Redox Reactions of Reduced Flavin Mononucleotide (FMN

Sep 17, 2012 - Pacific Northwest National Laboratory, P.O. Box 999, MS K8-96, Richland, Washington 99352, United .... Chemical Geology 2018 476, 272-2...
1 downloads 0 Views 440KB Size
Article pubs.acs.org/est

Redox Reactions of Reduced Flavin Mononucleotide (FMN), Riboflavin (RBF), and Anthraquinone-2,6-disulfonate (AQDS) with Ferrihydrite and Lepidocrocite Zhi Shi,* John M. Zachara,* Liang Shi, Zheming Wang, Dean A. Moore, David W. Kennedy, and Jim K. Fredrickson Pacific Northwest National Laboratory, P.O. Box 999, MS K8-96, Richland, Washington 99352, United States S Supporting Information *

ABSTRACT: Flavins are secreted by the dissimilatory ironreducing bacterium Shewanella and can function as endogenous electron transfer mediators. To assess the potential importance of flavins in Fe(III) bioreduction, we investigated the redox reaction kinetics of reduced flavin mononucleotide (i.e., FMNH2) and reduced riboflavin (i.e., RBFH2) with ferrihydrite and lepidocrocite. The organic reductants rapidly reduced and dissolved ferrihydrite and lepidocrocite in the pH range 4−8. The rate constant k for 2-line ferrihydrite reductive dissolution by FMNH2 was 87.5 ± 3.5 M−1·s−1 at pH 7.0 in batch reactors, and k was similar for RBFH2. For lepidocrocite, k was 500 ± 61 M−1·s−1 for FMNH2 and 236 ± 22 M−1·s−1 for RBFH2. The surface area normalized initial reaction rates (ra) were between 0.08 and 77 μmol·m−2·s−1 for various conditions in stopped-flow experiments. Initial rates (ro) were first-order with respect to iron(III) oxide concentration, and ra increased with decreasing pH. Poorly crystalline 2-line ferrihydrite yielded the highest ra, followed by more crystalline 6-line ferrihydrite and crystalline lepidocrocite. Compared to a previous whole-cell study with Shewanella oneidensis strain MR-1, our findings suggest that the reduction of electron transfer mediators by the Mtr (i.e., metal-reducing) pathway coupled to lactate oxidation is rate limiting, rather than heterogeneous electron transfer to the iron(III) oxide.



INTRODUCTION Iron(III) oxides are important mineral phases in subsurface environments that are used by dissimilatory iron-reducing bacteria (DIRB) as terminal electron acceptors for anaerobic respiration (e.g., refs 1 and 2). The most bioavailable iron(III) oxides are disordered (ferrihydrite) or crystalline forms of nanometer size (lepidocrocite and goethite), the so-called bioavailable ferric oxides (BAFOs).3 The reasons for their bioavailability are varied, but include free energy (e.g., reduction potential and solubility), surface site concentration, and physical accessibility.4−7 Soluble electron transfer mediators (ETMs)8 (or electron shuttles) are one of the three primary mechanisms through which DIRB transfer electron equivalents to iron oxides.7,9,10 Biogenic ETMs are soluble, redox-active shuttle molecules that are generated by microorganisms in their reduced state and that diffuse to nearby BAFO surfaces for subsequent oxidation through heterogeneous electron transfer. To be effective, the oxidized ETM must be regenerated through a reversible electron transfer reaction at the bacterial surface. Biogenic ETMs include phenazines produced by Pseudomonas species11,12 and flavins produced by Shewanella species.13,14 The metal-reducing bacterium Shewanella onedensis MR-1 (MR-1) utilizes the Mtr (i.e., metal-reducing) pathway for transferring electrons from the periplasm across the outer © 2012 American Chemical Society

membrane to external electron acceptors, including iron(III) oxides.15 The Mtr pathway includes periplasmic (MtrA) and outer membrane (MtrC) multiheme c-type cytochromes and a β-barrel, porin-like protein (MtrB) that spans the outer membrane.16 MtrA, MtrB, and MtrC form an integral complex that enables electron transfer from the periplasm to external electron acceptors.15−18 It has recently been shown that Shewanella excretes flavins which contribute to the overall reduction of iron(III) oxides through function as an ETM13,14,17 and that the Mtr pathway is essential for maintaining them in the reduced state.19 This finding is consistent with observations that MR-1 is capable of reducing iron(III) oxides without direct contact.20−22 In spite of these observations, there are no published heterogeneous reaction rates for flavins with iron(III) oxides that can be used to quantify their potential role as ETMs in dissimilatory iron reduction. Here we determine the reaction kinetics of chemically reduced flavin mononucleotide (FMN) and riboflavin (RBF) with a series of BAFO minerals of different free energies and Received: Revised: Accepted: Published: 11644

April 18, 2012 August 20, 2012 September 17, 2012 September 17, 2012 dx.doi.org/10.1021/es301544b | Environ. Sci. Technol. 2012, 46, 11644−11652

Environmental Science & Technology

Article

Table 1. Chemical Structures, Reduction Half-Reactions, and Relevant Properties of FMN, RBF, and AQDS

pH 7.0, half oxidized species and half reduced species, 25 °C. E values were calculated using the Nernst equation, if not labeled with sources. Extinction coefficients at the maximum absorption wavelength (in brackets below) at pH 7.0. X = oxidized form of redox couples, and H2X = reduced form of redox couples. cReference 28. dReference 29. eReference 30. fReference 27. gMeasured in the present work. hReference 31. i Reference 32. a b

Preparation of Reduced Flavins and AQDS. The reduced forms of FMN (FMNH2) and RBF (RBFH2) were prepared by dithionite reduction with subsequent purification inside an anaerobic glovebox. Bioreduced AH2DS was prepared by incubating AQDS with S. oneidensis strain MR-1 at pH 7.0 with subsequent filtration.24 The biotic method was used for AQDS reduction because we have observed that abiotic procedures involving palladium catalyst and hydrogen gas tend to cause irreversible damage to the molecule. The dithionite reduction method is not suitable for AQDS because all products are fully soluble at all pH conditions, preventing removal of residual dithionite. The UV−vis spectra of FMNH2 and AH2DS (Figure S3, Supporting Information) were identical to published spectra.26,27 Reducing equivalent tests verified that the synthesized FMNH2, RBFH2, and AH2DS were ∼100% in the reduced form, with negligible concentrations of residual dithionite or sulfite (Figure S4, Supporting Information). Upon reoxidation by air, the concentrations of stock solutions were determined by UV−vis analysis with known extinction coefficients of FMN, RBF, or AQDS (Table 1). Details of the reduction, characterization, and reducing equivalent tests are provided in the Supporting Information (sections S3 and S4). Batch Experimental Design and Data Analysis. Batch experiments were conducted in the dark using 125 mL polypropylene bottles that were continuously stirred with Teflon-coated stir bars in an anaerobic glovebox. pH was maintained at 7.0 ± 0.03 using 30 mM PIPES buffer, with verification by pH measurement before and after reaction. Iron(III) oxide stock suspension was added to the PIPES buffer (∼100 mL) and equilibrated for 1 h before reductant addition. Less than 0.5 mL of organic reductant stock solution was added to the iron(III) oxide suspension to initialize the reaction. Subsamples, collected as the reaction proceeded, were

crystallinilities: 2-line ferrihydrite, 2-line Si ferrihydrite, 6-line ferrihydrite, and lepidocrocite under stringent anaerobic conditions. The well-studied synthetic ETM anthraquinone2,6-disulfonate (AQDS) was included as a reference compound. The redox reactions were rapid, and a combination of batch reactors and a stopped-flow apparatus were used to evaluate the effects of the ETM reduction potential, Fe(III) oxide surface area, and pH on the reductive dissolution rate.



MATERIALS AND METHODS For details regarding deoxygenated distilled deionized water (DDDW), chemicals, and the anaerobic glovebox, see the Supporting Information (section S1). Syntheses and Characterization of Iron(III) Oxides. Two-line ferrihydrite (FH), 6-line ferrihydrite, and lepidocrocite (LEP) were synthesized according to the methods of Schwertmann and Cornell.23 Two-line Si ferrihydrite (Si-FH; 1.67 mol % Si) was synthesized according to Zachara et al.24 Coprecipitated Si stabilizes ferrihydrite in the environment and retards oxidative recrystallization.25 Details of the syntheses, washing, and characterization are provided in the Supporting Information (section S2). Transmission electron microscopy (TEM) revealed that the 2-line FH and 2-line Si-FH particles were highly aggregated after drying for UHV analysis (Figure S1, Supporting Information). The size of the aggregates ranged from 10 to 200 nm. The more crystalline 6-line FH particles were dispersed crystallites less than 10 nm in diameter. The Brunauer−Emmett−Teller (BET) surface areas of the 2- and 6-line ferrihydrites were 230 and 280 m2·g−1 (Table S1, Supporting Information). Crystalline lepidocrocite particles (γFeOOH) were lathlike with lengths ranging from 100 to 300 nm, widths ranging from 10 to 50 nm, and a surface area of 130 m2·g−1 (Table S1). 11645

dx.doi.org/10.1021/es301544b | Environ. Sci. Technol. 2012, 46, 11644−11652

Environmental Science & Technology

Article

immediately filtered through 0.2 μm Nylon syringe filter membranes (Whatman) to quench the reaction. Subsequently, aqueous organic species were analyzed by UV−vis spectrophotometry (HP8452, Agilent) inside the glovebox, and FeIIaq was measured by ferrozine assay.33 For selected samples, total dissolved Fe was measured by inductively coupled plasma optical emission spectroscopy (ICP-OES; Perkin-Elmer). The concentrations of the aqueous organic species were calculated from the UV−vis measurements.34 The redox reaction between FMNH2 and lepidocrocite (γ-FeOOH) serves as an example:

Stopped-Flow Experimental Design and Data Analysis. The rapid initial reaction stage was studied using a stopped-flow apparatus (BioLogic SF-4). The experiments were performed inside an anaerobic glovebox with reductant or product monitoring by a BioLogic MOS-250 spectrometer. For redox reactions with FMNH2 and RBFH2, the increase in absorbance at 450 nm (λmax,FMN or λmax,RBF) was measured, which allowed us to calculate the production of oxidized FMN or RBF as the reaction progressed. For reactions with AH2DS, the decrease in absorbance at 386 nm (λmax,AH2DS) was measured. FeIIaq was not determined because of the rapidity of the experiment and volumetric constraints. The light scattering/absorbance effect of the iron(III) oxide particles on the total UV−vis absorbance needed to be considered for quantification. At any given time of reaction 1, the total absorbance (At) measured at 450 nm is the sum of the absorbances of FMN, FMNH2, and the remaining lepidocrocite particles. The absorbances of all other components at this wavelength are negligible. The FMN concentration can be calculated as

FMNH 2 + 2Fe IIIOOH(s, LEP) + 4H+ = FMN + 2Fe2 + + 4H 2O

(1)

At any given time t, the total absorbance (At) measured at the maximum absorption of FMN (λmax,FMN) is the sum of the absorbances of FMN and the remaining FMNH2. The concentration of FMN can be determined by [FMN] =

A t − [FMNH 2]o εFMNH2 εFMN − εFMNH2

[FMN] =

(2)

where [FMNH2]o is the initial FMNH2 concentration at time zero and εFMN and εFMNH2 are the extinction coefficients of FMN and FMNH2. The concentrations of RBF and AQDS/ AH2DS can be calculated with similar measurements and relationships. The redox reactions were fast; therefore, few kinetic data satisfied the initial 10% dissolution criteria,35 preventing the calculation of initial reaction rates from the batch reaction data. We consequently used a second-order kinetic model to fit the time-series kinetic data. For redox reaction 1, we assumed that the reaction rate (r) was first-order with respect to the FMNH2 and lepidocrocite concentrations: r=

d[FMN] d[Fe 2 +] = = k[FMNH 2][FeOOH] dt 2 dt

where [FeOOH]ο is the initially added lepidocrocite concentration and ε FeOOH is the “extinction coefficient” of lepidocrocite at 450 nm measured separately. The concentrations of RBF and AH2DS were calculated with the same procedure. See the Supporting Information (section S5) for details. The initial rates (ro, μM·s−1) of organic reductant electron transfer were calculated from the slopes of the initial 10% portion of the time course curves. The slopes corresponded to a least-squares fit of more than 20 points (R2 > 0.95). These data points were collected within a few seconds of reaction after the stopped-flow reactions were triggered. In our typical stopped-flow experiments with a 100 μM iron(III) oxide concentrations, the surface area concentration of the four iron(III) oxide suspensions ranged between 1.15 and 2.99 m2·L−1 (Table S1, Supporting Information). Surface area normalized initial rates (ra, μmol·m−2·s−1) were calculated as ro ra = (7) surface area

(3)

[FMN] =



1 [FeOOH]o [FMNH 2]o (e[FeOOH]o kt − e 2[FMNH2]o kt ) 2 1 [FeOOH]o e[FeOOH]o kt − [FMNH 2]o e 2[FMNH2]o kt 2

RESULTS AND DISCUSSION Reaction Thermodynamics. Iron(III) oxides have relatively low reduction potentials (E′ between −300 and +60 mV at pH 7.0), and therefore, their reductive dissolution at neutral pH under anoxic conditions requires strong reductants and/or chelating reagents (e.g., refs 12, 27, 36, and 37). The redox reaction between hydroquinone (E′ = 0.286 at pH 7.0) and amorphous iron(III) oxide, for example, is thermodynamically unfavorable above pH 5.1, and reductive dissolution does not occur.38 The FMNH2/FMN, RBFH2/RBF, and AH2DS/AQDS redox couples have low reduction potentials (Table 1), and their reduced forms are strong reductants. Upon exposure to air, FMNH2, RBFH2, and AH2DS are instantly oxidized. Under our reaction conditions, the reduction potentials of ferrihydrite and lepidocrocite at pH 7.0 are at least 80 mV higher than those of the three organic redox couples. Thermodynamically, all three

(4)

[Fe2 +] = 2[FMN] [FeOOH]o [FMNH 2]o (e[FeOOH]o kt − e 2[FMNH2]o kt ) 1 [FeOOH]o e[FeOOH]o kt 2

εFMN − εFMNH 2 − 2εFeOOH (6)

where k is the second-order reaction rate constant (M−1·s−1). Solving the differential eq 3 with respect to the initial conditions yielded [FMN] and [Fe2+]:

=

A t − [FMNH 2]o εFMNH 2 − [FeOOH]o εFeOOH

− [FMNH 2]o e 2[FMNH2]o kt (5)

where the subscript “o” means the initial condition at time zero. Equations 4 and 5 are valid only when [FeOOH]o is not equal to 2[FMNH2]o. k is the only unknown parameter in these equations, which can be obtained by fitting experimental data with eqs 4 and 5. Similar equations were derived for the other reductants and oxides. For details regarding the batch experimental design, kinetic model, and curve fitting, see the Supporting Information (section S5). 11646

dx.doi.org/10.1021/es301544b | Environ. Sci. Technol. 2012, 46, 11644−11652

Environmental Science & Technology

Article

Figure 1. FeIIaq, FMNaq, and RBFaq production as a function of time during the redox reaction of 50 μM iron(III) oxide (2-line FH or LEP) with 20 μM reduced flavins (FMNH2 or RBFH2) at pH 7.0 (30 mM PIPES buffer). Reaction time course data were acquired using batch reactors. Experimental data were fitted with a second-order reaction model (solid curves). Model fitting parameters are listed in Table S2 (Supporting Information).

group and they are able to form metal−ligand complexes,43 the metal-binding process is either slow or insufficiently strong to influence the dissolution of solid metal (hydr)oxide phases. The ratio of FeIIaq to FMNaq or RBFaq was about 2:1 (Figure 1), consistent with the following reaction stoichiometry:

reductants are capable of reducing both ferrihydrite and lepidocrocite over a wide pH range (Figure S5, Supporting Information). The reaction free energy for reductive dissolution increases as the pH decreases. Crystalline lepidocrocite has a lower reduction potential than ferrihydrite and exhibits a smaller thermodynamic driving force for reductive dissolution. Batch Reaction Products, Stoichiometry, and Kinetics. The batch reactors initially contained 20 μM reductant concentrations (FMNH2 or RBFH2) and 50 μM iron(III) oxide (2-line FH or lepidocrocite) at pH 7.0. At the end of the redox reaction, the 20 μM FMNH2 or RBFH2 was fully oxidized to 20 μM FMN or RBF, and nearly 39 μM FeIIaq was produced (Figure 1). The measured maximum Fe II aq concentrations were slightly less than 40 μM, likely due to Fe2+ adsorption by the residual 10 μM iron(III) oxide.12 The adsorption of oxidized FMN or RBF by the residual iron(III) oxide was negligible under these conditions. Isosbestic points at 288 and 334 nm indicated stoichiometric oxidation of RBFH2 to RBF without indication of irreversible degradation (Figure S6, Supporting Information; ref 39). The total dissolved Fe concentrations (FeIIaq + FeIIIaq) measured by ICP-OES were the same as that of FeIIaq measured by ferrozine assay, indicating that reduction was the dominant dissolution mechanism.40 There was no evidence for aqueous Fe(III) complexes within the limits of our detection. Neither ligand-controlled dissolution 41 nor ligand-promoted reductive dissolution42 was involved in the studied reactions. Although FMN and RBF have the metal-binding α-iminoketo functional

H 2X + 2Fe III(OH)3 (s, FH) + 4H+ = X + 2Fe 2 + + 6H 2O

(8)

H 2X + 2Fe IIIOOH(s, LEP) + 4H+ = X + 2Fe 2 + + 4H 2O

(9)

where H2X represents the organic reductant (FMNH2 or RBFH2) and X represents the corresponding oxidized product (FMN or RBF). Reaction time courses of ferrihydrite and lepidocrocite had distinctive characteristics. Reaction rates with 2-line FH quickly declined as the reaction progressed, while reaction rates with lepidocrocite did not decline and reached a plateau in 10 min. With ferrihydrite, redox reaction half-lives were less than 5 min, and more than 10% of the initially added ferrihydrite was reduced and dissolved within 30 s. Reactions with lepidocrocite were faster: redox reaction half-lives were less than 1.5 min, and more than 20% of the solid mass was reduced and dissolved within 30 s. A second-order reaction model was used to fit the time-series kinetic data for batch experiments performed at pH 7. 11647

dx.doi.org/10.1021/es301544b | Environ. Sci. Technol. 2012, 46, 11644−11652

Environmental Science & Technology

Article

Figure 2. Representative reaction time courses acquired using a stopped-flow apparatus. Reaction conditions: 100 μM 2-line ferrihydrite (FH) or 100 μM lepidocrocite (LEP) with 50 μM FMNH2 at pH 7.0 (30 mM PIPES buffer). (a, b) Raw data; absorbance at 450 nm increased as a function of time. (c, d) FMN concentration derived from raw data curves by using extinction coefficients of FMN, FMNH2, 2-line FH, or LEP at 450 nm wavelength. Each panel contains six or seven stacked time courses from replicate experiments, and each time course has 5000 data points. Inset: expansion of the first 5 s of the time course.

We expected that ferrihydrite would be more redox reactive than lepidocrocite because (1) the reduction potential of ferrihydrite was 164 mV higher than that of lepidocrocite at pH 7.0 and (2) the surface area concentration of the ferrihydrite suspension was 1.8-fold higher than that of lepidocrocite. In contrast, the k values for ferrihydrite with both FMNH2 and RBFH2 were smaller than those for lepidocrocite. It is possible that ferrihydrite aggregation changed the effective reactive surface area of this phase, reducing its reactivity below that of lepidocrocite at pH 7.0. It is also possible that there is an inhibitory mechanism in the ferrihydrite reaction: Fe(II) generated by heterogeneous electron transfer may catalyze recrystallization to lower free energy, less reactive iron(III) oxides.24,46 Perhaps the adsorption of the oxidized flavins to ferrihydrite at an early stage of reaction restricted additional site access. A stronger inhibitory effect might be expected for FMNH2 for previously described reasons. The “rate constant” k that we fitted with the time-series kinetic data are a combination of all factors mentioned above, which may not allow a straightforward mechanistic interpretation. Reaction Kinetics Acquired by the Stopped-Flow Apparatus. The rate of FMN production from the reaction of 2-line FH and FMNH2 at pH 7 was almost constant over the first 5 s and then quickly decreased as the reaction progressed (Figure 2). The rate of FMN production decreased to 0 within 20 s, although only 45% of the initially added FMNH2 was oxidized at this point. The reductive dissolution of 2-line FH was apparently inhibited after an initial rapid stage of reaction, consistent with batch experiment observations. The aqueous reductant and the iron(III) oxide cannot be stirred in the stopped-flow apparatus after mixing; therefore, the redox reaction at later times may be controlled by the diffusion of the aqueous reductant to reactive surface sites. The FMN

Heterogeneous reaction kinetics are more complicated than aqueous homogeneous reaction kinetics. Mineral surface characteristics and redox reactivities change with reaction progression.44,45 Even though the time course data were well-fit with a second-order model (R2 values > 0.97), caution is needed in the interpretation of the parameters. The second-order reaction rate constant values (k) fitted with the FMN or RBF data were only 5−8% smaller than the k values fit with the FeIIaq values (Figure S7, Supporting Information). An exception was the reaction between FMNH2 and lepidocrocite, where k from the FMN production data was 1.16 times larger. With ferrihydrite, FMNH2 and RBFH2 yielded similar k values, while with lepidocrocite, FMNH2 was 2-fold more reactive than RBFH2. Lepidocrocite yielded 5.1 times and 6.3 times higher k values for FMNH2 than ferrihydrite, in spite of a lower surface area loading (Table S1, Supporting Information). The reduction potential of FMNH2 is only 8 mV lower than that of RBFH2 at pH 7.0 (Table 1). If electron transfer were the rate-limiting step, the redox reactivities of FMNH2 and RBFH2 would be similar. We should not have observed a 2-fold increase in the k value of FMNH2 with lepidocrocite simply due to an 8 mV lower reduction potential. FMNH2 and RBFH2 have almost identical chemical structures with the same redoxactive moiety, except that FMNH2 has a phosphate group at the 5′ position (Table 1). FMN adsorption to hydroxylated mineral surfaces is greater than that of RBF because of this phosphate group (Figure S8, Supporting Information). The similar k values of FMNH2 and RBFH2 with ferrihydrite suggested that adsorption was not the rate-limiting step. However, both adsorption and electron transfer appeared to influence the reduction rates of more strongly sorbing FMNH2 with lepidocrocite. 11648

dx.doi.org/10.1021/es301544b | Environ. Sci. Technol. 2012, 46, 11644−11652

Environmental Science & Technology

Article

Figure 3. Log−log plots of initial rates (ro) as a function of the iron(III) oxide concentration: (a) 50 μM FMNH2 with various concentrations of iron(III) oxide; (b) 50 μM FMNH2, RBFH2, or AH2DS with various concentrations of 2-line ferrihydrite (FH); (c) 50 μM FMNH2, RBFH2, or AH2DS with various concentrations of lepidocrocite. All reaction time courses were acquired using a stopped-flow apparatus. Error bars represent 1 standard deviation from at least three replicate experiments.

⎛ d[X] ⎞ ⎟ = ko[H 2X]o x [Fe(OH)3 ]o y ro = ⎜ ⎝ dt ⎠ o

production plateau between 20 and 50 s is consequently not a kinetic saturation stage. The redox reaction between lepidocrocite and FMNH2 yielded a different type of time course curve (Figure 2). In the first 5 s, the rate of FMN production was constant and slower than with ferrihydrite. The FMN production rate slowly decreased as the reaction progressed, but it did not decrease to 0 as it did with ferrihydrite. To avoid potential diffusion and inhibitory effects in the later stage of reaction, and to make meaningful experimental comparisons, we decided to use the time course data up to the point where 10% of the initially added iron(III) oxide was reduced and dissolved, i.e., where 10% of the initially added organic reductant was converted to its oxidized form (dashed lines in Figure 2). Within this range, the curvature in time course curves was slight, allowing a linear fit of the kinetic data with the slope equal to the initial reaction rate. The initial rate method is well-suited to the kinetic analysis of experiments performed with similar procedures but differing in initial conditions.12,35,38,47,48 The reaction of FMNH2 with ferrihydrite yielded a 3-fold higher surface area normalized initial rate ra (3.4 ± 0.3 μmol·m−2·s−1) than the reaction with lepidocrocite (ra = 1.1 ± 0.2 μmol·m−2·s−1) (Table S3, Supporting Information). The ra values from this initial reaction stage evaluation appeared more reasonable than those observed in batch experiment because ferrihydrite is more reactive toward reductive dissolution than lepidocrocite.49 Effect of the Iron(III) Oxide Concentration on Reaction Rates. Mineral surfaces have various sites with dissimilar structures and reactivities.45 For a given solid-phase reactant, it is reasonable to assume that an increase in surface area concentration results in a proportional increase in reactive site concentration under the same solution conditions. Reaction rates should therefore be proportional to the surface area and suspended oxide concentration.38 The following kinetic relationships hold under the assumption that the initial rate (ro) for the redox reaction between organic reductant (H2X) and ferrihydrite (Fe(OH)3) is xth order with respect to the initial reductant concentration and yth order with respect to the ferrihydrite concentration:

(10)

where ko is the rate constant for the initial reaction and X is the corresponding organic product. Taking the logarithm yields log ro = log ko + x log[H 2X]o + y log[Fe(OH)3 ]o

(11)

Analogous equations can be written for the redox reaction with lepidocrocite: ⎛ d[X] ⎞ ⎟ = ko[H 2X]o x [FeOOH]o y ro = ⎜ ⎝ dt ⎠ o

(12)

log ro = log ko + x log[H 2X]o + y log[FeOOH]o

(13)

Figure 3 presents log−log plots for the reaction between FMNH2, RBFH2, and AH2DS with various ferrihydrite forms and lepidocrocite at pH 7. In the first set of experiments, the iron(III) oxide concentration was varied from 10 to 200 μM at fixed initial FMNH2 concentration (Figure 3a). The leastsquares fit of each data set yielded the following y values: 1.03 (2-line FH), 1.02 (2-line Si-FH), 1.05 (6-line FH), and 0.90 (lepidocrocite), with R2 > 0.99. These y values were close to 1, indicating that the initial reaction rates were first-order with respect to the iron(III) oxide concentration. These results also indicated that mineral aggregation effects were negligible in the initial stage of the stopped-flow experiments. The initial rate (ro) of 2-line Si-FH reductive dissolution was slightly higher than that of 2-line FH, but their rates were almost the same after surface area normalization (ra) (Figure S9, Supporting Information). At pH 7.0, the initial redox reactivity of ferrihydrite was not affected by 1.67% Si substitution, which is known to stabilize ferrihydrite against reductive recrystallization.23,24 Six-line FH and lepidocrocite yielded 3−7 times smaller ro values than 2-line FH. After surface area normalization, the ra values of 6-line FH and lepidocrocite were similar (Figure S9). The redox reactivity of the more crystalline 6-line FH was similar to that of wellcrystalline lepidocrocite, and both were less reactive than poorly crystalline 2-line FH. 11649

dx.doi.org/10.1021/es301544b | Environ. Sci. Technol. 2012, 46, 11644−11652

Environmental Science & Technology

Article

Figure 4. Surface area normalized initial rates (ra) as a function of pH: (a) 50 μM FMNH2 and 100 μM iron(III) oxide; (b) 50 μM RBFH2 and 100 μM iron(III) oxide; (c) 50 μM AH2DS and 100 μM iron(III) oxide. Four types of iron(III) oxides were studied: 2-line FH, 2-line Si-FH, 6-line FH, and lepidocrocite. All reaction time courses were acquired using a stopped-flow apparatus. The following pH buffers (30 mM) were employed: acetate (pH 4.0 and 5.0), PIPES (pH 6.1 and 7.0), and HEPES (pH 8.0). Error bars represent 1 standard deviation from at least three replicate experiments. The aqueous solubility of RBFH2 was below 100 μM when pH < 7, preventing comparative kinetic studies below that pH.

respectively. At pH 4.0, the ra with 2-line FH was 4.7 and 48 times higher than those with 6-line FH and lepidocrocite. The reaction order with respect to [H+] was determined by calculating the slopes between successive pairs of points in Figure 4.12,35 Except for FMNH2 with lepidocrocite or 6-line FH between pH 7.0 and pH 8.0, all other slopes were much smaller than 1. The reaction order with respect to [H+] was consequently less than 1 for most pH conditions. The slopes decreased with decreasing pH, indicating less pH dependence at the lower pH conditions. For the reaction between AH2DS and 2-line Si-FH, the slope was 0.05 between pH 4 and pH 5, indicating that the reaction was almost independent of pH change. Hydrogen ion activity influences reductive dissolution reactions in multiple ways. First, the thermodynamic driving force (reaction free energy) is larger at lower pH (Figure S5, Supporting Information). Second, H+ is a stoichiometric participant in the redox reaction (i.e., reactions 8 and 9). Third, the pH affects organic reactant speciation, the mineral surface protonation level, the extent of organic reductant adsorption,40,50 and the adsorption of reaction product Fe(II).27 The effect of the pH on the extent of organic reductant adsorption is not easily studied because of difficulties in resolving the adsorption and subsequent fast electron transfer steps by experiment. However, useful insights can be obtained through comparisons of the chemical structures and functional groups of the organic reductants. FMNH2 has a phosphate group, while AH2DS has two sulfonate groups. A decrease in pH should have a greater effect on the adsorption of phosphate as compared to sulfonate groups due to their different ionization behaviors and affinity for the surface. RBFH2 (or RBF) does not have the phosphate functional group; therefore, its adsorption behavior is quite different from that of FMNH2 (or FMN), as shown in Figure S8 (Supporting Information). Implications. Flavins that are secreted by Fe(III)-respiring Shewanella species in pure culture enhance the rate of solidphase Fe(III) reduction.13,14,19 The flavins, dominated by RBF and FMN, are present at submicromolar to micromolar concentrations in rich (Luria−Bertani, LB) medium.19 They are believed to function as ETMs when the organisms are in close proximity to the oxide surface (e.g., ∼1 μm19,51), with their reduced forms regenerated by the extracellular Mtr respiratory pathway.18,19 Exogenous RBF and FMN enhanced the reduction rates of ferrihydrite by S. oneidensis by a factor of 1.73 over AQDS at pH 7 (all at 12 μM with 4.5 mM ferrihydrite19).

In our second set of experiments, we determined the reaction kinetics of the three organic reductants with 2-line FH at multiple 2-line FH concentrations from 10 to 200 μM. The reaction rates of the three organic reductants were compared at a fixed 50 μM initial concentration (Figure 3b). The ro values were still first-order with respect to the 2-line FH concentration (y values between 0.97 and 1.13). The ro values for the reaction of FMNH2, RBFH2, and AH2DS with 2-line FH were similar, consistent with batch experiment observations. Increases in the iron(III) oxide concentration caused proportionate rate increases for all three reductants. In our last set of experiments, we varied the lepidocrocite concentration from 10 to 200 μM and compared the reactivity of the three organic reductants at a fixed 50 μM initial concentration (Figure 3c). As expected, the ro values were firstorder with respect to the lepidocrocite concentration (y values between 0.89 and 1.27). In contrast to 2-line FH, FMNH2 yielded the highest initial rates. The reaction with RBFH2 was 1.7-fold slower, and that for AH2DS reactions was 12 ± 6 times slower. The rate differences decreased as the concentration of lepidocrocite increased. Effect of the pH on Initial Reaction Rates. The redox reactions were sensitive to pH change, and the initial rate ra increased with decreasing pH (Figure 4). For FMNH2 (Figure 4a), the ra observed at pH 4.0 was 63 times higher than at pH 8.0 for 2-line FH and 18 times higher for 2-line Si-FH (Table S3, Supporting Information). The differences in the reactivity of the various iron(III) oxides with FMNH2 amplified with decreasing pH. The ra with 2-line FH was similar to that with 2line Si-FH at pH 7.0, with both being 3-fold higher than that with lepidocrocite. The initial rate with 2-line Si-FH and lepidocrocite became slower than that with 2-line FH as the pH was decreased. At pH 4.0, the ra with 2-line FH was 2.9, 7.3, and 34 times higher than those with 2-line Si-FH, 6-line FH, and lepidocrocite, respectively. The redox reactivity of 2-line ferrihydrite was affected by 1.67% Si substitution between pH 4 and pH 6. The reductive dissolution rate promoted by AH2DS was less sensitive to the pH than was FMNH2 (Figure 4c). A decrease in pH from 8.0 to 4.0 yielded a 24-fold increase in ra for reaction with 2-line FH and an 8.9-fold increase for 2-line Si-FH (Tabld S3, Supporting Information). The differences in reactivity between the various iron(III) oxides were slightly less sensitive to pH change than observed for FMNH2, but followed the same trend. At pH 7.0, the ra with 2-line FH was 3.7 and 32 times higher than those with 6-line FH and lepidocrocite, 11650

dx.doi.org/10.1021/es301544b | Environ. Sci. Technol. 2012, 46, 11644−11652

Environmental Science & Technology

Article

Our results revealed that the reduced flavins RBFH2 and FMNH2 are kinetically competent to function as ETMs, displaying intrinsic heterogeneous electron transfer rates comparable to but generally exceeding those for AH2DS under all conditions evaluated. The flavins are significantly more powerful reductants than AH2DS when the crystallinity of the iron(III) oxide increases above that of 2-line ferrihydrite, such as observed for lepidocrocite. The effect appears greater than that attributable to free energy differences alone, which are small between the three studied compounds. Our reduction rates for RBFH2 with ferrihydrite at pH 7 in a batch reaction system, when compared to the results of Coursolle et al.19 noted above, are at least 50 times more rapid than the observed ferrihydrite reduction rate by S. oneidensis in the presence of riboflavin with lactate as the electron donor. While there are both physiologic and chemical factors that influence this comparison, this result suggests that ETM reduction by the Mtr pathway coupled to lactate oxidation is rate limiting, rather than heterogeneous electron transfer to the iron(III) oxide. Moreover, our results suggest that all other things being equal (e.g., physiologic sensitivity to the pH), the ETM function of flavins in Shewanella−iron(III) oxide systems will increase with decreasing pH below 7.



(3) Hansel, C. M.; Benner, S. G.; Nico, P.; Fendorf, S. Structural constraints of ferric (hydr)oxides on dissimilatory iron reduction and the fate of Fe(II). Geochim. Cosmochim. Acta 2004, 68, 3217−3229. (4) Bonneville, S.; Van Cappellen, P.; Behrends, T. Microbial reduction of iron(III) oxyhydroxides: effects of mineral solubility and availability. Chem. Geol. 2004, 212, 255−268. (5) Pedersen, H. D.; Postma, D.; Jakobsen, R.; Larsen, O. Fast transformation of iron oxyhydroxides by the catalytic action of aqueous Fe(II). Geochim. Cosmochim. Acta 2005, 69, 3967−3977. (6) Roden, E. E.; Zachara, J. M. Microbial reduction of crystalline iron(III) oxides: influence of oxide surface area and potential for cell growth. Environ. Sci. Technol. 1996, 30, 1618−1628. (7) Weber, K. A.; Achenbach, L. A.; Coates, J. D. Microorganisms pumping iron: anaerobic microbial iron oxidation and reduction. Nat. Rev. Microbiol. 2006, 4, 752−764. (8) O’Loughlin, E. J. Effects of electron transfer mediators on the bioreduction of lepidocrocite (γ-FeOOH) by Shewanella putrefaciens CN32. Environ. Sci. Technol. 2008, 42, 6876−6882. (9) Hernandez, M. E.; Newman, D. K. Extracellular electron transfer. Cell. Mol. Life Sci. 2001, 58, 1562−1571. (10) Kappler, A.; Straub, K. L. Geomicrobiological cycling of iron. In Molecular Geomicrobiology; Banfield, J. F., Cervini-Silva, J., Nealson, K. H., Eds. Rev. Mineral. Geochem. 2005, 59, 85−108. (11) Hernandez, M. E.; Kappler, A.; Newman, D. K. Phenazines and other redox-active antibiotics promote microbial mineral reduction. Appl. Environ. Microbiol. 2004, 70, 921−928. (12) Wang, Y.; Newman, D. K. Redox reactions of phenazine antibiotics with ferric (hydr)oxides and molecular oxygen. Environ. Sci. Technol. 2008, 42, 2380−2386. (13) Marsili, E.; Baron, D. B.; Shikhare, I. D.; Coursolle, D.; Gralnick, J. A.; Bond, D. R. Shewanella secretes flavins that mediate extracellular electron transfer. Proc. Natl. Acad. Sci. U.S.A. 2008, 105, 3968−3973. (14) von Canstein, H.; Ogawa, J.; Shimizu, S.; Lloyd, J. R. Secretion of flavins by Shewanella species and their role in extracellular electron transfer. Appl. Environ. Microbiol. 2008, 74, 615−623. (15) Shi, L.; Rosso, K. M.; Clarke, T. A.; Richardson, D. J.; Zachara, J. M.; Fredrickson, J. K. Molecular underpinnings of Fe(III) oxide reduction by Shewanella oneidensis MR-1. Front. Microbiol. 2012, 3, 50. (16) Richardson, D. J.; Butt, J. N.; Fredrickson, J. K.; Zachara, J. M.; Shi, L.; Edwards, M. J.; White, G.; Baiden, N.; Gates, A. J.; Marritt, S. J.; Clarke, T. A. The ‘porin-cytochrome’ model for microbe-to-mineral electron transfer. Mol. Microbiol. 2012, 85, 201−212. (17) Ross, D. E.; Brantley, S. L.; Tien, M. Kinetic characterization of OmcA and MtrC, terminal reductases involved in respiratory electron transfer for dissimilatory iron reduction in Shewanella oneidensis MR-1. Appl. Environ. Microbiol. 2009, 75, 5218−5226. (18) Clarke, T. A; Edwards, M. J.; Gates, A. J.; Hall, A.; White, G. F.; Bradley, J.; Reardon, C. L.; Shi, L.; Beliaev, A. S.; Marshall, M. J.; Wang, Z. M.; Watmough, N. J.; Fredrickson, J. K.; Zachara, J. M.; Butt, J. N.; Richardson, D. J. Structure of a bacterial cell surface decaheme electron conduit. Proc. Natl. Acad. Sci. U.S.A. 2011, 108, 9384−9389. (19) Coursolle, D.; Baron, D. B.; Bond, D. R.; Gralnick, J. A. The Mtr respiratory pathway is essential for reducing flavins and electrodes in Shewanella oneidensis. J. Bacteriol. 2010, 192, 467−474. (20) Nevin, K. P.; Lovley, D. R. Mechanisms for accessing insoluble Fe(III) oxide during dissimilatory Fe(III) reduction by Geothrix fermentans. Appl. Environ. Microbiol. 2002, 68, 2294−2299. (21) Lies, D. P.; Hernandez, M. E.; Kappler, A.; Mielke, R. E.; Gralnick, J. A.; Newman, D. K. Shewanella oneidensis MR-1 uses overlapping pathways for iron reduction at a distance and by direct contact under conditions relevant for biofilms. Appl. Environ. Microbiol. 2005, 71, 4414−4426. (22) Peretyazhko, T. S.; Zachara, J. M.; Kennedy, D. W.; Fredrickson, J. K.; Arey, B. W.; McKinley, J. P.; Wang, C. M.; Dohnalkova, A. C.; Xia, Y. Ferrous phosphate surface precipitates resulting from the reduction of intragrain 6-line ferrihydrite by Shewanella oneidensis MR1. Geochim. Cosmochim. Acta 2010, 74, 3751−3767.

ASSOCIATED CONTENT

* Supporting Information S

Three additional tables and nine additional figures, plus accompanying text and additional details about the materials, experimental design and results, and model fitting. This material is available free of charge via the Internet at http:// pubs.acs.org.



AUTHOR INFORMATION

Corresponding Author

*Phone: (509) 371-6263 (Z.S.); (509) 371-6355 (J.M.Z.). Fax: (509) 371-6354 (Z.S.); (509) 371-6354 (J.M.Z.). E-mail: [email protected] (Z.S.); [email protected] (J.M.Z.). Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS This research was supported by the Geosciences Research Program of the Office of Basic Energy Science, U.S. Department of Energy. The contributions of J.K.F. and L.S. were supported by the Pacific Northwest National Laboratory Scientific Focus Area (PNNL SFA), which is funded by the Department of Energy’s Office of Biological and Environmental Research (BER). A portion of the experiments were performed at the Environmental Molecular Sciences Laboratory (EMSL), a national scientific user facility sponsored by the BER and located at PNNL. PNNL is operated for the Department of Energy by Battelle. We thank three anonymous reviewers for insightful comments.



REFERENCES

(1) Nealson, K. H.; Saffarini, D. Iron and manganese in anaerobic respirationenvironmental significance, physiology, and regulation. Annu. Rev. Microbiol. 1994, 48, 311−343. (2) Fredrickson, J. K.; Zachara, J. M.; Kennedy, D. W.; Dong, H. L.; Onstott, T. C.; Hinman, N. W.; Li, S. M. Biogenic iron mineralization accompanying the dissimilatory reduction of hydrous ferric oxide by a groundwater bacterium. Geochim. Cosmochim. Acta 1998, 62, 3239− 3257. 11651

dx.doi.org/10.1021/es301544b | Environ. Sci. Technol. 2012, 46, 11644−11652

Environmental Science & Technology

Article

(23) Schwertmann, U.; Cornell, R. M. Iron Oxides in the Laboratory: Preparation and Characterization; Wiley-VCH: Weinheim, Germany, 2000; pp 93−95 and 104−105. (24) Zachara, J. M.; Kukkadapu, R. K.; Peretyazhko, T.; Bowden, M.; Wang, C. M.; Kennedy, D. W.; Moore, D. A.; Arey, B. The mineralogic transformation of ferrihydrite induced by heterogeneous reaction with bioreduced anthraquinone disulfonate (AQDS) and the role of phosphate. Geochim. Cosmochim. Acta 2011, 75, 6330−6349. (25) Kukkadapu, R. K.; Zachara, J. M.; Fredrickson, J. K.; Kennedy, D. W. Biotransformation of two-line silica-ferrihydrite by a dissimilatory Fe(III)-reducing bacterium: formation of carbonate green rust in the presence of phosphate. Geochim. Cosmochim. Acta 2004, 68, 2799−2814. (26) Macheroux, P. UV-visible spectroscopy as a tool to study flavoproteins. In Flavoprotein Protocols; Chapman, S. K., Reid, G. A., Eds.; Methods in Molecular Biology, Vol. 131; Humana Press Inc.: Totowa, NJ, 1999; pp 1−7. (27) Liu, C. X.; Zachara, J. M.; Foster, N. S.; Strickland, J. Kinetics of reductive dissolution of hematite by bioreduced anthraquinone-2,6disulfonate. Environ. Sci. Technol. 2007, 41, 7730−7735. (28) Clark, W. M. Oxidation-Reduction Potentials of Organic Systems; The Williams & Wilkins Co.: Baltimore, MD, 1960. (29) Sober, H. A., Ed. Handbook of Biochemistry: Selected Data for Molecular Biology; The Chemical Rubber Publishing Co.: Cleveland, OH, 1968; pp J27−J34. (30) Whitby, L. G. A new method for preparing flavin-adenine dinucleotide. Biochem. J. 1953, 54, 437−442. (31) Draper, R. D; Ingraham, L. L. A potentiometric study of the flavin semiquinone equilibrium. Arch. Biochem. Biophys. 1968, 125, 802−808. (32) Conant, J. B.; Kahn, H. M.; Fieser, L. F.; Kurtz, S. S. An electrochemical study of the reversible reduction of organic compounds. J. Am. Chem. Soc. 1922, 44, 1382−1396. (33) Stookey, L. L. Ferrozinea new spectrophotometric reagent for iron. Anal. Chem. 1970, 42, 779−781. (34) Wang, Z. M.; Liu, C. X.; Wang, X. L.; Marshall, M. J.; Zachara, J. M.; Rosso, K. M.; Dupuis, M.; Fredrickson, J. K.; Heald, S.; Shi, L. Kinetics of reduction of Fe(III) complexes by outer membrane cytochromes MtrC and OmcA of Shewanella oneidensis MR-1. Appl. Environ. Microbiol. 2008, 74, 6746−6755. (35) Stone, A. T. Reductive dissolution of manganese(III/IV) oxides by substituted phenols. Environ. Sci. Technol. 1987, 21, 979−988. (36) Afonso, M. D.; Stumm, W. Reductive dissolution of iron(111) (hydr)oxides by hydrogen sulfide. Langmuir 1992, 8, 1671−1675. (37) Debnath, S.; Hausner, D. B.; Strongin, D. R.; Kubicki, J. Reductive dissolution of ferrihydrite by ascorbic acid and the inhibiting effect of phospholipid. J. Colloid Interface Sci. 2010, 341, 215−223. (38) Shi, Z.; Stone, A. T. PbO2(s, plattnerite) reductive dissolution by aqueous manganous and ferrous ions. Environ. Sci. Technol. 2009, 43, 3596−3603. (39) Cohen, M. D.; Fischer, E. Isosbestic points. J. Chem. Soc. 1962, 3044−3052. (40) Suter, D.; Banwart, S.; Stumm, W. Dissolution of hydrous iron(111) oxides by reductive mechanisms. Langmuir 1991, 7, 809− 813. (41) Kraemer, S. M. Iron oxide dissolution and solubility in the presence of siderophores. Aquat. Sci. 2004, 66, 3−18. (42) Afonso, M. D.; Morando, P. J.; Blesa, M. A.; Banwart, S.; Stumm, W. The reductive dissolution of iron oxides by ascorbate. The role of carboxylate anions in accelerating reductive dissolution. J. Colloid Interface Sci. 1990, 138, 74−82. (43) Albert, A. The metal-binding properties of riboflavin. Biochem. J. 1950, 47 (3), xxvii. (44) Schwertmann, U. Solubility and dissolution of iron oxides. Plant Soil 1991, 130, 1−25. (45) Stumm, W. Chemistry of the Solid-Water Interface; WileyInterscience: New York, 1992.

(46) Hansel, C. M.; Benner, S. G; Fendorf, S. Competing Fe(II)induced mineralization pathways of ferrihydrite. Environ. Sci. Technol. 2005, 39, 7147−7153. (47) Shi, Z.; Stone, A. T. PbO2(s, plattnerite) reductive dissolution by natural organic matter: reductant and inhibitory subfractions. Environ. Sci. Technol. 2009, 43, 3604−3611. (48) Erbs, J. J.; Gilbert, B.; Penn, R. L. Influence of size on reductive dissolution of six-line ferrihydrite. J. Phys. Chem. C 2008, 112, 12127− 12133. (49) Larsen, O.; Postma, D. Kinetics of reductive bulk dissolution of lepidocrocite, ferrihydrite, and goethite. Geochim. Cosmochim. Acta 2001, 65, 1367−1379. (50) Zinder, B.; Furrer, G.; Stumm, W. The coordination chemistry of weathering: II. Dissolution of Fe(III) oxides. Geochim. Cosmochim. Acta 1986, 50, 1861−1869. (51) Picioreanu, C.; Head, I. M.; Katuri, K. P.; van Loosdrecht, M. C. M; Scott, K. A. Computational model for biofilm-based microbial fuel cells. Water Res. 2007, 41, 2921−2940.

11652

dx.doi.org/10.1021/es301544b | Environ. Sci. Technol. 2012, 46, 11644−11652