Physiological Evidence for Isopotential Tunneling in the Electron Transport Chain of Methane-Producing Archaea

ABSTRACT

Many, but not all, organisms use quinones to conserve energy in their electron transport chains. Fermentative bacteria and methane-producing archaea (methanogens) do not produce quinones but have devised other ways to generate ATP. Methanophenazine (MPh) is a unique membrane electron carrier found in Methanosarcina species that plays the same role as quinones in the electron transport chain. To extend the analogy between quinones and MPh, we compared the MPh pool sizes between two well-studied Methanosarcina species, Methanosarcina acetivorans C2A and Methanosarcina barkeri Fusaro, to the quinone pool size in the bacterium Escherichia coli. We found the quantity of MPh per cell increases as cultures transition from exponential growth to stationary phase, and absolute quantities of MPh were 3-fold higher in M. acetivorans than in M. barkeri. The concentration of MPh suggests the cell membrane of M. acetivorans, but not of M. barkeri, is electrically quantized as if it were a single conductive metal sheet and near optimal for rate of electron transport. Similarly, stationary (but not exponentially growing) E. coli cells also have electrically quantized membranes on the basis of quinone content. Consistent with our hypothesis, we demonstrated that the exogenous addition of phenazine increases the growth rate of M. barkeri three times that of M. acetivorans. Our work suggests electron flux through MPh is naturally higher in M. acetivorans than in M. barkeri and that hydrogen cycling is less efficient at conserving energy than scalar proton translocation using MPh.

IMPORTANCE Can we grow more from less? The ability to optimize and manipulate metabolic efficiency in cells is the difference between commercially viable and nonviable renewable technologies. Much can be learned from methane-producing archaea (methanogens) which evolved a successful metabolic lifestyle under extreme thermodynamic constraints. Methanogens use highly efficient electron transport systems and supramolecular complexes to optimize electron and carbon flow to control biomass synthesis and the production of methane. Worldwide, methanogens are used to generate renewable methane for heat, electricity, and transportation. Our observations suggest Methanosarcina acetivorans, but not Methanosarcina barkeri, has electrically quantized membranes. Escherichia coli, a model facultative anaerobe, has optimal electron transport at the stationary phase but not during exponential growth. This study also suggests the metabolic efficiency of bacteria and archaea can be improved using exogenously supplied lipophilic electron carriers. The enhancement of methanogen electron transport through methanophenazine has the potential to increase renewable methane production at an industrial scale.

INTRODUCTION

The question of how cells obtain energy to grow is fundamental to biology. There are several ways cells metabolize substrates to produce energy in the form of ATP, such as by coupling thermodynamically favorable reactions to the production of a chemiosmotic gradient, substrate-level phosphorylation, and transmembrane cycles, etc. (1–10). Among the interesting organisms that can grow with very little ATP yield are methane-producing archaea (methanogens). Methanogens grow by reducing carbon substrates to methane gas in a process known as methanogenesis (11). Methanogens are strict anaerobes found in a wide range of natural habitats, from freshwater lake sediment to deep sea hydrothermal vents. In these environments, end products of anaerobic fermentation or respiration such as acetate and/or C₁ compounds (CO₂ and hydrogen, formate, CO, methanol, methylamines, and methylsulfides, etc.) are used as the substrates by methanogens. It is estimated that 4% of all net fixed carbon on Earth is remineralized to methane gas in the global carbon cycle (12). In comparison to heterotrophic bacteria such as Escherichia coli, methanogens obtain very little ATP for their metabolic lifestyle. E. coli aerobically growing on glucose obtains 32 to 38 mol ATP per mole substrate, whereas methanogens obtain between 0.5 to 2 mol ATP per mole substrate, depending on whether they have the ability to synthesize a membrane electron carrier, methanophenazine (MPh) (12, 13).

Most methanogens that have been isolated and characterized are specialists that use CO₂ and hydrogen or formate as the substrates for growth. However, the Methanosarcinales order can use a wide array of substrates in addition to CO₂ and formate (12). Methanosarcina species, such as Methanosarcina barkeri Fusaro, use CO₂ and hydrogen as well as C₁ compounds and acetate to grow. Methanosarcina acetivorans C2A is a close relative of M. barkeri, but it does not synthesize hydrogenases and is incapable of hydrogenotrophic methanogenesis. Regardless of the substrate, electrons from the reduced electron carriers ferredoxin, deazaflavin F₄₂₀, or hydrogen gas are used to reduce MPh, a unique membrane-bound electron carrier found so far only in Methanosarcinales methanogens (14, 15). MPh has a tricyclic ring phenazine head group and a prenyl tail connected by an ether linkage (Fig. 1a). Reduced MPh is oxidized by the cytochrome-containing terminal oxidoreductase, coenzyme M (CoM)-S-S-coenzyme B (CoB) heterodisulfide reductase, HdrED (Fig. 1b) (14, 16–18). CoM-S-S-CoB is the terminal electron acceptor synthesized in the last step of the methanogenesis pathway (19). The reduction of CoM-S-S-CoB back to the coenzyme M (mercaptoethanesulfonate) and coenzyme B (2-mercaptoheptanoylthreonine phosphate) thiols results in scalar translocation of protons across the cell membrane, similar to a Q-loop mechanism found in organisms that use oxidative respiration (20). However, MPh has a lower redox potential (∼−150 mV for MPh versus −100 or +100 mV for quinones), which tailors the methanogen electron transport system to donate electrons to the CoM-S-S-CoB disulfide terminal acceptor (21).

FIG 1.

Methanophenazine is a membrane electron carrier in Methanosarcina. (a) Structure of methanophenazine (MPh). (b) MPh shuttles electrons between the MPh:CoM-S-S-CoB oxidoreductase HdrED and proton-pumping F₄₂₀:methanophenazine oxidoreductase (Fpo) or viologen-reducing MPh hydrogenase (Vho) in M. barkeri, or between HdrED and Fpo or sodium-pumping ferredoxin:MPh oxidoreductase (Rnf [Rhodobacter nitrogen fixation protein]) in M. acetivorans.

MPh is interesting from an evolutionary perspective, because it is structurally unique yet suggests the selective pressure for the thermodynamic superiority of utilizing a chemiosmotic membrane gradient to conserve energy is ancient (22). We wanted to compare the physiological similarities between MPh synthesized by Methanosarcina species and quinones synthesized by bacteria to explore the functional analogy between MPh and quinones (23–25).

RESULTS

Improved methods for methanophenazine extraction and chromatography.

If MPh plays the same role in methanogens as quinones do in bacteria, we would expect that MPh should be as abundant in methanogen cell membranes as quinones are in bacteria. To quantify MPh and quinones in genetically tractable Methanosarcina strains, we devised a procedure to efficiently extract both species from cell membranes. Careful quantification of MPh in cell membranes requires the use of an internal standard that possesses similar chemical characteristics yet can be distinguished from the molecule under study. It is important that both molecules have similar solvent behaviors to ensure that both molecules are extracted from the crude sample with similar yields. Taking these considerations into account, we opted to use menaquinone (MK₄) as an internal standard, because it is also a prenylated membrane electron carrier that is not synthesized by methanogen cells and can be commercially purchased (14, 24). To efficiently extract MPh and MK₄, we devised a protocol that uses smaller cell numbers (between 10⁷ and 10⁸) during small-scale batch growth than previous methods that require extraction from 10 g (dry weight) stationary-phase cells grown in a fermenter, adds hexane extraction, and eliminates an isooctane extraction step as previously described (see Table S1 in the supplemental material) (14). Extracted membrane carriers were then separated by reverse-phase C₁₈ column chromatography. The peak eluted after chromatography was collected and confirmed to have a ¹H nuclear magnetic resonance imaging (NMR) spectra consistent with the published spectra of MPh extracted from M. mazei (see Fig. S1) (14). Our procedure had a mean recovery of the MK₄ internal standard of 89% from technical replicates of exponentially growing and stationary-phase cultures, with a mean variance of 10.8%. The mean variance for MPh from technical replicates was 17% from exponentially growing and 8.9% from stationary-phase cultures, with a mean variance of 14.7% (Table S1).

Methanophenazine from closely related Methanosarcina species differs in prenyl saturation.

A wide variety of membrane electron carriers are synthesized in nature (26–29). To determine if similar but nonidentical MPh derivatives are synthesized by methanogens, we extracted lipophilic compounds from membranes of genetically tractable M. acetivorans and M. barkeri (Table 1). Reverse-phase chromatography showed that the elution profile of MPh from M. acetivorans is identical to the elution profile of the membrane electron carrier from M. mazei (14). However, the membrane electron carrier from M. barkeri eluted later, suggesting the membrane electron carrier in M. barkeri is slightly more hydrophobic (see Fig. S2) (30). The M. barkeri membrane electron carrier had an optical absorbance spectrum identical to that of MPh extracted from M. acetivorans (see Fig. S3b and d). As a change in the phenazine head group would be expected to change the optical absorbance spectrum, the difference between MPh in M. acetivorans and that in M. barkeri is therefore most likely due to a change in the prenyl tail saturation or length (30). To confirm this, MPh extracted from M. acetivorans and M. barkeri was subjected to liquid chromatography-tandem mass spectrometry (LC-MS/MS) to compare the molecular masses with the published mass for MPh from M. mazei (14). As anticipated, MPh from M. acetivorans is identical to that from M. mazei with a molecular mass of 541 Da (see Fig. S4a). The MPh extracted from M. barkeri has an observed molecular mass 2 Da smaller, which exactly matches the mass predicted for an MPh molecule that has only one saturation in the prenyl tail (Fig. S4b). We interpreted these results as an indication that MPh species synthesized by M. mazei and M. acetivorans have two saturated prenyl moieties on the pentaprenyl tail, whereas that from M. barkeri has only one saturated prenyl moiety, confirming the MPh prenyl tail saturation reported by others (31).

TABLE 1.

Organisms used

NB no.a	Organism	Genotype	Reference or source
3	Escherichia coli DH5α F′ lacIq	F′ proAB lacIq Δ(lacZ)M15 zzf::Tn10 (Tetr) fhuA2Δ(argF-lacZ)U169 phoA glnV44 ϕ80Δ(lacZ)M15 gyrA96 recA1 endA1 thi-1 hsdR17	New England BioLabs
34	Methanosarcina acetivorans C2A	Δhpt::ϕC31 int attP	59
32	Methanosarcina barkeri Fusaro	Δhpt::ϕC31 int attP	59

^aNB no., Buan laboratory strain collection number.

Archaea and bacteria adjust the quantity of membrane soluble electron carriers during batch growth.

We wanted to determine if methanogens regulate MPh content under conditions of changing substrate availability and population density. To do so, we compared MPh content per cell for M. acetivorans and M. barkeri incrementally during batch growth. Electron flux through MPh allows M. acetivorans to produce a scalar transmembrane ion gradient for ATP synthesis, which results in a higher metabolic efficiency than if electrons were simply donated to a cytoplasmic electron acceptor. Electrons can also flow through the faster albeit non-energy-conserving (nonbifurcating) heterodisulfide reductase (HdrABC) path (32). Therefore, the direction of electron flow through MPh or HdrABC has an effect on the growth rates of the organism and the population as a whole (32, 33).

If MPh synthesis was balanced with cell growth and was independent of population density, the number of MPh molecules per cell would not change regardless of the population density of the cell culture. If membrane carrier synthesis was upregulated, the result would be an increase in the average number of molecules per cell as the number of cells per ml in the culture increases. Conversely, if the cell adjusts energy metabolism in a cell density-dependent manner, then the number of electron carrier molecules per cell would be expected to decrease.

In addition, if the energy conservation mechanism required external diffusion of an electron carrier (such as hydrogen, sulfur compounds, flavins, or phenazines, etc.), the cell runs the risk of losing electrons to nearby competing cells. Because M. barkeri uses an obligate hydrogen cycle to conserve energy, at low cell densities during growth on methanol as the sole energy source, electrons are lost as H₂ to the surrounding medium before they can be efficiently captured by MPh-reducing hydrogenase (Vho) (34). Therefore, at low cell densities, cells might be expected to increase MPh pool sizes to maximize the rate of H₂ recapture by hydrogenase at the membrane. At higher cell densities, individual cells will lose electrons through H₂, but a near neighbor in the population could capture the H₂. Likewise, an H₂ molecule(s) from neighboring cells could be captured, and overall, there would be a net zero loss of electrons by the culture population. Accordingly, the M. barkeri MPh content per cell could be higher in exponentially growing low-cell-density cultures than in higher-density stationary-phase cultures to compensate for the loss of electrons by H₂ diffusion at a low cell density. M. acetivorans would not be expected to exhibit a population-specific MPh synthesis effect, because it does not express functional MPh-dependent hydrogenases (F₄₂₀H₂:MPh oxidoreductase [Fpo] or Vho). Therefore, despite the requirement for MPh by both M. acetivorans and M. barkeri, the two organisms could regulate MPh synthesis through different mechanisms based on their divergent energy conservation pathways.

We extracted MPh from M. acetivorans and M. barkeri grown on methanol as the sole carbon and energy sources during batch growth. Under these conditions, M. acetivorans and M. barkeri have similar growth rates, cell sizes, and shapes. Because Methanosarcina can produce gas vacuoles, we related the extracted MPh quantity to cell counts instead of biomass to account for changes in cell density that might occur. In M. acetivorans cultures, we observed that the MPh concentration per cell increased 1.67 ± 0.06-fold from the exponential to stationary phase (Fig. 2a). MPh content per cell increased linearly during exponential phase up to 4.10 × 10⁶ molecules per cell and sharply increased as cell counts decreased when cultures reached stationary phase. Therefore, two rates of MPh accumulation were observed, an exponential rate and a stationary rate (Fig. 3a). These data indicate that during exponential growth, the rate of MPh synthesis outstrips the rate of cell doubling (Fig. 3b). In stationary phase, the number of cells per ml decreases with time while the molecules of MPh per cell extracted increases. Possible explanations for the increase in MPh concentration per cell as the population stops dividing could be attributed to cell lysis and scavenging or to changing extractability (Fig. 3c). However, we cannot formally exclude the possibility that M. acetivorans cells may synthesize MPh at a higher rate in response to entering stationary phase in light of the reasonably good correlation coefficient between cell density and MPh molecules per cell (R² = 0.5215) (Fig. 3c).

FIG 2.

Archaea and bacteria regulate electron carrier density. (a) MPh extracted from M. acetivorans. (b) Quantification of monounsaturated (blue bars) and double-saturated MPh (orange bars) in M. barkeri cells grown on methanol. (c) Quantification of menaquinone (blue bars) and ubiquinone (orange bars) extracted from E. coli cells grown aerobically on LB glucose. (d) Total electron carrier density in cell membranes during exponential (blue bars) or stationary (orange bars) phase. Error bars represent standard deviations. exp, exponential growth phase; sta, stationary-phase cells. Mac, M. acetivorans; Mba, M. barkeri; Eco, E. coli. Student's two-tailed t tests: *, P < 0.01 versus exponentially growing cells; **, P < 0.01 versus stationary-phase M. acetivorans cells; #, P < 0.01 versus exponentially growing M. acetivorans cells.

FIG 3.

MPh synthesis by M. acetivorans during growth on methanol. (a) MPh synthesis is biphasic in M. acetivorans cultures (inset, corresponding growth curve). (b) MPh quantity per cell increases linearly with cell density during exponential growth. (c) In stationary phase, as culture density decreases (top right quadrant), MPh quantity per cell increases (top left quadrant).

When extracting MPh from M. barkeri cultures of different ages, we observed that the total MPh pool did not significantly change as culture turbidity increased, unlike what we observed in M. acetivorans cultures (Fig. 2b). Second, we noticed that unlike M. acetivorans, M. barkeri synthesizes two forms of MPh: a double-saturated form and the historically known monounsaturated form (Fig. 4). In exponential phase, the M. barkeri MPh pool is equally composed of double- and monosaturated forms of the electron carrier. However, as cultures enter stationary phase, the representation of each is drastically changed and is instead dominated (93%) by the more reduced double-saturated form of MPh. (Fig. 2b). We did not observe a biphasic increase in MPh content as cells entered stationary phase. Our data suggest M. barkeri cells initially produce monosaturated MPh and that prenyl reduction to produce the double-saturated form of MPh may be conditionally regulated.

FIG 4.

Monounsaturated MPh is reduced to the double-saturated form in M. barkeri cells during growth on methanol. (a) Inverse correlation between total and unsaturated MPh quantities in cells. (b) Positive correlation between total and saturated MPh amounts. (c) Inverse correlation between saturated and unsaturated MPh quantities in the cell.

Altogether, these data indicate M. acetivorans increases MPh content as the substrate availability decreases and population density increases, while M. barkeri keeps the total MPh content steady (though prenyl tail oxidation may be regulated with growth phase). The data support a model of balanced growth and constant MPh synthesis for M. barkeri but not for M. acetivorans.

Next, we addressed whether our observation of membrane carrier regulation in cultures of Methanosarcina species was specific to methane-producing archaea or part of a more general prokaryotic phenomenon. To do so, we quantified membrane electron carriers in E. coli bacteria grown aerobically on rich medium supplemented with glucose as the carbon and energy sources and compared the amounts of menaquinone to those of ubiquinone during exponential and stationary phases (35–37). During aerobic exponential growth, E. coli synthesizes eight times as much ubiquinone (orange bars) as menaquinone (blue bars), while the total membrane carrier concentration increases 1.85 ± 0.03-fold from exponential to stationary phase (Fig. 2c). Not only does E. coli regulate metabolism by sensing redox states of ubiquinone and menaquinone but the overall electron carrier pool size increases as cultures enter stationary phase, similar to what we observed with methanogen cultures (38). Our data support the idea that some archaea and bacteria increase the amount of electron carriers in their membranes as they transition from exponential to stationary phase growth.

Possibility for isopotential electron tunneling in methanogen membranes.

What effect could changing the amount of MPh in the membrane have on cellular energetics? To answer this question, we used the measured dimensions of methanogen and E. coli cells to calculate the density of each electron carrier in the membrane. We then arranged the data according to energetics from the most enthalpically driven to the least enthalpically driven (Fig. 2d). In doing so, we noticed a pattern when comparing the relative ratios of quinones and MPh in exponentially growing versus stationary-phase cell populations. Intriguingly, the MPh content profile in M. acetivorans was similar to that in E. coli cultures grown aerobically on glucose Lysis broth (LB) while M. barkeri cells had the lowest average membrane carrier concentration (Fig. 2d).

We then calculated the theoretical rate of isopotential electron tunneling from one reduced MPh (or quinone) to a nearby oxidized electron carrier (Fig. 5a) (39–42). Based on the average number of membrane carrier molecules we measured, it is possible that electrons from one reduced MPh can tunnel to an adjacent MPh in M. acetivorans (42). In M. barkeri, tunneling may occur in stationary phase but is less probable when cultures are growing exponentially. This can be interpreted to suggest a reduced electron carrier must be physically close to the electron acceptor (such as the terminal oxidase) for electrons to be quickly and efficiently shuttled (Fig. 4b and c). If not, the reduced membrane carrier must diffuse randomly through the membrane until it comes in close proximity to an acceptor. Any reduced electron carriers that diffuse in the membrane risk either losing the electron to random off-target reactions or a decay in the energy of the electron being transferred. By contrast, for E. coli grown aerobically on LB or M. acetivorans, regardless of the distance between the electron-donating oxidoreductase and the terminal oxidase, electrons may be able to rapidly tunnel from one membrane carrier to another through the membrane with negligible energy loss until it reaches the terminal oxidase (Fig. 5d and e). Such a situation suggests some Methanosarcinales may have electrically “quantized” membranes, in which case the entire membrane could act as one electrically conductive surface, such as a copper metal sheet, rather than a surface studded with individual anode and cathode terminals separated by a nonconductive rubber insulator. Our calculations suggest that in M. acetivorans, electrons may be able to tunnel from any one membrane donor to any membrane acceptor at any point on the cell surface through a conductive MPh layer.

FIG 5.

Isopotential electron tunneling in membranes. (a) Total electron carrier concentrations in membranes were used to calculate the theoretical rate of electron transfer for each organism. Dark gray box, rates consistent with quantum tunneling; light gray box, rates where quantum tunneling is possible but less probable. (b) At low electron carrier density (yellow hexagons), electrons must diffuse from the donor (D) to the acceptor (A), resulting in localized electron transfer and slow kinetics. (c) Electrons (white) can diffuse through the membrane (black) to an acceptor within a short radius before decaying to a lower energy state (bar). (d) At high electron carrier density, electrons can tunnel throughout the membrane, resulting in the entanglement of all possible electron donors and acceptors and faster kinetic rates. (e) Electrons (white) can tunnel through the membrane (black) to a more distant acceptor without losing energy (bar). Eco, E. coli; Mac, M. acetivorans; Mba, M. barkeri; exp, exponential growth phase; sta, stationary-phase cells.

We reasoned that if some Methanosarcinales have electrically quantized membranes while others do not, the addition of a membrane-soluble (hydrophobic) electron carrier compound with structural and redox properties similar to those of MPh (such as phenazines) to cultures should enhance the growth rate of organisms such as M. barkeri, which may not produce an optimal amount of MPh, but will have a modest effect on M. acetivorans, which already produces enough MPh to electrically quantize the membrane. Phenazines can be used as extracellular electron shuttles or as toxins that kill cells by generating reactive oxygen species (ROS) that damage proteins and nucleic acids (43, 44). Toxicity in susceptible organisms requires enzymes in the electron transport chain to reduce the phenazines, which then react with molecular oxygen to produce ROS (45). In the experiments we describe, the cells were not exposed to oxygen and remained under strict anaerobic conditions at all times, thus eliminating the possibility of generating hydroxyl radicals from molecular oxygen. Consistent with our prediction, phenazine improved the growth rates and yields for M. acetivorans and M. barkeri, but the magnitude of the enhancement was more pronounced for M. barkeri (Fig. 6). When 10 μM phenazine was added to cultures, the M. acetivorans growth rate increased 37% versus a 94% increase in M. barkeri cultures (Fig. 6a). Recent reports also show that the addition of phenol red and phenazines to anaerobic digesters and microbial fuel cells does in fact increase the growth of Methanosarcina species, demonstrating that this phenomenon is generally relevant to biotechnology (46).

FIG 6.

Effects of MPh mimics on growth rate. Increasing concentrations of phenazine (a) or 5,10-dioxyphenazine (b) were added to methanol-grown cultures. Mac, M. acetivorans; Mba, Methanosarcina barkeri. Data were collected from six independent cultures per treatment. Error bars represent standard deviations. Student's two-tailed t tests versus no added phenazine or 5,10-dioxyphenazine condition: *, P < 0.05; **, P < 0.01.

To determine whether hydrophobicity was critical for growth rate enhancement, we tested whether soluble 2-hydroxyphenazine affects the growth of M. acetivorans. 2-Hydroxyphenazine closely mimics a hydroxyl moiety in place of the prenyl tail that can be used in place of authentic MPh in MPh H₂:CoM-S-S-CoB heterodisulfide reductase (HdrED) and F₄₂₀H₂:MPh oxidoreductase (Fpo) enzyme assays (15, 18, 47). We expected the more hydrophilic 2-hydroxyphenazine to have a weaker influence on growth than the more hydrophobic phenazine. Consistent with this hypothesis and with other reports, we observed that the addition of 10 μM soluble 2-OH-phenazine did not have an effect on growth rate (data not shown) (18). However, we were able to detect that 2-hydroxyphenazine resulted in a 48.5% decrease in MPh extracted from stationary-phase cultures (2.38 ± 0.29 × 10⁶ molecules per cell versus 4.90 ± 0.54 × 10⁶ molecules per cell). This observation, along with our aforementioned results, supports the idea that M. acetivorans regulates MPh biosynthesis, though whether the mechanism of regulation is through metabolic, transcriptional, or translational control remains an open question.

We next asked whether the oxidation state of phenazine added to cultures can affect growth. 5,10-Dioxyphenazine is a soluble, fully oxidized phenazine that we expected to inhibit growth by becoming reduced by electrons from the electron transport chain (45). 5,10-Dioxyphenazine has been shown to be a selective hypoxic trigger cytotoxin in human cell lines that is nontoxic when oxidized but produces ROS (hydroxyl radicals) that damage cells when reduced (48). As predicted, despite the lack of molecular oxygen in the medium, 5,10-dioxyphenazine was toxic to both M. acetivorans and M. barkeri with 50% inhibitory concentrations (IC₅₀s) of 1.4 μM for M. acetivorans and 0.2 μM for M. barkeri (Fig. 6b). Our results suggest that hydrophobic phenazines, which may be synthetic or naturally produced by bacteria, have the capacity to be reduced by the methanogen electron transport chain (49, 50).

DISCUSSION

For single cells to grow and divide at the maximum rate, the efficiency of electron transfer must surpass the rate of off-target redox reactions and minimize entropic decay through chemical bond vibration and heat dissipation. How have methanogens evolved to minimize energy dissipation while maximizing growth and respiration/fermentation rates? Phylogeny suggests the first methanogens and contemporary obligately hydrogenotrophic methanogens used hydrogen gas to reduce CO₂ or formate and do not use a membrane electron carrier to shuttle electrons from the electron donors (hydrogen and intracellular electron carriers such as deazaflavins F₄₂₀, FAD, and NAD) to the terminal electron acceptor, CoM-S-S-CoB heterodisulfide (12). Instead, they physically couple the methyl-viologen reducing hydrogenase (Mvh) to the electron-bifurcating terminal oxidase, CoM-S-S-CoB heterodisulfide reductase, HdrABC (51–53). Ergo, the bulk of electron transfer reactions in obligately hydrogenotrophic methanogens occur intracellularly through a network of iron/sulfur clusters and enzyme-bound redox cofactors. In vivo protein cross-linking experiments with the obligate hydrogenotroph Methanococcus maripaludis suggest methanogenesis enzymes form multienzyme complexes that spatially constrain electron donors and enzyme active sites to minimize the relative entropy between methanogenesis pathway steps (51, 53). Forming multienzyme complexes is kinetically and energetically ideal for optimizing any biochemical reactions, and methanogens use the same strategy to maximize growth and metabolism from hydrogen and carbon dioxide.

Methanosarcinales methanogens are postulated to have acquired the ability to use acetate and methylotrophic substrates through lateral gene transfer(s) with Clostridia and/or other bacteria (54–56). The ability to use new substrates opened up new trophic niches and allowed the cells to extract more chemical energy for growth but required the cells to find a new way to oxidize the electron carriers. Simply reversing the hydrogenase-catalyzed reactions, while thermodynamically allowable at low hydrogen concentrations, results in the loss of reducing power and makes the cells dependent on establishing a scalar proton transport mechanism to conserve energy.

One solution to the problem of overreduced electron carriers is to evolve and/or acquire hydrogenases and a membrane electron carrier to accelerate scalar proton transport and recapture electrons from hydrogen, resulting in a true hydrogen-cycling mechanism of energy conservation (34, 57). In Methanosarcinales such as M. barkeri and Methanosarcina mazei, MPh is used to couple electrons recaptured from hydrogen by hydrogenases to the HdrED terminal oxidase. Therefore, growth is dependent on the efficiency of recapture of electrons by hydrogenases and the rate that those electrons can be donated to HdrED. In Methanosarcinales that do not express hydrogenases, the electron carriers must be balanced by electron transfer through the membrane to generate a proton motive force by scalar proton transport. The MPh step is therefore rate limiting in the central metabolism of methylotrophic substrates or acetate regardless of whether the methanogen expresses hydrogenases (58).

M. acetivorans appears to have maximized the rate of electron transfer through the membrane by synthesizing sufficient amounts of MPh to enable isopotential electron tunneling. We also observed that when growth slows during stationary phase, M. acetivorans cells respond by increasing the membrane MPh pool size, which would have the effect of compensating for reduced electron flux by increasing the rate of electron transfer. While the prenyl tail saturation of MPh in M. barkeri changes, the total MPh membrane content is constant between cells in exponential and stationary phases. Overall, M. barkeri membranes do not contain as much MPh as M. acetivorans membranes. How can two closely related Methanosarcina species with similar growth rates have such different MPh membrane profiles? Does regulating electron flux through the cell membrane require the regulation of MPh biosynthetic genes to increase MPh pool sizes? Perhaps it does not. There are multiple biophysical strategies that can be employed by cells to modulate the rate of electron transport through the membrane. To maintain efficient electron transfer rates with low contents of electron carriers such as MPh molecules in M. barkeri and M. mazei and quinones in E. coli grown in anaerobic conditions, the cells would need to do one or several things: (i) utilize electron donor and acceptor proteins/enzymes so that each electron carrier molecule has only a short distance to diffuse; (ii) constrain the diffusion of enzyme donors and acceptors by localizing them near each other (such as by forming supramolecular complexes); and/or (iii) express isopotential membrane redox proteins to serve as electrical relays in place of MPh molecules. Depending on the evolutionary pressures and genomic content of a cell lineage, any one (or several) of these strategies could be employed to result in maximal energy conservation efficiency.

MATERIALS AND METHODS

Microorganisms and culture conditions.

Organisms were obtained from the sources listed in Table 1 (59). Methanogens were grown in high-salt mineral medium (HS) at 37°C as described previously (60). Escherichia coli cells were grown aerobically in 0.5% glucose Lysis broth (LB) with shaking at 37°C (61). 2-OH-phenazine was synthesized de novo (62). All other chemicals were obtained from Fisher. Culture growth was measured using a Spectronic D spectrophotometer fitted with a Balch tube (18 mm) modification or using a Tecan Sunrise UV-Vis spectrophotometric plate reader. Cell numbers were measured microscopically with a Hausmann cell counting chamber using an EVOS FL Auto cell imaging system microscope in the University of Nebraska-Lincoln (UNL) Morrison microscopy core facility. Cell dimensions were measured using NIH ImageJ software (63).

Quinone and methanophenazine extraction.

Cells from 100-ml cultures were harvested by centrifugation at 4,000 × g for 30 min in a Sorvall Legend XTR tabletop centrifuge fitted with a TX-750 swinging bucket rotor (Thermo Fisher Scientific). Cells were washed with 100 ml of 0.4 M NaCl and resuspended in 1 ml of 0.4 M NaCl. Menaquinone (MK₄) was added as an internal standard (Sigma-Aldrich). The following steps were carried out in the dark to avoid photodegradation. Resuspended cells were lysed by freezing at −80°C 1 h to overnight and then thawing at 37°C in a water bath. Ethanol was added to 70% and the lysates were vortexed and then heated to 70°C for 10 min. Lysates were transferred to glass Hungate tubes and extracted three times with 1 ml hexane. After each addition of hexane, samples were vortexed for 30 s and centrifuged at 2,000 × g for 10 min in a TX-750 swinging bucket rotor. Upper hexane organic phases containing MPh and the MK₄ internal standard were combined and transferred to clean Hungate tubes and evaporated under N₂ at 45°C in the dark.

High-performance liquid chromatography.

MPh was separated and quantified by reverse-phase high-performance liquid chromatography (HPLC) using an Agilent HPLC system equipped with a UV-Vis detector. Samples were injected into a Supelcosil LC-18 5-μm column (15 cm by 4.6 cm) fitted with a SupelGuard C₁₈ guard column (Sigma-Aldrich). Samples were analyzed by isocratic chromatography using either 95% methanol (MeOH) and 5% H₂O or 90% MeOH and 10% hexane mobile phase at 1 ml · min⁻¹. Table S1 in the supplemental material shows the method evaluation of the new MPh extraction procedure with a mobile phase of 90% MeOH and 10% hexane. For quantification of MPh and quinones, absorbances were measured at 255 nm for MPh detection and at 260 nm for quinones (14).

Mass spectrometry.

MPh samples were analyzed by LC-MS/MS on an AB SCIEX Q-Trap 4000 integrated mass spectrometer integrated with an Agilent 1200/Dionex U3000 nano LC system in the UNL metabolomics core facility.

Nuclear magnetic resonance imaging.

Isolated samples were dissolved in deuterated methanol. Proton NMR data were acquired on a Bruker DRX-500 with a TXI Cryoprobe (500.13 MHz proton frequency). The following data were gathered with a standard 30 degree pulse, 128 scans, and 65,536 data points: ¹H NMR(CD₃OD) 8.54 (1.5H, s), 8.21 (d, 1H, J = 8.7 Hz), 8.16 (d, 1H, J = 8.6 Hz), 8.11 (d, 1H, J = 9.3 Hz), 7.91 (t, 1H, J = 8.3 Hz), 7.87 (t, 1H, J = 8.3 Hz), 7.60 (dd, 1H, J = 9.3 and 2.5 Hz), 7.40 (d, 1H, J = 2.6 Hz), 5.14 (t, 1H, J = 7.0 Hz), 5.07 (t, 1H, J = 7.0 Hz), 4.31 (m, 2H), 3.59 (s, 2H), 2.15 to 1.90 (m, 20 H, side chains and impurity), 1.88 (CH₃CN), 1.67 to 1.53 (methyl side chains), 1.28 (impurity, “grease”), and 0.88 (impurity, grease) (64).