as indicated by increased absorbance near 374 nm in conjunction with decreased absorbance at 454 nm and an isosbestic around 390 nm. In the second phase, absorbance intensity decreased especially at 374 nm but also at 454 nm indicating formation HQ from ASQ. These phases represent two sequential 1e reductions of a higher- FAD (ET FAD). Further deflation, in a third phase affecting A454 more than A374, revealed 2e reduction of the lower- FAD (Bf FAD). The absence of detectable SQ at the midpoint of this phase indicates that the Bf FAD has crossed potentials and a suppressed SQ (
      • Nitschke W.
      • Russell M.J.
      Redox bifurcations: Mechanisms and importance to life now, and at its origin.
      ).
      The R273-RpaETF variants displayed differences in the maximum amount of ASQ accumulated between phases 1 and 2. The extent of ASQ formation was determined from the slope of the plot of A374 versus. A454 (inset to Fig. 3). The slope early in phase 1 (right edge of inset in Fig. 3) reveals the extents to which 1e and 2e reduction contribute to consumption of OX. A second estimate of the fraction of sites undergoing 1e reduction was obtained from the slope late in phase 2. Obtained slopes were compared with slopes derived from authentic cases of 1e or 2e reduction of OX flavin, or 1e or 2e formation of HQ flavin, to obtain the extent to which each process contributed. (Details are in supporting information.) Based on the two values obtained in each case, Table 3 summarizes the extent to which 1e ET occurs, expressed as maximum population of ASQ formed ASQmax, in each of the variants.
      Table 3Maximum population of ASQ in different variants
      RpaETF variantASQmaxKsqE OX/ASQ - E ASQ/HQ (mV)
      uncertainties are ≤5 mV, See Table S1.
      WT0.84 ± 0.0
      Reference value, calculated from the separation between E°s determined here.
      ,
      Two independent repetitions.
      110122
      R273K0.84 ±0.0
      Two independent repetitions.
      110127
      R273A0.52±0.03
      Two independent repetitions.
      4.6940
      R273H0.71 ±0.00
      Both sets of slopes produced the same values.
      23.9882
      R165K0.84 ± 0.0
      Two independent repetitions.
      110125
      T9497A0.74 ± 0.0
      Two independent repetitions.
      32.489
      a uncertainties are ≤5 mV, See Table S1.
      b Reference value, calculated from the separation between s determined here.
      c Two independent repetitions.
      d Both sets of slopes produced the same values.
      Our analysis reveals maximal ASQ accumulation varying from 0.84 (almost all sites) to 0.52. Since an ASQmax of 1/3 is expected when the two 1e s are equal (see supporting information), our variants all retain uncrossed potentials (OX/ASQ > ASQ/AHQ). Thus, interactions additional to those of Arg-273 must participate in converting the ET flavin from being predominantly a 2e site (as in free flavin, with crossed s) to a 1e site (uncrossed s). Nevertheless, the role of Arg-273 is large, as replacement of Arg by Ala almost halves the amount of ASQ that accumulates.
      Replacement of Arg-273 with Lys had no effect on ASQ yield, confirming that the presence of a positive charge is paramount. Among R273A-RpaETF and R273H-RpaETF, the R273H-RpaETF variant better stabilizes ASQ, possibly acquiring a proton in some fraction of sites in conjunction with flavin reduction. Nevertheless, lowering the pH did not detectably increase the yield of ASQ for R273H-RpaETF. Similarly, stepwise reduction of R273H-RpaETF at pH 7 did not result in more ASQ, or visible neutral SQ formation, arguing against proton acquisition by the flavin either. Thus, we have no evidence for outright protonation of His-273 or the flavin in this system and therefore speculate that the His-273 might replicate H-bonding or π stacking between Arg-273 and flavin ASQ that does not occur in R273A-RpaETF. It is also possible that R273A-RpaETF and R273H-RpaETF have different propensities for the two conformations of ETF, thereby explaining the different amounts of ASQ on the basis of a large-scale factor rather than interactions with the flavin. Either of these explanations is consistent with Sucharitakul's report that binding to partner protein eliminated thermodynamic population of ASQ at the ET site (
      • Sucharitakul J.
      • Buttranon S.
      • Wongnate T.
      • Chowdhury N.P.
      • Prongjit M.
      • Buckel W.
      • Chaiyen P.
      Modulations of the reduction potentials of flavin-based electron bifurcation complexes and semiquinone stabilities are key to control directional electron flow.
      ).
      Replacement of Arg-165 in the Bf site with Lys did not alter the 1e versus 2e reactivity of the ET flavin; however, absence of the Bf flavin from T94,97A-RpaETF slightly diminished the dominance of 1e reactivity. This small drop in stabilization of ASQ is nevertheless interesting, as it provides a second instance of remote effects of the Bf site on the ET flavin.

      Determination of reduction midpoint potentials

      To understand whether diminished yields of ASQ reflect less favorable formation of ASQ, or more favorable consumption, we measured individual s. Given that several of the variants accumulated significantly substoichiometric amounts of ASQ, the potentials halfway through phases 1 and 2 are not the individual 1e s, but their average is still the 2e (
      • Clark W.M.
      Oxidation-Reduction Potentials of Organic Systems.
      ). For each variant, we used the average value of the half-way points, and the separation between 1e s calculated from the ASQ yield, to obtain the individual 1e s (see Experimental procedures).
      Phase 1 reduction of WT-, R273K-, R273A-, or R273H-RpaETF was monitored at 454 nm, that of R165K-RpaETF was monitored at 458 nm. The total change associated with phase 1 was initially treated as representing reduction of a full equivalent of ET flavin from OX to ASQ to calculate the log([FAD]OX/[FAD]ASQ), for plotting versus log([DYE]OX/[DYE]Red) (Fig. 4). The resulting log/log plot revealed slopes of 0.5 as anticipated for 1e reductions, consistent with the mathematical treatment used (Experimental procedures and Table S1). The intercepts were then used to obtain potentials corresponding to the half-way points of this first reduction (
      • Mohamed-Raseek N.
      • Duan H.D.
      • Mroginski M.A.
      • Miller A.F.
      Spectroscopic, thermodynamic and computational evidence of the locations of the FADs in the nitrogen fixation-associated electron transfer flavoprotein.
      ). Analogous procedures yielded the half-way points of the reduction of ASQ to HQ during coreduction with Nile blue (Table S1) based on log/log plot slopes of 0.5 indicative of 1e transfer reactions in all cases. The two half-way points were averaged to obtain the average of the two 1e s (Table S1).
      Figure thumbnail gr4
      Figure 4Log/log plots of populations present during coreduction of ETF and dye in reductive titration of R165K-RpaETF. Log([oxidized FAD]/[reduced FAD]) for the flavins of R165K-RpaETF is plotted versus. the analogous quantity for the dye, showing linear behavior in each of the three phases of reduction, consistent with the Nernst equation. The slopes near 0.5 for phases 1 and 2 are consistent with 1e events and the slope near 1.0 for phase 3 indicates a 2e event, because all the dyes used underwent 2e reduction, see . FAD, flavin adenine dinucleotide; RpaETF, ETF of Rhodopseudomonas palustris.
      Using the average of the 2 s, as well as the separation between them calculated from the maximal yields of ASQ (Table 3), we calculated values for OX/ASQ and ASQ/HQ at pH 8 for each of the variants (Table 4). Our data indicate no significant difference between the OX/ASQ values of R273K-RpaETF and R165K-RpaETF, but they are 6 to 8 mV lower than the value obtained for WT. In contrast, introduction of Ala or His in place of the Arg-273 in the ET site dramatically decreased the EOX/ASQ. We surmise that this reflects loss of residue 273's positive charge in conjunction with the fact that the OX/ASQ couple creates anionic flavin, because there was little effect on the ASQ/HQ couples, which do not change the charge of the flavin.
      Table 4Effects on E° of replacing the arginine residues at pH 8
      RpaETF variantFlavin contentE OX/ASQE ASQ/HQE OX/HQ (mV)
      WT2−61± 3 a−182 ± 2
      Determined by coreduction with Nile blue, NB (E°’ = −116 mV, 2e/H+ at pH 7, calculated E° = −146 mV pH 8) (19, 61). λmax of NB is at 636 nm.
      −273 ± 5 (n=2)
      Determined by coreduction with phenosafranin (PS, E°’ = −252 mV, 2e/H+ at pH 7, calculated E° = −282 mV pH 8) (19, 61).
      R273K2−69 ± 4
      EOX/ASQ was measured by coreduction with new methylene blue (NMB) with the calculated potential of −51 mV (E°’ = −21 mV, 2e/H+ at pH 7) (66) and λmax of NMB is at 632 nm.
      −185 ± 5
      Determined by coreduction with Nile blue, NB (E°’ = −116 mV, 2e/H+ at pH 7, calculated E° = −146 mV pH 8) (19, 61). λmax of NB is at 636 nm.
      −273 ± 5 (n=2)
      Determined by coreduction with phenosafranin (PS, E°’ = −252 mV, 2e/H+ at pH 7, calculated E° = −282 mV pH 8) (19, 61).
      R273H2−119 ± 2
      Determined by coreduction with resorufin, RS (E°’ = −51 mV, 2e/2H+ at pH 7, calculated E° = −111 mV pH 8) (67, 68) using A572 nm for RS.
      −201 ± 2
      Determined by coreduction with Nile blue, NB (E°’ = −116 mV, 2e/H+ at pH 7, calculated E° = −146 mV pH 8) (19, 61). λmax of NB is at 636 nm.
      −310 ± 2 (n=2)
      Safranin O, E°’ = −289 mV, 2e/H+ at pH 7, calculated E° = −319 mV pH 8) (22, 61).
      R273A2−137
      Determined by coreduction with resorufin, RS (E°’ = −51 mV, 2e/2H+ at pH 7, calculated E° = −111 mV pH 8) (67, 68) using A572 nm for RS.
      −178 ± 2
      Determined by coreduction with Nile blue, NB (E°’ = −116 mV, 2e/H+ at pH 7, calculated E° = −146 mV pH 8) (19, 61). λmax of NB is at 636 nm.
      −313 ± 2 (n=2)
      Safranin O, E°’ = −289 mV, 2e/H+ at pH 7, calculated E° = −319 mV pH 8) (22, 61).
      R165K2−67.6
      EOX/ASQ was measured by coreduction with new methylene blue (NMB) with the calculated potential of −51 mV (E°’ = −21 mV, 2e/H+ at pH 7) (66) and λmax of NMB is at 632 nm.
      −182 ± 2
      Determined by coreduction with Nile blue, NB (E°’ = −116 mV, 2e/H+ at pH 7, calculated E° = −146 mV pH 8) (19, 61). λmax of NB is at 636 nm.
      −271 ± 5 (n=2)
      Determined by coreduction with phenosafranin (PS, E°’ = −252 mV, 2e/H+ at pH 7, calculated E° = −282 mV pH 8) (19, 61).
      a Determined by coreduction with Nile blue, NB (E°’ = −116 mV, 2e/H+ at pH 7, calculated  = −146 mV pH 8) (
      • Duan H.D.
      • Lubner C.E.
      • Tokmina-Lukaszewska M.
      • Gauss G.H.
      • Bothner B.
      • King P.W.
      • Peters J.W.
      • Miller A.F.
      Distinct flavin properties underlie flavin-based electron bifurcation within a novel electron-transferring flavoprotein FixAB from Rhodopseudomonas palustris.
      ,
      • Miller A.F.
      • Duan H.D.
      • Varner T.A.
      • Mohamed-Raseek N.
      Reduction midpoint potentials of bifurcating electron transfer flavoproteins.
      ). λmax of NB is at 636 nm.
      b Determined by coreduction with phenosafranin (PS, E°’ = −252 mV, 2e/H+ at pH 7, calculated  = −282 mV pH 8) (
      • Duan H.D.
      • Lubner C.E.
      • Tokmina-Lukaszewska M.
      • Gauss G.H.
      • Bothner B.
      • King P.W.
      • Peters J.W.
      • Miller A.F.
      Distinct flavin properties underlie flavin-based electron bifurcation within a novel electron-transferring flavoprotein FixAB from Rhodopseudomonas palustris.
      ,
      • Miller A.F.
      • Duan H.D.
      • Varner T.A.
      • Mohamed-Raseek N.
      Reduction midpoint potentials of bifurcating electron transfer flavoproteins.
      ).
      c EOX/ASQ was measured by coreduction with new methylene blue (NMB) with the calculated potential of −51 mV (E°’ = −21 mV, 2e/H+ at pH 7) (
      • Lei C.
      • Zhang Z.
      • Liu H.
      • Kong J.
      • Deng J.
      Biosensoring of hydrogen peroxide using new methylene blue N incorporated in a montmorillonite-modified horseradish peroxidase immobilization matrix as an electron shuttle.
      ) and λmax of NMB is at 632 nm.
      d Determined by coreduction with resorufin, RS (E°’ = −51 mV, 2e/2H+ at pH 7, calculated  = −111 mV pH 8) (
      • Oja S.M.
      • Guerrette J.P.
      • David M.R.
      • Zhang B.
      Fluorescence-enabled electrochemical microscopy with dihydroresorufin as a fluorogenic indicator.
      ,
      • Twigg R.S.
      Oxidation-reduction aspects of resazurin.
      ) using A572 nm for RS.
      e Safranin O, E°’ = −289 mV, 2e/H+ at pH 7, calculated  = −319 mV pH 8) (
      • Schut G.J.
      • Mohamed-Raseek N.R.
      • Tokmina-Lukaszewska M.
      • Mulder D.E.
      • Nguyen D.M.N.
      • Lipscomb G.L.
      • Hoben J.P.
      • Patterson A.
      • Lubner C.E.
      • King P.W.
      • Peters J.W.
      • Bothner B.
      • Miller A.F.
      • Adams M.W.W.
      The catalytic mechanism of electron bifurcating electron transfer flavoproteins (ETFs) involves an intermediary complex with NAD+.
      ,
      • Miller A.F.
      • Duan H.D.
      • Varner T.A.
      • Mohamed-Raseek N.
      Reduction midpoint potentials of bifurcating electron transfer flavoproteins.
      ).
      Whereas replacements of Arg-273 could affect the ET flavin via local effects as well as large-scale conformational change, the Bf flavin would be affected only by conformational change. To determine the OX/HQ of Bf FAD, we again used the xanthine/xanthine oxidase system, to drive coreduction of the Bf flavin with the reference dye safranin O (SO) for R273A-RpaETF and R273H-RpaETF, or phenosafranin (PS) for WT-RpaETF, R273K-RpaETF, and R165K-RpaETF, in the third phase of reduction (Fig. 3, left-hand side of inset). The slopes obtained from the log/log plots were 1.0 (Fig. 4 and Table S1) and calculated potentials are in Table 4. The Bf OX/HQ couples of R273A-RpaETF and R273H-RpaETF were 40 mV lower than those of WT-RpaETF, R273K-RpaETF, and R165K-RpaETF. Thus, the R273A and R273H substitutions in the ET site exert similar long-range effects on the Bf flavin, possibly by altering the conformational equilibrium.

      A possible role for kinetics

      Canonical ETF from W3A1 and MmeETF form ASQ but resist further reduction to the HQ state (
      • Yang K.Y.
      • Swenson R.P.
      Modulation of the redox properties of the flavin cofactor through hydrogen-bonding interactions with the N(5) atom: Role of alpha Ser254 in the electron-transfer flavoprotein from the methylotrophic bacterium W3A1.
      ,
      • Talfournier F.
      • Munro A.W.
      • Basran J.
      • Sutcliffe M.J.
      • Daff S.
      • Chapman S.K.
      • Scrutton N.S.
      Alpha Arg-237 in Methylophilus methylotrophus (sp. W3A1) electron-transferring flavoprotein affords approximately 200-millivolt stabilization of the FAD anionic semiquinone and a kinetic block on full reduction to the dihydroquinone.
      ). This tends to magnify the amount of ASQ observed (
      • Burgess S.G.
      • Messiha H.L.
      • Katona G.
      • Rigby S.E.J.
      • Leys D.
      • Scrutton N.S.
      Probing the dynamic interface between trimethylamine dehydrogenase (TMADH) and electron transferring flavoprotein (ETF) in the TMADH-2ETF complex: Role of the Arg-α237 (ETF) and Tyr-442 (TMADH) residue pair.
      ). In our case, since substoichiometric ASQ was observed, we were instead concerned that formation of ASQ might be impeded, despite our use of slow reductions via the xanthine/XO system. To directly probe the possibility that our diminished ASQ yields reflect a kinetic barrier to formation of ASQ in R273A-RpaETF and R273H-RpaETF, we added an amount of dithionite calculated to yield maximal ASQ and monitored the system for 2 to 3 h. Accumulation of ASQ was complete in 2 min, after which no further accumulation of ASQ was observed. This argues against kinetic suppression of ASQ formation and instead indicates that removal of Arg-273’s positive charge at the ET site diminishes the amount of ASQ formed via thermodynamic effects.

      Behavior specific to the R165H-RpaETF

      Unexpectedly, the R165H-RpaETF displayed broad absorbance from 500 to 750 nm (Fig. 2). The rest of the spectrum resembled the ET FAD spectrum more than that of the Bf FAD, on the basis of its shallow dip between bands I and II and its λmax value for band II that is closer to the larger value typical of ET flavin than the shorter value of Bf flavin. (R165H-RpaETF's band II λmax is 6 nm shorter than the shortest ET flavin value but 8 nm longer than the longest Bf flavin value, Table 2). The long wavelength absorbance was present before reconstitution with FAD, when only one flavin was present (see below), so a single FAD is responsible, and it is the one that copurifies in this variant. Normally these criteria would identify it as the more tightly bound Bf flavin, but the fact that we had modified the Bf site with the R165H substitution cast this assumption in doubt.

      Identity of the FAD retained by R165H-RpaETF

      The long wavelength and breadth of the new absorbance are suggestive of a charge-transfer (CT) interaction, for example, with a nearby amino acid side chain or substrate analog (
      • Massey V.
      • Palmer A.G.
      Charge transfer complexes of lipoyl dehydrogenase and free flavins.
      ,
      • He T.-F.
      • Guo L.
      • Guo X.
      • Chang C.-W.
      • Wang L.
      • Zhong D.
      Femtosecond dynamics of short-range protein electron transfer in flavodoxin.
      ,
      • Abramovitz A.S.
      • Massey V.
      Interaction of phenols with old yellow enzyme. Physical evidence for charge-transfer complexes.
      ). Since the band is not present in the WT, the introduced His-165 is a candidate participant. However, CT with His-165 would require that this variant retain FAD in the Bf site, even though the substitution in the Bf site seems more likely to disrupt Bf flavin binding than ET flavin binding. An alternative interpretation is that the long-wavelength band represents a flavin semiquinone (SQ). This would allow that the retained flavin resides in the ET site, but seems surprising in these air-equilibrated samples.
      To learn which flavin is retained in R165H-RpaETF, we exploited the distinct visible CD of the two flavins. The visible CD of R165H-RpaETF clearly contained the signature of the ET FAD at short wavelength, but also possessed a weak version of the Bf FAD's signature negative CD at longer wavelengths (425–500 nm) (Fig. S3). The simplest interpretation is that considerable ET FAD is present, but some Bf FAD accompanies it. However, the long-wavelength CD could also reflect perturbation of the ET flavin’s environment, even though the substitution is located in the other domain of the ETF. A similar observation has been reported for human ETF, where mutation of Tyr-16 to Ala in domain III significantly changed the longer wavelength flavin CD for FAD bound in domain II, consistent with Tyr-16 reaching between domains (
      • Dwyer T.M.
      • Zhang L.
      • Muller M.
      • Marrugo F.
      • Frerman F.E.
      The functions of the flavin contact residues αArg249 and βTyr16, in human electron transfer flavoprotein.
      ).
      As an additional test of the presence of Bf FAD in R165H-RpaETF, it was titrated with NADH, because NADH interacts with the Bf FAD site to reduce the system, whereas ETFs lacking Bf flavin are not subject to reduction by NADH (
      • Sato K.
      • Nishina Y.
      • Shiga K.
      Interaction between NADH and electron-transferring flavoprotein from Megasphaera elsdenii.
      ,
      • Chowdhury N.P.
      • Mowafy A.M.
      • Demmer J.K.
      • Upadhyay V.
      • Koelzer S.
      • Jayamani E.
      • Kahnt J.
      • Hornung M.
      • Demmer U.
      • Ermler U.
      • Buckel W.
      Studies on the mechanism of electron bifurcation catalyzed by electron transferring flavoprotein (Etf) and butyryl-CoA dehydrogenase (Bcd) of Acidaminococcus fermentans.
      ,
      • Mohamed-Raseek N.
      • Duan H.D.
      • Mroginski M.A.
      • Miller A.F.
      Spectroscopic, thermodynamic and computational evidence of the locations of the FADs in the nitrogen fixation-associated electron transfer flavoprotein.
      ). Upon addition of NADH, R165H-RpaETF was reduced in a 2e manner initially, but after approximately 10% of the flavin was converted to HQ, the rest of the flavin was unreactive (Fig. 5B). Because the long wavelength band did not gain strength in conjunction with reduction of Bf flavin with NADH, we can rule out a CT band between Bf flavin HQ and NAD+. Full reduction was nevertheless achieved by sodium dithionite, but displayed predominantly two phases, corresponding to phases 1 and 2 based on a plot of A374 versus. A450 (inset of Fig. 5A). Thus, some 10% of Bf sites contain FAD, but some 90% of the flavin present is bound in the ET site based on the fraction of absorbance at 454 nm susceptible to reduction by NADH.
      Figure thumbnail gr5
      Figure 5Reductive titrations of R165H with NADH or dithionite. A, Spectra observed in the course of reduction with sodium dithionite (DT). B, spectra observed during reduction by NADH. The inset in A provides the absorbance at 374 nm versus. that at 450 nm. A374 reports on ASQ formation and then consumption as the ETF is reduced (see also inset to ). ASQ, anionic semiquinone.
      The reductive titrations also assign the broad long-wavelength band to the ET flavin, as the band was diminished by reduction with dithionite in conjunction with reduction of the ASQ of the ET flavin, in phase 2. Thus, we propose that it reflects the SQ state of a modified flavin, since it is not observed in WT-RpaETF. The ASQ state of 8-formyl flavin accumulates in air (
      • Augustin P.
      • Toplak M.
      • Fuchs K.
      • Gerstmann E.C.
      • Prassl R.
      • Winkler A.
      • Macheroux P.
      Oxidation of the FAD cofactor to the 8-formyl-derivative in human electron-transferring flavoprotein.
      ) or illumination (
      • Robbins J.M.
      • Geng J.
      • Barry B.A.
      • Gadda G.
      • Bommarius A.S.
      Photoirradiation generates an ultrastable 8-formyl FAD semiquinone radical with unusual properties in formate oxidase.
      ), so it can be present in the starting spectrum.

      The ET flavin undergoes chemical modification in the R165Q-RpaETF and R165H-RpaETF

      The spectrum of R165H-RpaETF can be better understood with reference to that of R165Q-RpaETF (Fig. 6A). Both of these variants possess substoichiometric flavin and both have similar long-wavelength bands. Thus, by extension of our finding for R165H-RpaETF, we understand that R165Q-RpaETF contains predominantly the ET flavin. The spectra of R165Q-RpaETF differ significantly from the ET flavin spectra of the other variants (Fig. 2), but bear resemblance to the ET FAD spectrum of T94,97A-RpaETF at pH 9 (Fig. 6A). Both T94,97A-RpaETF at pH 9 and R165Q-RpaETF display the distinctive signature of 8-formyl flavin semiquinone, as documented by Macheroux' team for human ETF (see Fig. 4B in their paper (
      • Augustin P.
      • Toplak M.
      • Fuchs K.
      • Gerstmann E.C.
      • Prassl R.
      • Winkler A.
      • Macheroux P.
      Oxidation of the FAD cofactor to the 8-formyl-derivative in human electron-transferring flavoprotein.
      )). Indeed, mass spectrometric analysis of our T94,97A-, R156H-, and R165Q-RpaETF reveals presence of 8-formyl flavin, and EPR detects a radical consistent with 8-formyl flavin ASQ in R165Q-RpaETF.
      Figure thumbnail gr6
      Figure 6Spectra of ETF R165-RpaETF variants and their released flavins. Spectra scaled to correspond to flavin concentrations of 30 μM and compared for (A) ETF variants and (B) cofactors released from them. R165H- and R165Q-RpaETF purified and maintained at pH 8 are compared with T94,97A-RpaETF purified and maintained at pH 9, revealing a sharp feature near 425 nm. This is attributed to the flavin itself rather than the environment since displacement and replacement with fresh FAD greatly diminishes it ('reconstitution'). Nevertheless, all spectra retain absorbance at long wavelengths. Spectra of released flavins (B) confirm that the flavin itself sustains chemical modification in R165Q-RpaETF at pH 8, similar to that seen in T94,97A-RpaETF at pH 9. RpaETF, ETF of Rhodopseudomonas palustris.
      Fractional production of 8-formyl flavin can also explain the as-isolated spectrum of R165H-RpaETF at pH 8 (before incubation with FAD, Fig. 6A). This R165H displays a small lump at 425 nm where WT-RpaETF has a dip between the two flavin bands but 8-formyl flavin ASQ has a sharp peak. The same feature is also seen in R165Q-RpaETF at pH 8 and T94,97A-RpaETF at pH 9. 8-formyl flavin ASQ furthermore explains R165H-RpaETF's broad absorbance from 500 to 700 nm, also prominently displayed by R165Q-RpaETF (compare with Figs. 4 and 6 of Augustin et al. (
      • Augustin P.
      • Toplak M.
      • Fuchs K.
      • Gerstmann E.C.
      • Prassl R.
      • Winkler A.
      • Macheroux P.
      Oxidation of the FAD cofactor to the 8-formyl-derivative in human electron-transferring flavoprotein.
      ). The high OX/ASQ of 8-formyl flavin explains the accumulation of this species even in air-equilibrated samples (
      • Augustin P.
      • Toplak M.
      • Fuchs K.
      • Gerstmann E.C.
      • Prassl R.
      • Winkler A.
      • Macheroux P.
      Oxidation of the FAD cofactor to the 8-formyl-derivative in human electron-transferring flavoprotein.
      ,
      • Robbins J.M.
      • Souffrant M.G.
      • Hamelberg D.
      • Gadda G.
      • Bommarius A.S.
      Enzyme-mediated conversion of flavin adenine dinucleotide (FAD) to 8-formyl FAD in formate oxidase results in a modified cofactor with enhanced catalytic properties.
      ). The spectra of the released flavins also bear the signatures of modified flavin, in which band I is shifted to a longer wavelength of 462 nm in flavin released from R165Q (Fig. 6B). However, we note that the amplitude of band II relative to band I deviates from that expected from pure 8-formyl flavin, indicating that modifications other than or additional to oxidation of the 8-methyl may also be present. Additional work is required to sort out this complicated system, but data in hand demonstrate that the long-wavelength band in R165H-RpaETF and R165Q-RpaETF, as well as the unusual spectrum of R165Q-RpaETF, reflects modification of the ET flavin, substantially to 8-formyl flavin. This appears to affect only a minority of sites in the case R165H-RpaETF (in which spectra also retain the signature of unmodified flavin), but a majority for R165Q-RpaETF. Thus, it appears that a positively charged residue 165 in the Bf site or presence of the Bf FAD has the long-range effect of suppressing modification of the ET flavin.
      Formation of 8-formyl flavin is well-known in ETFs which lack the Bf FAD, based on occurrence in human ETF, and T94,97A-RpaETF at pH 9. Yet these variants retain a positive residue at position 165. Thus, we propose that presence of modified ET flavin is a consequence of the lack of Bf flavin in R165H-RpaETF and R165Q-RpaETF, and that the primary effect of the R165H and R165Q substitutions is to disfavor acquisition of Bf FAD. Nevertheless, accumulation of modified flavin is faster in R165Q-RpaETF than in T94,97A-RpaETF, suggesting that positive charge at position 165 also plays a role.

      The Bf site responds to pH changes

      Since Arg or Lys at position 165 enables full FAD content, it appears that His and Gln's lack of positive charge and/or their shorter nonpolar stalks is/are responsible for poor FAD binding. To distinguish between these, we examined the pH dependence of FAD content in the R165H-RpaETF as a means of modulating His-273's charge. At pH 8, only one FAD was bound, whereas at pH 9, 1.6 FADs were bound per ETF (Table 5). This is opposite to the behavior one would expect if protonation of His-165 and acquisition of positive charge were required to mediate FAD binding in the Bf site. However, it argues that the shorter stalk of His compared with Lys or Arg is less important.
      Table 5pH dependence of flavin occupancy in Bf site variants
      RpaETF VariantFAD content (per dimer, at pH 8)AMP content (per dimer, at pH 8)FAD content (per dimer, at pH 9)
      WT20.02
      R165K20.02
      R165H0.9 ± 0.10.2 ± 0.01.6
      R165Q0.26 ± 0.000.8 ± 0.21
      The pH dependence of FAD content was also tested for the R165Q-RpaETF to determine whether it pertains to His-165 or something else. Interestingly, even R165Q-RpaETF contains more FAD at pH 9 than at pH 8. However, both variants remain less stable than WT-RpaETF or R165K-RpaETF even at pH 9, consistent with their substoichiometric FAD contents and the location of the Bf flavin at a pivotal location in the interface between domains I and III. This pH dependence in conjunction with the benefit of a cationic Arg or Lys at position 165 suggests that some entity should be deprotonated and anionic for optimal FAD binding. Two candidates are the FAD's pyrophosphate or the isoalloxazine ring itself. The pKa for deprotonation of the isoalloxazine at N3H is 10.3 for free FAD (
      • Schuman M.
      • Massey V.
      Purification and characterization of glycolic acid oxidase from pig liver.
      ). However, because the visible spectrum of the Bf flavin in the WT-RpaETF and R165K-RpaETF did not display the strong blue shift in band II reported for flavin deprotonated at N3 (
      • Schuman M.
      • Massey V.
      Purification and characterization of glycolic acid oxidase from pig liver.
      ,
      • Macheroux P.
      • Massey V.
      • Thiele D.J.
      Expression of spinach glycolate oxidase in Saccharomyces cerevisiae: Purification and characterization.
      ), we infer that the pyrophosphate is most likely the anion whose interaction with Arg-165 favors binding.
      Given R165H-RpaETF and R165Q-RpaETF's low FAD contents, we assessed the possibility that AMP could be bound in lieu of FAD in the Bf site, as in canonical ETFs (
      • Roberts D.L.
      • Frerman F.E.
      • Kim J.J.
      Three-dimensional structure of human electron transfer flavoprotein to 2.1-A resolution.
      ,
      • Ku H.-K.
      • Lim H.-M.
      • Oh K.-H.
      • Yang H.-J.
      • Jeong J.-S.
      • Kim S.-K.
      Interpretation of protein quantitation using the Bradford assay: Comparison with two calculation models.
      ). Because our assay detects the adenine moiety's absorbance at 260 nm, ADP could also be responsible for excess A260, but AMP is deemed more likely because it is the natural ligand of canonical ETFs. For ligands released from R165H-RpaETF, the ratio of A260 to A450 was only slightly higher than that of FAD, indicating 0.2 AMP bound per ETF. This contrasts with the situation in T94,97A-RpaETF where the Bf site is fully populated with AMP (and the ET site is replete with FAD) (
      • Mohamed-Raseek N.
      • Duan H.D.
      • Mroginski M.A.
      • Miller A.F.
      Spectroscopic, thermodynamic and computational evidence of the locations of the FADs in the nitrogen fixation-associated electron transfer flavoprotein.
      ). R165Q-RpaETF had almost as much AMP, but the FAD content was so low that AMP could have been bound in either site. In the R165H-RpaETF and R165Q-RpaETF, the sum of the FAD and AMP stoichiometries was closer to the amount required to populate a single site, the ET site based on our results above. Absence of a nucleotide from the Bf site is consistent with the low stability of the R165H/Q-RpaETF since canonical ETFs lacking AMP are unstable (
      • Sato K.
      • Nishima Y.
      • Shiga K.
      In vitro assembly of FAD, AMP, and the two subunits of electron-transferring flavoprotein: An important role of AMP related with the conformational change of the apoprotein.
      ). Because AMP is present in T94,97A-RpaETF but not R165H-RpaETF or R165Q-RpaETF, it appears that a cation at position 165 is needed for AMP binding in the ET site. We propose that favorable electrostatic interaction between cationic residue 165 and the phosphate(s) aids in binding even the AMP of canonical ETFs, not just the FAD of Bf ETFs. This can explain the high conservation of Arg at this position among ETFs in general.

      Discussion

      In this study, we explored the roles of two Arg residues that are conserved, one in each flavin-binding site. Nevertheless, the sites produce contrasting flavin reactivity. The salient difference between the two flavins is that the ET flavin undergoes reduction by two sequential 1e steps, whereas the Bf flavin undergoes a single 2e reduction. Therefore, we have characterized the effects on redox reactivity of replacing either Arg-273 in the ET site or Arg-165 in the Bf site of the RpaETF.
      In both sites, replacement of Arg by similarly charged Lys did not alter (s) or occupancy of the nearby FAD; however, 1e reactivity of the ET flavin was diminished when residue 273 was neutral and nucleotide binding in the Bf site was abrogated when residue 165 was neutral. This demonstrates that in both sites, positive charge at the location of the Arg is a critical element of the protein's interaction with FAD, even though nature and consequences of the interaction are different in the two sites.

      Redox tuning of the ET flavin for sequential 1e transfer activity

      A mechanism preventing the ET FAD from engaging in 2e redox chemistry is important in both canonical and Bf ETFs, and this is the proposed significance of positive charge at Arg-273, since WT-RpaETF, R273K-RpaETF, and R165K-RpaETF (retaining Arg-273) all undergo two sequential 1e reductions. However our data demonstrate that in R273A-RpaETF and R273H-RpaETF, OX ET FAD can be directly converted to HQ in a substantial fraction of sites (Table 3). ET FAD sites that accept both electrons from the Bf FAD could mediate wasteful 2e reduction of quinone by NADH, uncoupled from reduction of flavodoxin/ferredoxin.
      Quantitatively, the consequences of removal of Arg-273's positive charge conformed with our expectations and strong literature precedent (summarized in Table 6). The ET flavin's elevated E°'OX/ASQ of −61 mV in WT-RpaETF, compared with that of free flavin (−313 mV (
      • Mayhew S.G.
      The effects of pH and semiquinone formation on the oxidation-reduction potentials of flavin mononucleotide: A reappraisal.
      )), is by no means unique. Indeed, E°'OX/ASQ is more elevated in other Bf ETFs. Arg-273's charge is clearly critical since R273A-RpaETF's and R273H-RpaETF's E°’OX/ASQ values are lower by more than 60 mV, showing a 5.8 kJ/mol (1.4 kcal/mol) lower driving force for reduction of OX to ASQ compared with WT-RpaETF, R273K-RpaETF, and R165K-RpaETF. This confirms work by Scrutton and coworkers on MmeETF (
      • Burgess S.G.
      • Messiha H.L.
      • Katona G.
      • Rigby S.E.J.
      • Leys D.
      • Scrutton N.S.
      Probing the dynamic interface between trimethylamine dehydrogenase (TMADH) and electron transferring flavoprotein (ETF) in the TMADH-2ETF complex: Role of the Arg-α237 (ETF) and Tyr-442 (TMADH) residue pair.
      ,
      • Talfournier F.
      • Munro A.W.
      • Basran J.
      • Sutcliffe M.J.
      • Daff S.
      • Chapman S.K.
      • Scrutton N.S.
      Alpha Arg-237 in Methylophilus methylotrophus (sp. W3A1) electron-transferring flavoprotein affords approximately 200-millivolt stabilization of the FAD anionic semiquinone and a kinetic block on full reduction to the dihydroquinone.
      ,
      • Jang M.-H.
      • Scrutton N.S.
      • Hille R.
      formation of W3A1 electron-transferring flavoprotein (ETF) hydroquinone in the trimethylamine dehydrogenase·ETF protein complex.
      ) where replacement of the Arg corresponding to Arg-273 by an Ala lowered E°’OX/ASQ by 200 mV from +153 mV to −47 mV. In our case the largest effect of replacing Arg-273 was a 76 mV depression of OX/ASQ observed upon substitution with Ala. The elevated E°'OX/ASQ of the WT systems can be understood substantially in terms of favorable electrostatic interaction of the formed ASQ. Thus, our data agree well with prior work but are the first to extend it to a Bf ETF, where stabilization of the ET flavin in the ASQ state in vivo would limit it to accepting only a single electron (
      • Schut G.J.
      • Mohamed-Raseek N.R.
      • Tokmina-Lukaszewska M.
      • Mulder D.E.
      • Nguyen D.M.N.
      • Lipscomb G.L.
      • Hoben J.P.
      • Patterson A.
      • Lubner C.E.
      • King P.W.
      • Peters J.W.
      • Bothner B.
      • Miller A.F.
      • Adams M.W.W.
      The catalytic mechanism of electron bifurcating electron transfer flavoproteins (ETFs) involves an intermediary complex with NAD+.
      ,
      • Sucharitakul J.
      • Buckel W.
      • Chaiyen P.
      Rapid kinetics reveal surprising flavin chemistry in the bifurcating electron transfer flavoprotein from Acidaminococcus fermentans.
      ).
      Table 6Comparison of E°’s of 1 and 2-FAD systems
      values are adjusted to pH 7, see also (14, 19).
      .
      SystemFlavin contentE°’OX/ASQ; ASQ/HQ; OX/HQ (mV)Ref.
      Free flavinN/A−313; −101; −207(
      • Mayhew S.G.
      The effects of pH and semiquinone formation on the oxidation-reduction potentials of flavin mononucleotide: A reappraisal.
      )
      2-FAD ETFs
      PaeETF2+33; −94; −285(
      • Schut G.J.
      • Mohamed-Raseek N.R.
      • Tokmina-Lukaszewska M.
      • Mulder D.E.
      • Nguyen D.M.N.
      • Lipscomb G.L.
      • Hoben J.P.
      • Patterson A.
      • Lubner C.E.
      • King P.W.
      • Peters J.W.
      • Bothner B.
      • Miller A.F.
      • Adams M.W.W.
      The catalytic mechanism of electron bifurcating electron transfer flavoproteins (ETFs) involves an intermediary complex with NAD+.
      )
      AfeETF2+134; −36; −249(
      • Sucharitakul J.
      • Buttranon S.
      • Wongnate T.
      • Chowdhury N.P.
      • Prongjit M.
      • Buckel W.
      • Chaiyen P.
      Modulations of the reduction potentials of flavin-based electron bifurcation complexes and semiquinone stabilities are key to control directional electron flow.
      )
      WT RpaETF2−47; −83; −223(
      • Duan H.D.
      • Lubner C.E.
      • Tokmina-Lukaszewska M.
      • Gauss G.H.
      • Bothner B.
      • King P.W.
      • Peters J.W.
      • Miller A.F.
      Distinct flavin properties underlie flavin-based electron bifurcation within a novel electron-transferring flavoprotein FixAB from Rhodopseudomonas palustris.
      ,
      • Mohamed-Raseek N.
      • Duan H.D.
      • Mroginski M.A.
      • Miller A.F.
      Spectroscopic, thermodynamic and computational evidence of the locations of the FADs in the nitrogen fixation-associated electron transfer flavoprotein.
      )
      −61; −122; −243This work
      R273K-RpaETF2−69, −125, −243This work
      R273H-RpaETF2−119, −141, −280This work
      R273A-RpaETF2−137, −118, −283This work
      R165K-RpaETF2−67.6, −122, −241This work
      T94,97A-RpaETF1−7, −96This work
      MelETF2+81; −136; −279(
      • Sato K.
      • Nishina Y.
      • Shiga K.
      Interaction between NADH and electron-transferring flavoprotein from Megasphaera elsdenii.
      )
      1-FAD ETFs
      T94,97A RpaETF1−7; −96(
      • Mohamed-Raseek N.
      • Duan H.D.
      • Mroginski M.A.
      • Miller A.F.
      Spectroscopic, thermodynamic and computational evidence of the locations of the FADs in the nitrogen fixation-associated electron transfer flavoprotein.
      ), This work
      MmeETF WT1+153; <−250; N.A.(
      • Burgess S.G.
      • Messiha H.L.
      • Katona G.
      • Rigby S.E.J.
      • Leys D.
      • Scrutton N.S.
      Probing the dynamic interface between trimethylamine dehydrogenase (TMADH) and electron transferring flavoprotein (ETF) in the TMADH-2ETF complex: Role of the Arg-α237 (ETF) and Tyr-442 (TMADH) residue pair.
      ,
      • Talfournier F.
      • Munro A.W.
      • Basran J.
      • Sutcliffe M.J.
      • Daff S.
      • Chapman S.K.
      • Scrutton N.S.
      Alpha Arg-237 in Methylophilus methylotrophus (sp. W3A1) electron-transferring flavoprotein affords approximately 200-millivolt stabilization of the FAD anionic semiquinone and a kinetic block on full reduction to the dihydroquinone.
      )
      MmeETF αR237A1−43; −31; N.A.(
      • Burgess S.G.
      • Messiha H.L.
      • Katona G.
      • Rigby S.E.J.
      • Leys D.
      • Scrutton N.S.
      Probing the dynamic interface between trimethylamine dehydrogenase (TMADH) and electron transferring flavoprotein (ETF) in the TMADH-2ETF complex: Role of the Arg-α237 (ETF) and Tyr-442 (TMADH) residue pair.
      ,
      • Talfournier F.
      • Munro A.W.
      • Basran J.
      • Sutcliffe M.J.
      • Daff S.
      • Chapman S.K.
      • Scrutton N.S.
      Alpha Arg-237 in Methylophilus methylotrophus (sp. W3A1) electron-transferring flavoprotein affords approximately 200-millivolt stabilization of the FAD anionic semiquinone and a kinetic block on full reduction to the dihydroquinone.
      )
      MmeETF αR237K1−62; <−250; N.A.(
      • Burgess S.G.
      • Messiha H.L.
      • Katona G.
      • Rigby S.E.J.
      • Leys D.
      • Scrutton N.S.
      Probing the dynamic interface between trimethylamine dehydrogenase (TMADH) and electron transferring flavoprotein (ETF) in the TMADH-2ETF complex: Role of the Arg-α237 (ETF) and Tyr-442 (TMADH) residue pair.
      )
      αR237C1−57; −145; N.A.(
      • Burgess S.G.
      • Messiha H.L.
      • Katona G.
      • Rigby S.E.J.
      • Leys D.
      • Scrutton N.S.
      Probing the dynamic interface between trimethylamine dehydrogenase (TMADH) and electron transferring flavoprotein (ETF) in the TMADH-2ETF complex: Role of the Arg-α237 (ETF) and Tyr-442 (TMADH) residue pair.
      )
      αR237E1−59; −199; N.A.(
      • Burgess S.G.
      • Messiha H.L.
      • Katona G.
      • Rigby S.E.J.
      • Leys D.
      • Scrutton N.S.
      Probing the dynamic interface between trimethylamine dehydrogenase (TMADH) and electron transferring flavoprotein (ETF) in the TMADH-2ETF complex: Role of the Arg-α237 (ETF) and Tyr-442 (TMADH) residue pair.
      )
      Human WT1+22; −42; N.A.(
      • Dwyer T.M.
      • Zhang L.
      • Muller M.
      • Marrugo F.
      • Frerman F.E.
      The functions of the flavin contact residues αArg249 and βTyr16, in human electron transfer flavoprotein.
      )
      Human αR249K1−39: −124; N.A.(
      • Dwyer T.M.
      • Zhang L.
      • Muller M.
      • Marrugo F.
      • Frerman F.E.
      The functions of the flavin contact residues αArg249 and βTyr16, in human electron transfer flavoprotein.
      )
      Pig liver

      Sus domesticus
      1+4; −20; N.A.(
      • Husain M.
      • Stankovich M.T.
      • Fox B.G.
      Measurement of the oxidation-reduction potentials for one-electron and two-electron reduction of electron-transfer flavoprotein from pig liver.
      )
      Paracoccus denitrificans1−6; −36; N.A.(
      • Griffin K.J.
      • Dwyer T.M.
      • Manning M.C.
      • Meyer J.D.
      • Carpenter J.F.
      • Frerman F.E.
      alphaT244M mutation affects the redox, kinetic, and in vitro folding properties of Paracoccus denitrificans electron transfer flavoprotein.
      )
      a values are adjusted to pH 7, see also (
      • Sato K.
      • Nishina Y.
      • Shiga K.
      Interaction between NADH and electron-transferring flavoprotein from Megasphaera elsdenii.
      ,
      • Duan H.D.
      • Lubner C.E.
      • Tokmina-Lukaszewska M.
      • Gauss G.H.
      • Bothner B.
      • King P.W.
      • Peters J.W.
      • Miller A.F.
      Distinct flavin properties underlie flavin-based electron bifurcation within a novel electron-transferring flavoprotein FixAB from Rhodopseudomonas palustris.
      ).
      In contrast with our findings, work from the teams of Scrutton and Frerman found that replacing the ET site Arg with a Lys decreased the E°’OX/ASQ by 60 mV or more (
      • Burgess S.G.
      • Messiha H.L.
      • Katona G.
      • Rigby S.E.J.
      • Leys D.
      • Scrutton N.S.
      Probing the dynamic interface between trimethylamine dehydrogenase (TMADH) and electron transferring flavoprotein (ETF) in the TMADH-2ETF complex: Role of the Arg-α237 (ETF) and Tyr-442 (TMADH) residue pair.
      ,
      • Dwyer T.M.
      • Zhang L.
      • Muller M.
      • Marrugo F.
      • Frerman F.E.
      The functions of the flavin contact residues αArg249 and βTyr16, in human electron transfer flavoprotein.
      ), but their resulting E°'s agree with that of our R273K-RpaETF (for the corresponding R to K variants, E°’OX/ASQ = −62 mV for MmeETF, −39 mV for human ETF versus. -69 mV for R273K-RpaETF). In contrast the WT-RpaETF's OX/ASQ is 80 to 200 mV lower than those of WT-human and WT-MmeETF, suggesting that the Arg-273 in RpaETF is less supportive of ASQ than the corresponding Args in MmeETF and human ETF. We propose that the anomalous E°'OX/ASQ of our WT ET site reflects incomplete engagement of RpaETF's Arg-273 with the ASQ flavin under our conditions. Crystal structures show the Arg corresponding to Arg-273 interacting with Glu residues in the partner protein (e.g., Glu 198 in butyryl CoA dehydrogenase (Bcd) of Clostridium difficile) (
      • Demmer J.K.
      • Chowdhury N.P.
      • Selmer T.
      • Ermler U.
      • Buckel W.
      The semiquinone swing in the bifurcating electron transferring flavoprotein/butyryl-coA dehydrogenase complex from Clostridium difficile.
      ) or the other ETF domain (
      • Toogood H.S.
      • van Thiel A.
      • Scrutton N.S.
      • Leys D.
      Stabilization of non-productive conformations underpins rapid electron transfer to electron transferring flavoprotein.
      ). Thus, Arg-273 is prone to multiple interactions that could dilute its stabilizing effect on the ET flavin ASQ and anionic HQ, depending on the protein conformation. Arg-273's tendency to “dabble” with residues in other domains or partners could be a cause of the observed dramatic change in the ET flavin's s when AfeETF has access to its partner Bcd (
      • Sucharitakul J.
      • Buttranon S.
      • Wongnate T.
      • Chowdhury N.P.
      • Prongjit M.
      • Buckel W.
      • Chaiyen P.
      Modulations of the reduction potentials of flavin-based electron bifurcation complexes and semiquinone stabilities are key to control directional electron flow.
      ).

      Coupling between flavin reactivities in the two sites, a role for conformation?

      By extending substitutions of Arg-273 to Bf ETFs, we can address the new question of consequences for one flavin site of amino acid substitutions in the other. Some cross talk is expected because the two flavins must function together. Indeed, studies of pre-steady-state kinetics revealed that the oxidation state of the ET flavin affects the rate of Bf flavin reduction by NADH, as well as its kinetic isotope effect, in AfeETF (
      • Sucharitakul J.
      • Buckel W.
      • Chaiyen P.
      Rapid kinetics reveal surprising flavin chemistry in the bifurcating electron transfer flavoprotein from Acidaminococcus fermentans.
      ). Similarly, we found that replacement of Arg-273 with neutral residues depressed E°OX/HQ of the Bf flavin. Conversely, Sato et al. found that the ET flavin's optical signature depended on the status of the Bf site (
      • Sato K.
      • Nishina Y.
      • Shiga K.
      Purification of electron-transferring flavoprotein from Megasphaera elsdenii and binding of additional FAD with an unusual absorption spectrum.
      ), and we found that absence of the Bf flavin was associated with a shift to shorter wavelengths of band II of the ET flavin and covalent modification. This was observed even when Arg-165 was retained (as in T94,97A).
      Prior work on canonical ETFs concluded that Arg-273 forms interactions stabilizing a “closed” (=B) conformation of isolated ETF (
      • Toogood H.S.
      • Leys D.
      • Scrutton N.S.
      Dynamics driving function : New insights from electron transferring flavoproteins and partner complexes.
      ) and its mutation to other residues removes a kinetic impediment to full reduction (
      • Burgess S.G.
      • Messiha H.L.
      • Katona G.
      • Rigby S.E.J.
      • Leys D.
      • Scrutton N.S.
      Probing the dynamic interface between trimethylamine dehydrogenase (TMADH) and electron transferring flavoprotein (ETF) in the TMADH-2ETF complex: Role of the Arg-α237 (ETF) and Tyr-442 (TMADH) residue pair.
      ). Thus, we speculate that our R273A-RpaETF and R273H-RpaETF populate the D conformation more than the WT-RpaETF does and infer that the OX/HQ of the Bf flavin may be lower in the D conformation than in the B conformation. We note that this would act to reinforce the forward catalysis by disfavoring electron transfer back from the ET flavin to the Bf flavin (but see(23)). The lower E°’OX/HQ of the Bf FAD than the ASQ/HQ of the ET FAD is critical for bifurcation, and this is preserved in all our variants.
      Regarding ET flavin modification observed when the Bf flavin was absent, proclivity to modification is a long-recognized property of the ET flavin of ETFs (
      • Lehman T.C.
      • Thorpe C.
      A new form of mammalian electron transfer flavoprotein.
      ,
      • Mayhew S.G.
      • Whitfield C.D.
      • Ghisla S.
      • S.-J. M.
      Identification and properties of new flavins in electron-transferring flavoprotein from Peptostreptococcus elsdenii and pig-liver glycolate oxidase.
      ,
      • Ghisla S.
      • Massey V.
      • Mayhew S.G.
      Studies on the active centers of flavoproteins: Binding of 8-hydroxy-FAD and 8-hydroxy-FMN to apoproteins.
      ,
      • O’Neill H.
      • Mayhew S.G.
      • Butler G.
      Cloning and analysis of the genes for a novel electron-transferring flavoprotein from Megasphaera elsdenii.
      ). Indeed, at pH 9, we observed flavin modification in the T94,97A-RpaETF that retains only the ET FAD (
      • Mohamed-Raseek N.
      • Duan H.D.
      • Mroginski M.A.
      • Miller A.F.
      Spectroscopic, thermodynamic and computational evidence of the locations of the FADs in the nitrogen fixation-associated electron transfer flavoprotein.
      ). Spectra of T94,97A-RpaETF maintained at pH 9, and of FAD released from it, demonstrated formation of 8-formyl flavin (Fig. 6B) as described by Macheroux' team (
      • Augustin P.
      • Toplak M.
      • Fuchs K.
      • Gerstmann E.C.
      • Prassl R.
      • Winkler A.
      • Macheroux P.
      Oxidation of the FAD cofactor to the 8-formyl-derivative in human electron-transferring flavoprotein.
      ). However, maintenance of T94,97A-RpaETF at pH 8 prevented formation of the modified FAD, hence our use of pH 8 in the current work (
      • Mohamed-Raseek N.
      • Duan H.D.
      • Mroginski M.A.
      • Miller A.F.
      Spectroscopic, thermodynamic and computational evidence of the locations of the FADs in the nitrogen fixation-associated electron transfer flavoprotein.
      ), whereas earlier work on RpaETF has been conducted at pH 9 for stability reasons. Augustin et al. (
      • Augustin P.
      • Toplak M.
      • Fuchs K.
      • Gerstmann E.C.
      • Prassl R.
      • Winkler A.
      • Macheroux P.
      Oxidation of the FAD cofactor to the 8-formyl-derivative in human electron-transferring flavoprotein.
      ) found that substitutions favoring the “open” conformation (=D) accelerated formation of 8-formyl flavin. Thus, T94,97A-RpaETF's greater accumulation of 8-formyl flavin than WT-RpaETF could indicate a higher population of the open conformation. Canonical ETFs that naturally lack Bf FAD could spend more time in the open conformation where Arg-273 will be less committed to interactions with other side chains and thus more engaged with the ET FAD ASQ, when partner protein is not present. This would make sense of canonical ETFs' higher E°'OX/ASQs than that of WT-RpaETF.

      Nucleotide binding and stabilization of the dimer interface between domains I and III

      Our replacement of Arg-165 with Lys in RpaETF yielded protein containing two FADs and with a WT-like Tm. However, R165Q-RpaETF and R165H-RpaETF incorporated much less FAD plus AMP and were too unstable for measurement of s and other detailed characterizations. The more extreme substitution, R165A, was expressed into insoluble fractions only and could not be recovered for study. Thus, a positive charge is best at position 165, a polar H-bond enables protein folding but not nucleotide binding, and absence of both positive charge and H-bonding even precludes folding of the protein. The significance of H-bonding by Arg-165 is substantiated by the vibrational structure visible on band I of Bf FAD in WT-RpaETF, but not in R165K-RpaETF, indicating destabilization of H-bonding interactions in the variant, even though the Bf flavin is bound. We attribute this to Arg's multifunctional head group that is better able to engage in several H-bonds at once.
      Although the instability of the R165H-RpaETF and R165Q-RpaETF precluded measurements, we note that the depressed OX/HQ of the Bf flavin in WT-RpaETF reveals that the anionic HQ and ASQ states, in particular, are strongly destabilized by this site. This has been difficult to reconcile with the fact that a nearby Arg should favor formation of the anionic states of the flavin. Since R165H-RpaETF (and R165Q-RpaETF) appears to not even bind AMP well (whereas T94,97A-RpaETF does), we propose that the anion stabilized is actually the phosphate(s) associated with the (di)nucleotide, whether FAD or AMP. Since the AMP portion of FAD is bound in domain III (chain B), whereas Arg-165 extends from domain I (chain A), this interaction would act as a link between domains, consistent with AMP's importance to the assembly of canonical ETFs (
      • Sato K.
      • Nishima Y.
      • Shiga K.
      In vitro assembly of FAD, AMP, and the two subunits of electron-transferring flavoprotein: An important role of AMP related with the conformational change of the apoprotein.
      ).
      Such an interaction between domains appears to be mediated via a stretch of peptide backbone from chain B that would not neutralize Arg-165's charge, but instead would propagate it. The peptide is oriented such as to present carbonyl Os to Arg-165's side chain (
      • Kayastha K.
      • Vitt S.
      • Buckel W.
      • Ermler U.
      Flavins in the electron bifurcation process.
      ). This in turn constrains the peptide to present its backbone NH groups toward the phosphate(s), acting as a zipper to hold together domains I and III (Fig. 7). This electrostatically polarized network of H-bond is seen in the crystal structures of each of the ETFs that has been crystallized (Fig. S4).
      Figure thumbnail gr7
      Figure 7Extended electrostatically polarized network of H-bonds bridging the domain interface and mediated by the conserved peptide zipper. Figure is based on the crystal structure of AfeETF, 4KPU.pdb (
      • Chowdhury N.P.
      • Mowafy A.M.
      • Demmer J.K.
      • Upadhyay V.
      • Koelzer S.
      • Jayamani E.
      • Kahnt J.
      • Hornung M.
      • Demmer U.
      • Ermler U.
      • Buckel W.
      Studies on the mechanism of electron bifurcation catalyzed by electron transferring flavoprotein (Etf) and butyryl-CoA dehydrogenase (Bcd) of Acidaminococcus fermentans.
      ). RpaETF residue numbering is used, for consistency with text, but in AfeETF the EtfA residues shown are Arg-146, Ala-129 (in blue), and the EtfB residues shown are Gly-123, Asp-124, Thr-125, and Ala-126 (in gray). AfeETF, ETF from Acidaminococcus fermentans; RpaETF, ETF of Rhodopseudomonas palustris.
      The peptide zipper represents one of the conserved signatures of Bf ETFs (
      • Garcia Costas A.M.
      • Poudel S.
      • Miller A.-F.
      • J S.G.
      • Ledbetter R.N.
      • Fixen K.
      • Seefeldt L.C.
      • Adams M.W.
      • Harwood C.S.
      • Boyd E.S.
      • Peters J.W.
      Defining electron bifurcation in the electron transferring flavoprotein family.
      ,
      • Chowdhury N.P.
      • Kahnt J.
      • Buckel W.
      Reduction of ferredoxin or oxygen by flavin-based electron bifurcation in Megasphaera elsdenii.
      ). Considering the four residues shown in Figures 7 and S4 as well as the two flanking residues, DGDTAQ (the four depicted are in bold), the percent conservation in a set of 216 sequences of Bf ETFs(13) is 100% D, 95% G, 85% D (with E making up an additional 10%), 100% T, 86% A (with G as another 11%) and 93% Q (with H as 7%). Meanwhile in 449 sequences of canonical (group 1) ETFs, the corresponding sequence is 100% D, 68% D, 98% D, 36% S, 98% N, 92% Q. The residues remain highly conserved, but differently, with an “extra” Asp that could play the electrostatic role of the second phosphate of FAD that is absent from AMP in canonical ETFs. The Arg corresponding to Arg-165 remains Arg in 447 of 449 canonical ETFs, with Gln or Ser present in the two outliers.
      The peptide zipper appears to interact with both phosphates of FAD, suggesting that it could interact similarly with the AMP that replaces FAD in canonical ETFs. Indeed, the motif is evident in the crystal structures of canonical ETFs (Fig. S4) (
      • Roberts D.L.
      • Salazar D.
      • Fulmer J.P.
      • Frerman F.E.
      • Kim J.J.
      Crystal structure of paracoccus denitrificans electron transfer flavoprotein: Structural and electrostatic analysis of a conserved flavin binding domain.
      ), providing an explanation for the importance of AMP to assembly of the ETF subunits (
      • Sato K.
      • Nishina Y.
      • Shiga K.
      Electron-transferring flavoprotein has an AMP-binding site in addition to the FAD-binding site.
      ). Thus, we identify a widely shared conserved structural motif linking the subunits and domains of ETF whereby Arg-165 stabilizes the domain interface in conjunction with the nucleotide, via long-range electrostatics mediated by a highly conserved stretch of peptide backbone (the “zipper”).
      The proposed extended interaction stabilizing the domain interface can explain the conservation of Arg-165 even in canonical ETFs that lack FAD and moreover canonical ETFs' retention of AMP (
      • DuPlessis E.R.
      • Rohlfs R.J.
      • Hille R.
      • Thorpe C.
      Electron-transferring flavoproteins from pig and the methylotrophic bacterium W3A1 contains AMP as well as FAD.
      ). It also makes sense of the otherwise baffling conservation of Arg in a site that must destabilize ASQ. However, we speculate that there may be important functional significance to the use of this electrostatic mechanism to stabilize a particular conformation of the domain interface. We propose that upon reduction of Bf flavin, after release of NAD+ when the Bf flavin is anionic HQ, the Bf flavin’s charge could supersede that of the more distant phosphates and effectively capture the Arg, thereby releasing Arg's constraints on the peptide zipper. Indeed, the CryoEM structure displays partial inversion of the polarity of the peptide backbone in the loop, consistent with our proposal, and partial reduction of the Bf flavin by the electron beam (
      • Feng X.
      • Schut G.J.
      • Lipscomb G.
      • Li H.Y.
      • Adams M.W.W.
      Cryoelectron microscopy structure and mechanism of the membrane-associated electron-bifurcating flavoprotein Fix/EtfABCX.
      ). The freeing of this protein element could allow domain III to reposition itself relative to domain I, providing a means of coupling the interconversion of B and D conformations to the oxidation state of the Bf flavin.
      Thus, besides the H-bond between the side chain of Arg-165 and the oxidized flavin N5 (
      • Demmer J.K.
      • Bertsch J.
      • Oppinger C.
      • Wohlers H.
      • Kayastha K.
      • Demmer U.
      • Ermler U.
      • Muller V.
      Molecular basis of the flavin-based electron-bifurcating caffeyl-CoA reductase reaction.
      ,
      • Kayastha K.
      • Vitt S.
      • Buckel W.
      • Ermler U.
      Flavins in the electron bifurcation process.
      ), we now propose that Arg-165 in domain I interacts with the phosphate(s) of FAD or AMP of domain III via a conserved peptide zipper when the Bf FAD is oxidized. We furthermore speculate that upon Bf FAD reduction, reorganization of these interactions in the dimer interface could precipitate conformational change. If confirmed experimentally, this would be an elegant way to couple acquisition of electrons to domain-scale motion, via electrostatics.

      Concluding remarks

      Our study reports consequences of replacing each of the conserved active site Args of a Bf ETF and provides a first biophysical glimpse of the Bf flavin's interactions with a nearby residue. We have demonstrated that Arg-273 plays a vital role in stabilizing ET FAD ASQ relative to OX in a Bf ETF, as in canonical ETFs. However, we note that a divided interaction between Arg-273 and the ET flavin in RpaETF could explain the latter's lower OX/ASQ. We find that substitutions of neutral residues for Arg-273 also lower the OX/HQ of the Bf flavin, and absence of a flavin in the Bf site hastens modification of the ET flavin, demonstrating coupling between the two sites that may be mediated by conformational equilibria. Finally, we have demonstrated that Arg-165 plays an important role in binding nucleotide in the interface between domains I and III. This is consistent with the crystal structures of Bf ETFs in which Arg-165’s side chain not only forms an H-bond with the flavin N5, but also forms interactions bridging the domain interface and anchored by the negatively charged phosphate(s) of AMP or FAD. The long-range electrostatic interaction mediated by H-bonding by a conserved stretch of peptide backbone is also observed in canonical ETFs, explaining the stabilizing role of AMP in canonical ETFs and the universal conservation of Arg-165. Moreover, we speculate that the use of electrostatics to stabilize the dimer interface serves to make the interface inherently responsive to the oxidation state of the Bf flavin.

      Experimental procedures

      Site-directed mutagenesis

      Note that the A, B naming of the ETF subunits is opposite to their naming in the Fix convention, such that EtfA is FixB, and vice versa. Thus, the fixA gene encodes EtfB and fixB encodes EtfA. Mutations were introduced into fixB augmented with an N-terminal His tag in the pMCSG21 plasmid (spectinomycin resistant) (
      • Duan H.D.
      • Lubner C.E.
      • Tokmina-Lukaszewska M.
      • Gauss G.H.
      • Bothner B.
      • King P.W.
      • Peters J.W.
      • Miller A.F.
      Distinct flavin properties underlie flavin-based electron bifurcation within a novel electron-transferring flavoprotein FixAB from Rhodopseudomonas palustris.
      ). Primers encoding the desired amino acid substitutions were designed using NEB base changer (Table S2) and employed according to the vendor’s recommendations to introduce the mutations via polymerase chain reactions using Q5 High Fidelity DNA polymerase (New England Biolabs). The mutated construct encoding EtfA and plasmid encoding native EtfB were transformed into competent Nico21(DE3) Escherichia coli (New England Biolabs), along with pGro7 (Takara Bio) for coexpression of molecular chaperones GroES/EL. Plasmids expressing fixB with mutations encoding the R165H or R165K substitutions were obtained from Dr H. Diessel Duan.

      Protein expression and purification

      Cells were grown in 1 l of Terrific Broth supplemented with 20 mg/l riboflavin and 2 mM MgSO4 along with carbenicillin (100 μg/ml), spectinomycin (100 μg/ml), and chloramphenicol (50 μg/ml) at 37 °C, shaking at 250 rpm, to an OD600 of ∼1–2. After cooling the culture to 18 °C, groES/EL gene expression was induced using 1 mg/ml L-arabinose. After 30 min of growth, ETF gene expression was induced using 0.1 mM isopropyl 1- thio-β-D-galactopyranoside (IPTG). Cultures were then grown for an additional 20 h at 18 °C. Cells were harvested by centrifugation at 4500 rpm, 4 °C for 30 min, and the pellet was stored at −80 °C after washing once in PBS (10 mM Na2HPO4, 1.8 mM KH2PO4, 137 mM NaCl, 2.7 mM KCl, pH 7.4).
      Frozen cell pellet (∼30 g) was thawed and resuspended in 60 ml of BugBuster (Cat no.70923–4 EMD Millipore) containing 1 mM 4-(2-aminoethyl) benzenesulfonyl fluoride hydrochloride (Cat no. 21250, Chem-Impex International), 1 mM FAD (Cat. No. 00151, Chem-Impex International), 2 μl of Benzonase Nuclease HC (Cat. No. 71205–3, EMD Millipore), and 2 μl of rLysozyme solution (Cat. No. 71110–4, EMD Millipore) and further incubated at 4 °C for 2 h with stirring. After centrifugation at 15,000 × g for 30 min at 4 °C, the clarified protein solution was mixed with 1.5 ml of pre-equilibrated nickel-nitrilotriacetic acid resin (Cat no.88222, Thermo Fisher) and incubated at 4 °C with stirring for 30 min. Then the mixture was transferred to a column at 4 °C. After collecting the flow-through, the column was washed with 20 column volumes of 20 mM Tris, pH 7.8, 500 mM KCl containing 20 mM imidazole. Finally, the column was developed with two column volumes of 20 mM Tris, pH 7.8, 500 mM KCl containing 100 mM imidazole, and the eluate was collected in three fractions. After SDS-PAGE analysis, fractions containing the purest most concentrated protein were pooled and passed over a 10DG column (Bio-Rad) equilibrated with working buffer (20 mM bis-Tris Propane, pH 8.0, 200 mM KCl, 10% (w/v) glycerol), to remove imidazole. To minimize covalent modification of the ET flavin, the current work was conducted at pH 8 unless stated otherwise. Coexpression of WT-RpaETF in the presence of chaperone and equilibration of the protein with the Ni-NTA resin for a shortened time interval (
      • Mohamed-Raseek N.
      • Duan H.D.
      • Mroginski M.A.
      • Miller A.F.
      Spectroscopic, thermodynamic and computational evidence of the locations of the FADs in the nitrogen fixation-associated electron transfer flavoprotein.
      ,
      • Duan H.D.
      • Mohamed-Raseek N.
      • Miller A.F.
      Spectroscopic evidence for direct flavin-flavin contact in a bifurcating electron transfer flavoprotein.
      ) enhanced the yield of ETF by tenfold compared with previous work (
      • Duan H.D.
      • Lubner C.E.
      • Tokmina-Lukaszewska M.
      • Gauss G.H.
      • Bothner B.
      • King P.W.
      • Peters J.W.
      • Miller A.F.
      Distinct flavin properties underlie flavin-based electron bifurcation within a novel electron-transferring flavoprotein FixAB from Rhodopseudomonas palustris.
      ).
      FAD lost during purification was reconstituted by overnight incubation in 1 mM FAD at 4 °C in darkness, followed by gel filtration over a 10DG column to remove unbound FAD. This resulted in protein containing two FAD per ETF dimer, although Bf ETFs purified without the two exposures to FAD often possessed as little as one (
      • Sato K.
      • Nishina Y.
      • Shiga K.
      Purification of electron-transferring flavoprotein from Megasphaera elsdenii and binding of additional FAD with an unusual absorption spectrum.
      ,
      • Chowdhury N.P.
      • Mowafy A.M.
      • Demmer J.K.
      • Upadhyay V.
      • Koelzer S.
      • Jayamani E.
      • Kahnt J.
      • Hornung M.
      • Demmer U.
      • Ermler U.
      • Buckel W.
      Studies on the mechanism of electron bifurcation catalyzed by electron transferring flavoprotein (Etf) and butyryl-CoA dehydrogenase (Bcd) of Acidaminococcus fermentans.
      ). Proteins were used the same day or flash-frozen in liquid nitrogen and stored at −80 °C for later use. The visible spectra of WT and variants other than R165Q-RpaETF indicate that they were purified predominantly in the oxidized state.

      Determination of protein and flavin concentrations

      The protein concentration (μg/ml) was determined using the Pierce 660 nm protein assay with bovine serum albumin as the standard (Thermo Fisher scientific). ETF mM concentration was calculated based on the molecular mass of 74545 Da (
      • Duan H.D.
      • Lubner C.E.
      • Tokmina-Lukaszewska M.
      • Gauss G.H.
      • Bothner B.
      • King P.W.
      • Peters J.W.
      • Miller A.F.
      Distinct flavin properties underlie flavin-based electron bifurcation within a novel electron-transferring flavoprotein FixAB from Rhodopseudomonas palustris.
      ).
      To quantify flavin and adenine, individual 1.5 ml microcentrifuge tubes containing 350 μl of ∼30 μM ETF in working buffer were wrapped in aluminum foil to exclude light and prevent photochemical transformation of cofactors. Cofactors were released by denaturing the ETF by heating at 100 °C for 10 min after which denatured protein was removed by centrifugation at 14,000 × g for 10 min after cooling the samples. Supernatant was transferred to a quartz cuvette and the optical spectrum was recorded on a HP 8453 spectrophotometer. Absorbance at 450 nm was used to determine the released flavin concentration (ℇ450 = 11.3 mM−1 cm−1) (
      • Sato K.
      • Nishina Y.
      • Shiga K.
      Interaction between NADH and electron-transferring flavoprotein from Megasphaera elsdenii.
      ). From this and FAD’s (ℇ260 = 35.93 mM−1 cm-1), we calculated the absorbance due to FAD at 260 nm. The excess A260 was used to determine the concentration of AMP or ADP (ℇ260 = 15.0 mM−1 cm−1) (
      • Sato K.
      • Nishima Y.
      • Shiga K.
      In vitro assembly of FAD, AMP, and the two subunits of electron-transferring flavoprotein: An important role of AMP related with the conformational change of the apoprotein.
      ).

      Far-UV CD for Tm measurement, and visible CD of the flavins

      Far-UV CD spectra of all the variants (17–30 μM) were recorded using a JASCO J-800 series spectropolarimeter from 250 to 190 nm starting at 20 °C in a 0. 2 cm path length quartz cuvette with the following parameters: bandwidth= 1.00 nm, scanning speed = 50 nm min-1, and three accumulations. The molar ellipticity was calculated using the equation, [θ] = θ/(c × l), in which θ is the ellipticity in millidegrees, c is the concentration in mM, and l is the cell path length in cm. CD measurements were taken in working buffer starting from 20 °C to 60 °C at 2 °C intervals.
      Visible CD spectra of 98 μM of R165H-RpaETF and 68 μM of WT-RpaETF were recorded from 600 to 300 nm at 4 °C in a 1.0 cm path length quartz cuvette with the following parameters: scan speed = 100 nm min−1, bandwidth= 2.00 nm, scanning speed = 100 nm min−1, and three accumulations.

      Spectral deconvolution

      Individual spectra of each of the flavins were deduced based on their separate sequential reductions in anaerobic titrations with dithionite. Experimental details follow. Plots of A374 versus. A454 (or absorbances at appropriate λmax values) were used to identify the last spectrum of phase 2, in which the ET flavin is HQ and the Bf flavin is OX (ETHQ/BfOX). This was subtracted from the first spectrum when both flavins are OX (ETOX/BfOX) to yield a difference spectrum from which the contribution of BfOX is removed, to retain only ETOX minus ETHQ (ETOX-HQ in Fig. S2). Separately, the endpoint spectrum of the titration corresponds to ETHQ/BfHQ so its subtraction from the spectrum at the end of phase 2 (ETHQ/BfOX) removes contributions from the ET flavin and retains only the BfOX minus BfHQ difference spectrum of the Bf flavin (BfOX-HQ in Fig. S2). To correct for the subtractive contribution of an HQ flavin spectrum in both cases, we exploited the fact that the HQ spectrum is weaker than that of OX except at short wavelengths, so any site-specific differences will have a weaker influence. We therefore divided the endpoint spectrum corresponding to ETHQ/BfHQ by two to yield the HQ spectrum of a single average flavin and added that to each of the two difference spectra (ETOX minus ETHQ, and BfOX minus BfHQ) to deduce ETOX and BfOX. This method was adopted because it consistently gave flat baselines and did not produce any negative absorbances. Sample spectra from an example are provided as Fig. S2.

      Extent of ASQ formation

      This determination was also based on reductive titrations in inert atmosphere (described below). The extent of ASQ formation was determined from the slope of the plot of A374 versus. A454 early in phase 1 since this reveals the extents to which of 1e and 2e reduction contribute to consumption of OX (before enough ASQ accumulates to support a significant contribution from reduction of ASQ to HQ). A second estimate of the fraction of sites undergoing 1e reactivity was obtained from the slope late in phase 2, when reduction of OX to ASQ is no longer significant so the slope represents the combination of 1e reduction of ASQ to HQ and 2e reduction of OX to HQ. The three reference slopes are known from the WT case where the phases are well separated and almost pure. Details are provided in the supporting information, including our correction to account for the 16% of WT sites that undergo 2e reduction.

      Reductive titrations

      Reductive titrations were performed in inert atmosphere, monitored using a HP 8452A spectrophotometer (Agilent technologies) equipped with an OLIS controller, in an inert atmosphere afforded by a glove box (Belle Technology), using a 1 cm path length self-masking quartz cuvette at 25 °C with the temperature controlled by a Quantum NorthWest Peltier cuvette holder. ETF was reduced stepwise by addition of small aliquots amounting 1 μM sodium dithionite (ℇ315 = 8.04 mM−1 cm−1) (
      • Yousafzai F.K.
      • Eady R.R.
      Dithionite reduction kinetics of the dissimilatory copper-containing nitrite reductase of alcalegenes xylosoxidans : The SO radical binds to the substrate binding type 2 copper site before the type 2 copper is reduced.
      ). This low concentration provides sufficient driving force to fully reduce the FAD at pH 8 (
      • Miller A.F.
      • Duan H.D.
      • Varner T.A.
      • Mohamed-Raseek N.
      Reduction midpoint potentials of bifurcating electron transfer flavoproteins.
      ).

      Reduction midpoint potential determinations

      To obtain the dense stable datasets needed for accurate determination of reduction midpoint potentials, xanthine oxidase in combination with xanthine (420 μM) was used to provide a slow continuous delivery of reducing equivalents (
      • Massey V.
      A simple method for determination of redox potentials.
      ). The reaction mixture contained 10 μM MV (methyl viologen) as a redox mediator, ∼20 to 30 μM of ETF, and 3 to 5 μM of an appropriate indicator dye with an close to that of the ETF so that the two could equilibrate with one-another during reduction (Table S3). For example, our initial choice of methylene blue (MB, E°’ = 11 mV, 2e/H+ at pH 7, calculated  = −19 mV at pH 8) (
      • Miller A.F.
      • Duan H.D.
      • Varner T.A.
      • Mohamed-Raseek N.
      Reduction midpoint potentials of bifurcating electron transfer flavoproteins.
      ) proved nonoptimal for the current variants, as the s determined in the presence of MB were ∼ −50 mV for WT-, R165K-, and R273K-RpaETF. Whereas for R273A- and R273H-RpaETF, the OX/ASQs around ∼90 mV at pH 8 were more than 30 mV from MB’s potential of −19 mV at pH 8 (see Table S3). In addition, we observed that the last steps of reduction of R273A- and R273H-RpaETF coincided with formation of significant amounts of MV radical, even before full reduction of the ETF was achieved, suggesting a very low or slow equilibration with the indicator dye in later steps of reduction. Our final choice of indicator dyes resolved this issue too.
      The reaction mixture was in working buffer at pH 8 to suppress covalent modification of the ET flavin (
      • Augustin P.
      • Toplak M.
      • Fuchs K.
      • Gerstmann E.C.
      • Prassl R.
      • Winkler A.
      • Macheroux P.
      Oxidation of the FAD cofactor to the 8-formyl-derivative in human electron-transferring flavoprotein.
      ) and thus ensure that changes observed are attributable to the residue being varied, Arg-273 and Arg-165, not other factors. However, this precluded study of R165H-RpaETF with both FADs bound. The reaction was initiated by addition of 3 to 10 nM xanthine oxidase depending on the ETF and of the phase under investigation. Spectra were recorded every 1 min.
      OX/ASQ characterizing reduction of OX FAD to ASQ was calculated by relating the extent of reduction of FAD to the extent of concurrent reduction of the reference dye. Conversion of OX to ASQ was quantified based on absorbance changes at 454 nm, while the reduction of the dye was quantified based on loss of absorbance at λmax of the dye.
      For WT-RpaETF, the point in the titration corresponding to the maximum ASQ concentration was identified and treated as the endpoint of the first 1-e reduction and assigned to full conversion to ASQ based on the agreement between the shape of the spectrum and that of authentic samples of 100% ASQ. Each spectrum up to that point in the titration was converted to a difference spectrum by subtracting the endpoint spectrum, to reveal the remaining OX population. Absorbance at 454 nm was used to find the absorbance change due to reduction of one equivalent of FAD and thereby to calculate the fraction of OX converted to ASQ. [FADOX]/[FADASQ] at each point in the titration was calculated from the fractional populations. Absorbance at λmax of the dye was used to calculate the extent to which oxidized dye, DyeOX, was converted to reduced dye, DyeRED. The ratio of oxidized to reduced dye along with that of the FAD was employed in the Nernst equation,
      log([FADOX][FADASQ])=nFADnDyelog([DyeOX][DyeRED])+nFADF(EDyeoEFADo)2.303RT
      (1)


      where nFAD, nDye denote the number of electrons acquired by the flavin and the dye respectively during the reaction under study, and F, R, and T are Faraday's constant, the ideal gas constant, and the temperature, respectively. A plot of log([FADOX][FADASQ]) versus log([DyeOX][DyeRED]), was used to determine nFAD from the known value of nDye = 2 (Table S1).
      The intercept of the plot then permits calculation of FAD for the ET FAD OX/ASQ couple.
      intercept=nFADF(EDyeoEFADo)2.303RT
      (2)


      Similar steps were followed to calculate the ASQ/HQ of all the ETFs. However, in determining OX/HQ from phase 3 of reduction in the presence of PS and SO, the strong absorbances of the dyes at 454 nm interferes with the flavin spectra. Thus, the extent of flavin oxidation was assessed at the isosbestic point of the dye’s reduction at 406 nm (for SO dye), where the extinction coefficients of oxidized and reduced dye are equal.

      Structural model of WT-RpaETF

      No crystal structure of RpaETF is available but both subunits have over 40% identity with the corresponding subunits of the AfeETF. Therefore, we generated a structural model of RpaETF based on the crystallographic coordinates 4KPU.pdb of AfeETF using Swiss Model (
      • Waterhouse A.
      • Bertoni M.
      • Bienert S.
      • Studer G.
      • Tauriello G.
      • Gumienny R.
      • Heer F.T.
      • de Beer T.A.P.
      • Rempfer C.
      • Bordoli L.
      • Lepore R.
      • Schwede T.
      SWISS-MODEL: Homology modelling of protein structures and complexes.
      ). The active site residues of RpaETF are shared by AfeETF, and indeed the coordinates of AfeETF's FAD moieties could be added to the protein model produced by Swiss Model with only minor clashes affecting a few side chains. To eliminate artifacts resulting from amino acid side chain differences, we subjected the initial RpaETF model to molecular mechanics geometry optimization by energy minimization using CHARMM (
      • Huang J.
      • Rauscher S.
      • Nawrocki G.
      • Ran T.
      • Feig M.
      • de Groot B.L.
      • Grubmüller H.
      • MacKerell Jr., A.D.
      CHARMM36m: An improved force field for folded and intrinsically disordered proteins.
      ), and then validated the resulting active sites by quantum mechanical molecular mechanical optimization followed by computation of electronic spectra for each of the two flavins using time-dependent density functional theory (TD-DFT) using B3LYP/6-31++G(d,p) (
      • Mohamed-Raseek N.
      • Duan H.D.
      • Mroginski M.A.
      • Miller A.F.
      Spectroscopic, thermodynamic and computational evidence of the locations of the FADs in the nitrogen fixation-associated electron transfer flavoprotein.
      ). The computed OX spectra replicate the differences between the observed spectra of OX ET FAD and Bf FAD, indicating that the optimized flavin-binding sites capture the environmental factors and interactions that are most important determinants of the flavin electronic structures.

      Data availability

      Data will be made available upon reasonable request, made to A.-F. Miller [email protected] .

      Supporting information

      This article contains supporting information (
      • Demmer J.K.
      • Bertsch J.
      • Oppinger C.
      • Wohlers H.
      • Kayastha K.
      • Demmer U.
      • Ermler U.
      • Muller V.
      Molecular basis of the flavin-based electron-bifurcating caffeyl-CoA reductase reaction.
      ,
      • Demmer J.K.
      • Chowdhury N.P.
      • Selmer T.
      • Ermler U.
      • Buckel W.
      The semiquinone swing in the bifurcating electron transferring flavoprotein/butyryl-coA dehydrogenase complex from Clostridium difficile.
      ,
      • Roberts D.L.
      • Frerman F.E.
      • Kim J.J.
      Three-dimensional structure of human electron transfer flavoprotein to 2.1-A resolution.
      ,
      • Chowdhury N.P.
      • Mowafy A.M.
      • Demmer J.K.
      • Upadhyay V.
      • Koelzer S.
      • Jayamani E.
      • Kahnt J.
      • Hornung M.
      • Demmer U.
      • Ermler U.
      • Buckel W.
      Studies on the mechanism of electron bifurcation catalyzed by electron transferring flavoprotein (Etf) and butyryl-CoA dehydrogenase (Bcd) of Acidaminococcus fermentans.
      ,
      • Duan H.D.
      • Lubner C.E.
      • Tokmina-Lukaszewska M.
      • Gauss G.H.
      • Bothner B.
      • King P.W.
      • Peters J.W.
      • Miller A.F.
      Distinct flavin properties underlie flavin-based electron bifurcation within a novel electron-transferring flavoprotein FixAB from Rhodopseudomonas palustris.
      ,
      • Mohamed-Raseek N.
      • Duan H.D.
      • Mroginski M.A.
      • Miller A.F.
      Spectroscopic, thermodynamic and computational evidence of the locations of the FADs in the nitrogen fixation-associated electron transfer flavoprotein.
      ,
      • Miller A.F.
      • Duan H.D.
      • Varner T.A.
      • Mohamed-Raseek N.
      Reduction midpoint potentials of bifurcating electron transfer flavoproteins.
      ,
      • Lei C.
      • Zhang Z.
      • Liu H.
      • Kong J.
      • Deng J.
      Biosensoring of hydrogen peroxide using new methylene blue N incorporated in a montmorillonite-modified horseradish peroxidase immobilization matrix as an electron shuttle.
      ,
      • Oja S.M.
      • Guerrette J.P.
      • David M.R.
      • Zhang B.
      Fluorescence-enabled electrochemical microscopy with dihydroresorufin as a fluorogenic indicator.
      ,
      • Twigg R.S.
      Oxidation-reduction aspects of resazurin.
      ,
      • Meng E.C.
      • Pettersen E.F.
      • Couch G.S.
      • Huang C.C.
      • Ferrin T.E.
      Tools for integrated sequence-structure analysis with UCSF Chimera.
      ,
      • Toogood H.S.
      • van Thiel A.
      • Basran J.
      • Sutcliffe M.J.
      • Scrutton N.S.
      Extensive domain motion and electron transfer in the human electron transferring flavoprotein.medium chain Acyl-CoA dehydrogenase complex.
      ).

      Conflict of interest

      The authors declare that they have no conflicts of interest with the contents of this article.

      Acknowledgments

      The authors are grateful to the National Science Foundation for support under awards numbers CLP-1808433 and CLP-2108134 . The authors thank H. D. Duan for providing R165K- and R165H-RpaETF constructs, T. Creamer for access to the CD spectrapolarimeter, A. Sebesta and B. Gess for maintenance of vital research infrastructure, S. Wang and A. Iyer for access to mass spectrometry, K. Ranguelova for guidance with EPR, and M. Ilham.

      Author contributions

      N. M.-R. and A.-F. M. conceptualization; A.-F. M. funding acquisition; N. M.-R. investigation; A.-F. M. project administration; A.-F. M. supervision; N. M.-R. and A.-F. M. visualization; N. M.-R. writing—original draft; A.-F. M. writing—reviewing and editing.

      Supporting information

      References

        • Toogood H.S.
        • Leys D.
        • Scrutton N.S.
        Dynamics driving function : New insights from electron transferring flavoproteins and partner complexes.
        FEBS J. 2007; 274: 5481-5504
        • Toogood H.S.
        • van Thiel A.
        • Scrutton N.S.
        • Leys D.
        Stabilization of non-productive conformations underpins rapid electron transfer to electron transferring flavoprotein.
        J. Biol. Chem. 2005; 280: 30361-30366
        • Demmer J.K.
        • Bertsch J.
        • Oppinger C.
        • Wohlers H.
        • Kayastha K.
        • Demmer U.
        • Ermler U.
        • Muller V.
        Molecular basis of the flavin-based electron-bifurcating caffeyl-CoA reductase reaction.
        FEBS Lett. 2018; 592: 332-342
        • Demmer J.K.
        • Chowdhury N.P.
        • Selmer T.
        • Ermler U.
        • Buckel W.
        The semiquinone swing in the bifurcating electron transferring flavoprotein/butyryl-coA dehydrogenase complex from Clostridium difficile.
        Nat. Commun. 2017; 8: 1577
        • Thorpe C.
        Electron-Transferring flavoproteins.
        in: Müller F. Chemistry and Biochemistry of Flavoenzymes. CRC press, Boca Raton FL1991: 471-486
        • Jones M.
        • Basran J.
        • Sutcliffe M.J.
        • Grossmann J.G.
        • Scrutton S., N.
        X-ray scattering studies of Methylophilus methylotrophus (sp W(3)A(1)) electron-transferring flavoprotein - evidence for multiple conformational states and an induced fit mechanism for assembly with trimethylamine dehydrogenase.
        J. Biol. Chem. 2000; 275: 21349-21354
        • Roberts D.L.
        • Frerman F.E.
        • Kim J.J.
        Three-dimensional structure of human electron transfer flavoprotein to 2.1-A resolution.
        Proc. Natl. Acad. Sci. U. S. A. 1996; 93: 14355-14360
        • Roberts D.L.
        • Salazar D.
        • Fulmer J.P.
        • Frerman F.E.
        • Kim J.J.
        Crystal structure of paracoccus denitrificans electron transfer flavoprotein: Structural and electrostatic analysis of a conserved flavin binding domain.
        Biochemistry. 1999; 38: 1977-1989
        • Burgess S.G.
        • Messiha H.L.
        • Katona G.
        • Rigby S.E.J.
        • Leys D.
        • Scrutton N.S.
        Probing the dynamic interface between trimethylamine dehydrogenase (TMADH) and electron transferring flavoprotein (ETF) in the TMADH-2ETF complex: Role of the Arg-α237 (ETF) and Tyr-442 (TMADH) residue pair.
        Biochemistry. 2008; 47: 5168-5181
        • Dwyer T.M.
        • Zhang L.
        • Muller M.
        • Marrugo F.
        • Frerman F.E.
        The functions of the flavin contact residues αArg249 and βTyr16, in human electron transfer flavoprotein.
        Biochim. Biophys. Acta. 1999; 1433: 139-152
        • Watmough N.J.
        • Frerman F.E.
        The electron transfer flavoprotein: Ubiquinone oxidoreductases.
        Biochim. Biophys. Acta. 2010; 1797: 1910-1916
        • Sato K.
        • Nishina Y.
        • Shiga K.
        Purification of electron-transferring flavoprotein from Megasphaera elsdenii and binding of additional FAD with an unusual absorption spectrum.
        J. Biochem. 2003; 134: 719-729
        • Garcia Costas A.M.
        • Poudel S.
        • Miller A.-F.
        • J S.G.
        • Ledbetter R.N.
        • Fixen K.
        • Seefeldt L.C.
        • Adams M.W.
        • Harwood C.S.
        • Boyd E.S.
        • Peters J.W.
        Defining electron bifurcation in the electron transferring flavoprotein family.
        J. Bacteriol. 2017; 199e00440-17
        • Sato K.
        • Nishina Y.
        • Shiga K.
        Interaction between NADH and electron-transferring flavoprotein from Megasphaera elsdenii.
        J. Biochem. 2013; 153: 565-572
        • Herrmann G.
        • Jayamani E.
        • Mai G.
        • Buckel W.
        Energy conservation via electron-transferring flavoprotein in anaerobic bacteria.
        J. Bacteriol. 2008; 190: 784-791
        • Weidenhaupt M.
        • Rossi P.
        • Beck C.
        • Fischer H.-M.
        • Hennecke H.
        Bradyrhizobium japonicum possesses two discrete sets of electron transfer flavoprotein genes:fixA, fixB and etfS, etfL.
        Arch. Microbiol. 1996; 165: 169-178
        • Ledbetter R.N.
        • Garcia Costas A.M.
        • Lubner C.E.
        • Mulder D.E.
        • Tokmina-Lukaszewska M.
        • Artz J.H.
        • Patterson A.
        • Magnuson T.S.
        • Jay Z.J.
        • Duan H.D.
        • Miller J.
        • Plunkett M.H.
        • Hoben J.P.
        • Barney B.M.
        • Carlson R.P.
        • et al.
        The electron bifurcating FixABCX protein complex from Azotobacter vinelandii: Generation of low-potential reducing equivalents for nitrogenase catalysis.
        Biochemistry. 2017; 56: 4177-4190
        • Chowdhury N.P.
        • Mowafy A.M.
        • Demmer J.K.
        • Upadhyay V.
        • Koelzer S.
        • Jayamani E.
        • Kahnt J.
        • Hornung M.
        • Demmer U.
        • Ermler U.
        • Buckel W.
        Studies on the mechanism of electron bifurcation catalyzed by electron transferring flavoprotein (Etf) and butyryl-CoA dehydrogenase (Bcd) of Acidaminococcus fermentans.
        J. Biol. Chem. 2014; 289: 5145-5157
        • Duan H.D.
        • Lubner C.E.
        • Tokmina-Lukaszewska M.
        • Gauss G.H.
        • Bothner B.
        • King P.W.
        • Peters J.W.
        • Miller A.F.
        Distinct flavin properties underlie flavin-based electron bifurcation within a novel electron-transferring flavoprotein FixAB from Rhodopseudomonas palustris.
        J. Biol. Chem. 2018; 293: 4688-4701
        • Yang K.Y.
        • Swenson R.P.
        Modulation of the redox properties of the flavin cofactor through hydrogen-bonding interactions with the N(5) atom: Role of alpha Ser254 in the electron-transfer flavoprotein from the methylotrophic bacterium W3A1.
        Biochemistry. 2007; 46: 2289-2297
        • Nitschke W.
        • Russell M.J.
        Redox bifurcations: Mechanisms and importance to life now, and at its origin.
        Bioessays. 2012; 34: 106-109
        • Schut G.J.
        • Mohamed-Raseek N.R.
        • Tokmina-Lukaszewska M.
        • Mulder D.E.
        • Nguyen D.M.N.
        • Lipscomb G.L.
        • Hoben J.P.
        • Patterson A.
        • Lubner C.E.
        • King P.W.
        • Peters J.W.
        • Bothner B.
        • Miller A.F.
        • Adams M.W.W.
        The catalytic mechanism of electron bifurcating electron transfer flavoproteins (ETFs) involves an intermediary complex with NAD+.
        J. Biol. Chem. 2019; 294: 3271-3283
        • Sucharitakul J.
        • Buttranon S.
        • Wongnate T.
        • Chowdhury N.P.
        • Prongjit M.
        • Buckel W.
        • Chaiyen P.
        Modulations of the reduction potentials of flavin-based electron bifurcation complexes and semiquinone stabilities are key to control directional electron flow.
        FEBS J. 2020; 288: 1008-1026
        • Mohamed-Raseek N.
        • Duan H.D.
        • Mroginski M.A.
        • Miller A.F.
        Spectroscopic, thermodynamic and computational evidence of the locations of the FADs in the nitrogen fixation-associated electron transfer flavoprotein.
        Chem. Sci. 2019; 10: 7762-7772
        • Massey V.
        • Hemmerich P.
        Active site probes of flavoproteins.
        Biochem. Soc. Trans. 1980; 8: 246-257
        • Lostao A.
        • Gomez-Moreno C.
        • Mayhew S.G.
        • Sancho J.
        Differential stabilization of the three FMN redox forms by tyrosine 94 and tryptophan 57 in flavodoxin from Anabaena and its influence on the redox potentials.
        Biochemistry. 1997; 36: 14334-14344
        • Bradley L.H.
        • Swenson R.P.
        Role of hydrogen bonding interactions to N(3)H of the flavin mononucleotide cofactor in the modulation of the redox potentials of the Clostridium beijerinckii flavodoxin.
        Biochemistry. 2001; 40: 8686-8695
        • Talfournier F.
        • Munro A.W.
        • Basran J.
        • Sutcliffe M.J.
        • Daff S.
        • Chapman S.K.
        • Scrutton N.S.
        Alpha Arg-237 in Methylophilus methylotrophus (sp. W3A1) electron-transferring flavoprotein affords approximately 200-millivolt stabilization of the FAD anionic semiquinone and a kinetic block on full reduction to the dihydroquinone.
        J. Biol. Chem. 2001; 276: 20190-20196
        • Peters J.W.
        • Miller A.F.
        • Jones A.K.
        • King P.W.
        • Adams M.W.
        Electron bifurcation.
        Curr. Opin. Chem. Biol. 2016; 31: 146-152
        • Sato K.
        • Nishima Y.
        • Shiga K.
        In vitro assembly of FAD, AMP, and the two subunits of electron-transferring flavoprotein: An important role of AMP related with the conformational change of the apoprotein.
        J. Biochem. 1997; 121: 477-486
        • Kayastha K.
        • Vitt S.
        • Buckel W.
        • Ermler U.
        Flavins in the electron bifurcation process.
        Arch. Biochem. Biophys. 2021; 701: 108796
        • Vigil J.,W.
        • Niks D.
        • Franz-Badur S.
        • Chowdhury N.
        • Buckel W.
        • Hille R.
        Spectral deconvolution of redox species in the crotonyl-CoA-dependent NADH:ferredoxin oxidoreductase from Megasphaera elsdenii. A flavin-dependent bifurcating enzyme.
        Arch. Biochem. Biophys. 2021; 701: 108793
        • Stanley R.J.
        Advances in flavin and flavoprotein optical spectroscopy.
        Antiox Redox Sign. 2001; 3: 847-866
        • Kar R.K.
        • Chasen S.
        • Mroginski M.A.
        • Miller A.F.
        Tuning the quantum chemical properties of flavins via modification at C8.
        J. Phys. Chem. B. 2021; 125: 12654-12669
        • Yagi K.
        • Ohishi N.
        • Nishimoto K.
        • Choi J.D.
        • Song P.-S.
        Effect of hydrogen bonding on electronic spectra and reactivity of flavins.
        Biochemistry. 1980; 19: 1553-1557
        • Cerda J.F.
        • Koder Jr., R.L.
        • Lichtenstein B.R.
        • Moser C.M.
        • Miller A.-F.
        • Dutton P.L.
        Hydrogen bond-free flavin redox properties: Managing flavins in extreme aprotic solvents.
        Org. Biomol. Chem. 2008; 6: 2204-2212
        • Mayhew S.G.
        The effects of pH and semiquinone formation on the oxidation-reduction potentials of flavin mononucleotide: A reappraisal.
        Eur. J. Biochem. 1999; 265: 698-702
        • Massey V.
        A simple method for determination of redox potentials.
        in: Curti B. Ronchi S. Zanetti G. Flavins and Flavoproteins. Walter de Gruyer, Berlin1991: 59-66
        • Clark W.M.
        Oxidation-Reduction Potentials of Organic Systems.
        Williams and Wilkins, Baltimore MD1960
        • Massey V.
        • Palmer A.G.
        Charge transfer complexes of lipoyl dehydrogenase and free flavins.
        J. Biol. Chem. 1962; 237: 2347-2358
        • He T.-F.
        • Guo L.
        • Guo X.
        • Chang C.-W.
        • Wang L.
        • Zhong D.
        Femtosecond dynamics of short-range protein electron transfer in flavodoxin.
        Biochemistry. 2013; 52: 9120-9128
        • Abramovitz A.S.
        • Massey V.
        Interaction of phenols with old yellow enzyme. Physical evidence for charge-transfer complexes.
        J. Biochel. Chem. 1976; 251: 5327-5336
        • Augustin P.
        • Toplak M.
        • Fuchs K.
        • Gerstmann E.C.
        • Prassl R.
        • Winkler A.
        • Macheroux P.
        Oxidation of the FAD cofactor to the 8-formyl-derivative in human electron-transferring flavoprotein.
        J. Biol. Chem. 2018; 293: 2829-2840
        • Robbins J.M.
        • Geng J.
        • Barry B.A.
        • Gadda G.
        • Bommarius A.S.
        Photoirradiation generates an ultrastable 8-formyl FAD semiquinone radical with unusual properties in formate oxidase.
        Biochemistry. 2018; 57: 5818-5826
        • Robbins J.M.
        • Souffrant M.G.
        • Hamelberg D.
        • Gadda G.
        • Bommarius A.S.
        Enzyme-mediated conversion of flavin adenine dinucleotide (FAD) to 8-formyl FAD in formate oxidase results in a modified cofactor with enhanced catalytic properties.
        Biochemistry. 2018; 56: 3800-3807
        • Schuman M.
        • Massey V.
        Purification and characterization of glycolic acid oxidase from pig liver.
        Biochim. Biophys. Acta. 1971; 227: 500-520
        • Macheroux P.
        • Massey V.
        • Thiele D.J.
        Expression of spinach glycolate oxidase in Saccharomyces cerevisiae: Purification and characterization.
        Biochemistry. 1991; 30: 4612-4619
        • Ku H.-K.
        • Lim H.-M.
        • Oh K.-H.
        • Yang H.-J.
        • Jeong J.-S.
        • Kim S.-K.
        Interpretation of protein quantitation using the Bradford assay: Comparison with two calculation models.
        Anal. Biochem. 2013; 434: 178-180
        • Jang M.-H.
        • Scrutton N.S.
        • Hille R.
        formation of W3A1 electron-transferring flavoprotein (ETF) hydroquinone in the trimethylamine dehydrogenase·ETF protein complex.
        J. Biol. Chem. 2000; 275: 12546-12552
        • Sucharitakul J.
        • Buckel W.
        • Chaiyen P.
        Rapid kinetics reveal surprising flavin chemistry in the bifurcating electron transfer flavoprotein from Acidaminococcus fermentans.
        J. Biol. Chem. 2021; 296: 100124
        • Lehman T.C.
        • Thorpe C.
        A new form of mammalian electron transfer flavoprotein.
        Arch. Biochem. Biophys. 1992; 292: 594-599
        • Mayhew S.G.
        • Whitfield C.D.
        • Ghisla S.
        • S.-J. M.
        Identification and properties of new flavins in electron-transferring flavoprotein from Peptostreptococcus elsdenii and pig-liver glycolate oxidase.
        Eur. J. Biochem. 1974; 44: 570-591
        • Ghisla S.
        • Massey V.
        • Mayhew S.G.
        Studies on the active centers of flavoproteins: Binding of 8-hydroxy-FAD and 8-hydroxy-FMN to apoproteins.
        in: Singer T.P. Flavins and Flavoproteins. Elsevier, Amsterdam1976: 334-340
        • O’Neill H.
        • Mayhew S.G.
        • Butler G.
        Cloning and analysis of the genes for a novel electron-transferring flavoprotein from Megasphaera elsdenii.
        J. Biol. Chem. 1998; 273: 21015-21024
        • Chowdhury N.P.
        • Kahnt J.
        • Buckel W.
        Reduction of ferredoxin or oxygen by flavin-based electron bifurcation in Megasphaera elsdenii.
        FEBS J. 2015; 282: 3149-3160
        • Sato K.
        • Nishina Y.
        • Shiga K.
        Electron-transferring flavoprotein has an AMP-binding site in addition to the FAD-binding site.
        J. Biochem. 1993; 114: 215-222
        • DuPlessis E.R.
        • Rohlfs R.J.
        • Hille R.
        • Thorpe C.
        Electron-transferring flavoproteins from pig and the methylotrophic bacterium W3A1 contains AMP as well as FAD.
        Biochem. Mol. Biol. Int. 1994; 32: 195-199
        • Feng X.
        • Schut G.J.
        • Lipscomb G.
        • Li H.Y.
        • Adams M.W.W.
        Cryoelectron microscopy structure and mechanism of the membrane-associated electron-bifurcating flavoprotein Fix/EtfABCX.
        Proc. Natl. Acad. Sci. U. S. A. 2021; 118e2016978118
        • Duan H.D.
        • Mohamed-Raseek N.
        • Miller A.F.
        Spectroscopic evidence for direct flavin-flavin contact in a bifurcating electron transfer flavoprotein.
        J. Biol. Chem. 2020; 295: 12618-12634
        • Yousafzai F.K.
        • Eady R.R.
        Dithionite reduction kinetics of the dissimilatory copper-containing nitrite reductase of alcalegenes xylosoxidans : The SO radical binds to the substrate binding type 2 copper site before the type 2 copper is reduced.
        J. Biol. Chem. 2002; 277: 34067-34073
        • Miller A.F.
        • Duan H.D.
        • Varner T.A.
        • Mohamed-Raseek N.
        Reduction midpoint potentials of bifurcating electron transfer flavoproteins.
        Meth. Enzymol. 2019; 620: 365-398
        • Waterhouse A.
        • Bertoni M.
        • Bienert S.
        • Studer G.
        • Tauriello G.
        • Gumienny R.
        • Heer F.T.
        • de Beer T.A.P.
        • Rempfer C.
        • Bordoli L.
        • Lepore R.
        • Schwede T.
        SWISS-MODEL: Homology modelling of protein structures and complexes.
        Nucleic Acids Res. 2014; 46: W296-W303
        • Huang J.
        • Rauscher S.
        • Nawrocki G.
        • Ran T.
        • Feig M.
        • de Groot B.L.
        • Grubmüller H.
        • MacKerell Jr., A.D.
        CHARMM36m: An improved force field for folded and intrinsically disordered proteins.
        Nat. Methods. 2017; 14: 71-73
        • Pettersen E.F.
        • Goddard T.D.
        • Huang C.C.
        • Couch G.S.
        • Greenblatt D.M.
        • Meng E.C.
        • Ferrin T.E.
        UCSF Chimera - a visualization system for exploratory research and analysis.
        J. Comput. Chem. 2004; 25: 1605-1612
        • Vdovin A.
        • Slenczka A.
        • Dick B.
        Electronic spectroscopy of lumiflavin in superfluid helium nanodroplets.
        Chem. Phys. 2013; 422: 195-203
        • Lei C.
        • Zhang Z.
        • Liu H.
        • Kong J.
        • Deng J.
        Biosensoring of hydrogen peroxide using new methylene blue N incorporated in a montmorillonite-modified horseradish peroxidase immobilization matrix as an electron shuttle.
        Analytica Chim. Acta. 1996; 332: 73-81
        • Oja S.M.
        • Guerrette J.P.
        • David M.R.
        • Zhang B.
        Fluorescence-enabled electrochemical microscopy with dihydroresorufin as a fluorogenic indicator.
        Anal. Chem. 2014; 86: 6040-6048
        • Twigg R.S.
        Oxidation-reduction aspects of resazurin.
        Nature. 1945; 155: 401-402
        • Husain M.
        • Stankovich M.T.
        • Fox B.G.
        Measurement of the oxidation-reduction potentials for one-electron and two-electron reduction of electron-transfer flavoprotein from pig liver.
        Biochem. J. 1984; 219: 1043-1047
        • Griffin K.J.
        • Dwyer T.M.
        • Manning M.C.
        • Meyer J.D.
        • Carpenter J.F.
        • Frerman F.E.
        alphaT244M mutation affects the redox, kinetic, and in vitro folding properties of Paracoccus denitrificans electron transfer flavoprotein.
        Biochemistry. 1997; 36: 4194-4202
        • Meng E.C.
        • Pettersen E.F.
        • Couch G.S.
        • Huang C.C.
        • Ferrin T.E.
        Tools for integrated sequence-structure analysis with UCSF Chimera.
        BMC Bioinformatics. 2006; 7: 339
        • Toogood H.S.
        • van Thiel A.
        • Basran J.
        • Sutcliffe M.J.
        • Scrutton N.S.
        Extensive domain motion and electron transfer in the human electron transferring flavoprotein.medium chain Acyl-CoA dehydrogenase complex.
        J. Biol. Chem. 2004; 279: 32904-32912