Abiotic Process for Fe(II) Oxidation and Green Rust ... - ACS Publications

Thermal intensification of microbial Fe(II)/Fe(III) redox cycling in a pristine shallow sand aquifer on the Canadian Shield. V.L. Shirokova , A.M.L. E...
0 downloads 0 Views 3MB Size
Article pubs.acs.org/est

Abiotic Process for Fe(II) Oxidation and Green Rust Mineralization Driven by a Heterotrophic Nitrate Reducing Bacteria (Klebsiella mobilis) Marjorie Etique,†,‡ Frédéric P. A. Jorand,*,†,‡ Asfaw Zegeye,†,‡,§ Brian Grégoire,†,‡ Christelle Despas,†,‡ and Christian Ruby†,‡ †

Université de Lorraine, Laboratoire de Chimie Physique et Microbiologie pour l’Environnement, UMR 7564, Institut Jean Barriol, Villers-lès-Nancy, F-54601, France ‡ CNRS, Laboratoire de Chimie Physique et Microbiologie pour l’Environnement, UMR 7564, Villers-lès-Nancy, F-54601, France S Supporting Information *

ABSTRACT: Green rusts (GRs) are mixed Fe(II)−Fe(III) hydroxides with a high reactivity toward organic and inorganic pollutants. GRs can be produced from ferric reducing or ferrous oxidizing bacterial activities. In this study, we investigated the capability of Klebsiella mobilis to produce iron minerals in the presence of nitrate and ferrous iron. This bacterium is well-known to reduce nitrate using an organic carbon source as electron donor but is unable to enzymatically oxidize Fe(II) species. During incubation, GR formation occurred as a secondary iron mineral precipitating on cell surfaces, resulting from Fe(II) oxidation by nitrite produced via bacterial respiration of nitrate. For the first time, we demonstrate GR formation by indirect microbial oxidation of Fe(II) (i.e., a combination of biotic/abiotic processes). These results therefore suggest that nitrate-reducing bacteria can potentially contribute to the formation of GR in natural environments. In addition, the chemical reduction of nitrite to ammonium by GR is observed, which gradually turns the GR into the end-product goethite. The nitrogen mass-balance clearly demonstrates that the total amount of ammonium produced corresponds to the quantity of bioreduced nitrate. These findings demonstrate how the activity of nitrate-reducing bacteria in ferrous environments may provide a direct link between the biogeochemical cycles of nitrogen and iron.



INTRODUCTION

nitrite toward the oxidation of Fe(II), have not been clearly defined.18 Among the iron hydroxide solids, green rusts (GRs) have been recognized as promising minerals in environmental remediation processes.19 GRs are Fe(II)−Fe(III) double hydroxysalt minerals composed of positively charged hydroxide layers separated by an interlayer of anions, An− (e.g., CO32−, SO42−, Cl−) and water molecules.20 GRs have the general formula {FeII(1−x)FeIIIx(OH)2}x+·{(x/n)An−mH2O}x−. The redox flexibility of GR allows many organic21 and inorganic pollutants22 to be reduced, with the Fe(II) content of the mineral determining whether a particular GR is a preferred electron donor or electron acceptor.23 GR’s ability to reduce relevant pollutants such as nitrate19 or nitrite,24 their environmental significance,22,25 and their potential role in remediation processes18 have naturally attracted growing interest from the scientific community. Thus, GRs are considered to play an important role in the iron biogeochemical

By taking part in the biogeochemical cycling of iron under both oxic and anoxic conditions, the microbial activity of ironoxidizing bacteria may play a major role in the geochemistry of the Earth’s crust.1−3 At neutral pH, many microorganisms (chemotrophic or phototrophic) can utilize Fe(II) as a source of electrons in oxic environments,4 whereas under anoxic conditions several bacteriadiscovered over the last 20 yearsare able to use Fe(II) as an electron donor.5−7 Among these bacteria, iron-oxidizing nitrate-reducing bacteria combine the oxidation of Fe(II) with the reduction of nitrate.8−11 Depending on the geochemical conditions, a variety of secondary iron minerals can be precipitated during this process, such as ferrihydrite,12,13 goethite (α-FeOOH),14 and mixed Fe(II)−Fe(III) minerals such as magnetite (Fe I I Fe I I I 2 O 4 ) 1 5 or carbon at ed green rust (e.g., FeII4FeIII2(OH)12CO3·3H2O).16 Due to their low solubility at circum-neutral pH, these secondary iron minerals precipitate rapidly in the immediate vicinity of cells, leading to their encrustation which apparently limits further cell growth and Fe(II) oxidation.14,17 However, the parameters that control the formation of these minerals, and the contribution of nitrate and © 2014 American Chemical Society

Received: Revised: Accepted: Published: 3742

July 29, 2013 March 3, 2014 March 8, 2014 March 8, 2014 dx.doi.org/10.1021/es403358v | Environ. Sci. Technol. 2014, 48, 3742−3751

Environmental Science & Technology

Article

Table 1. Data for Experiments Performed in the Presence of Various Concentrations of Nitrate, Fe(II), and Cell Densities, and with Unstarved Cells (Harvested Immediately at the End of the Exponential Growth Phase in an Oxic Rich Medium) or Prestarved Cells (2 Days Starvation in an Anoxic NaCl Medium)a Fe(II) oxidation rate constants exp. nb

stoichiometric ratio −

solid analysis

cells mL−1 × 108

r0 (nM s−1)

kobs (10−7 s−1)

ΔNO3 (mM)

ΔFe(II) (mM)

ΔNH4 (mM)

Rexp

Rexp(I)

6d

18 d

5.0 ± 0.2 r2 = 0.98 9.0 ± 0.2 r2 = 0.98 4.1 ± 0.3 r2 = 0.97 19.4 ± 0.4 r2 = 0.97 1.7 ± 0.4 r2 = 0.99 33.5 ± 0.2 r2 = 0.92 n.d. n.a.

−0.5

−3.5

+0.49

7.0 ± 0.4

6.0 ± 0.6

G

G

−1.2

−8.2

+1.24

6.8 ± 0.5

6.3 ± 0.2

G

G

−0.5

−3.8

+0.43

7.7 ± 0.4

5.8 ± 0.5

GR + ∼ G

G

−1.4

−10.8

+1.29

7.6 ± 0.3

6.7 ± 0.3

GR

G

−0.3

−1.8

+0.34

6.0 ± 0.3

6.0 ± 0.4

S, GR

G

−0.3

−2.1

+0.40

7.0 ± 0.5

5.8 ± 0.6

S, GR

GR

n.d. +0.3

n.d. −0.3

n.d. 0.02

n.d. n.a.

n.d. n.a.

GR ∼S

GR + G ∼S

#1

unstarved

13.7 ± 0.3

14 ± 2

#2

prestarved

8.6 ± 0.3

24 ± 4

#3

unstarved

13.7 ± 0.3

12 ± 3

#4

prestarved

8.6 ± 0.3

56 ± 7

#5

unstarved

2.8 ± 0.1

5±2

#6

unstarved

0.7 ± 0.1

20 ± 4

#7 control

unstarved no cells

2.8 ± 0.1 /

n.d. n.a.

+

Δx is the difference in the concentration of x at the beginning and end of the reaction (x: NO3−, NH4+, Fe(II)). Rexp corresponds to the ratio between ΔFe(II) and ΔNO3−. Rexp(I) is Rexp determined on the first four days. Solid analysis was performed at day 6 and day 18 by Raman spectroscopy, except #7 by TEM at days 7 and 24. The control was done without bacterial cells. Nitrite was analyzed but non-quantifiable in the initial and final media. All data are means of duplicate experiments (except #6, n = 4). See Table S1 for initial and final concentrations of nitrate, ammonium, and Fe(II). G = goethite, GR = green rust, S = siderite, n.a.= not applicable, n.d. = not determined. a

cycle26 and may be found in environments in which biooxidation and bioreduction of iron both occur.25,27,28 While the mechanism of GR formation is still under investigation,29 published data suggest that iron-oxidizing nitrate-reducing bacteria may play a significant role during the biological oxidation of Fe(II) in the presence of nitrate.16,29,30 These bacteria couple the biooxidation of Fe(II) with the reduction of nitrate with the final products assumed to be N2O or N2. However, the formation and the interaction of secondary iron minerals with biogenic N products, such as NO2−, have not yet been taken into account. Moreover, Fe(II) and GR species are well-known to be reactive with NO3− and NO2−.18,19,24 Some recent work has suggested that nitrite is often accumulated in the medium or in the vicinity of cells during Fe(II) oxidation by iron-oxidizing nitrate-reducing bacteria, such as various Acidovorax sp.5,14,31 In order to evaluate the role of nitrate and biogenic nitrite on abiotic Fe(II) oxidation and GR formation, we used Klebsiella mobilis as a model nitrate-reducing and nondenitrifying bacteria,32 since this bacterium is unable to enzymatically oxidize Fe(II).33 K. mobilis is an Enterobacteriaceae which can use several electron donors aerobically or anaerobically (e.g.: glucose, lactate, glycerol). It reduces nitrate to nitrite, which is actively expulsed from the cytoplasmic compartment.34 During growth in a rich medium, K. mobilis is able to store glycogen as an endogenous source of carbon and electrons, thus extending its viability and activity under starvation conditions.35 According to Morita,36 most of the biosphere is oligotrophic and most of the bacteria are in the starvation-survival mode. Therefore, the flux of electron donors is assumed to be very low. Thus, to simulate such an electron flux, we chose to incubate bacteria without exogenous organic carbon, and instead the bacteria should use their own storage of organic carbon (“endogenous organic electrons”). In this case, Fe(II) oxidation linked to the activity of K. mobilis is rather slow,

which enables identification of intermediate iron mineral phases formed during the process.



MATERIALS AND METHODS Preparation of the Resting Cell Suspension. Klebsiella mobilis (syn. Enterobacter aerogenes) was retrieved from lab stock (20% glycerol at −80 °C) and revived under oxic conditions on trypcase soy agar (TSA, BioMérieux). Trypcase soy broth (TSB, BioMérieux) was inoculated by colonies of this culture plate and incubated at 30 °C under orbital agitation (150 rpm). The cells were harvested in the early stationary growth phase after 24 h and centrifuged (10 000g at 22 °C for 10 min), washed twice, and resuspended in 9‰ NaCl to give a final cell density of 2.2 ± 0.5 × 1010 cells mL−1. The cell suspension was purged for 30 min with N2 (99.99%) and corresponds to the “unstarved cells” in the text below. The composition of the mineral medium was as follows: 2‰ NaCl (w/v), 4−40 mM NaNO3, 30 mM FeSO4·7H2O, 1.7 mM NH4Cl, 2.4 mM NaH2PO4, 7.3 mM NaHCO3, 1 mL L−1 vitamin solution (SL-10, DSMZ GmbH, 2010), and 1 mL L−1 trace elements solution (SL-10, DSMZ GmbH, 2010), and the pH was adjusted to 7.0 ± 0.2 with HCl or NaOH solution. A white fluffy precipitate appeared immediately after the addition of Fe(II), due to precipitation of Fe(II)-carbonates.37 A total of 80 mL of the mineral medium were filtered on a 0.2 μm membrane and dispensed into sterile flasks and crimp sealed with butyl rubber stoppers. The flasks were inoculated by the cell suspension, in duplicate or quadruplicate, with final concentrations ranging from 0.7 ± 0.1 × 108 to 13.7 ± 0.3 × 108 cells mL−1 and were incubated at 30 °C in the dark with 5.2 to 29 mM Fe(II) and 4.2 mM to 38.1 mM nitrate (Table 1, Table S1). Such nitrate concentrations can be found in polluted aquifers and other anthropogenic environments.38,39 No organic carbon was added as an electron donor (or carbon source) in order to keep cells at starvation. The dissolved 3743

dx.doi.org/10.1021/es403358v | Environ. Sci. Technol. 2014, 48, 3742−3751

Environmental Science & Technology

Article

Figure 1. Time courses of total (dissolved plus particular) Fe(II) (circles), nitrate (squares), and ammonium (diamonds) in an anoxic incubation medium with a resting cell suspension of Klebsiella mobilis. The graphs a, b, c, d, e, and f refer to the corresponding runs #1, #2, #3, #4, #5, and #6 from Table 1. Error bars are based on duplicate experiments. The vertical dashed line separates the two-step process where stage I is the reaction between Fe(II) and nitrite and stage II is the reaction between Fe(II) and nitrate.

Live/Dead BacLight viability kit (Molecular Probes, Invitrogen) to distinguish membrane-damaged bacteria (red fluorescing cells) from nondamaged bacteria (green fluorescing cells).40 Nitrate and nitrite concentrations were determined by ion chromatography (IC) after dilution of the sample in pure water (Purelab Option-Q, Elga LabWater, Antony, France) and filtering (0.22 μm, Acrodisc Supor, Gelman, Pall Corporation). IC analyses were carried out with a Metrohm 882 Compact IC plus instrument equipped with a high-pressure pump,

organic carbon content (NF EN 1484) of the medium inoculated by K. mobilis was 49 ± 1 mg L−1 (n = 3). Assays run with cells named below as “pre-starved cells” were obtained by the same method except that the cell suspension was left to stand for 48 h in 9‰ NaCl under N2 prior to inoculation. Cell Counts, Chemical Analyses, and Instrumentation. Cells were enumerated using an epifluorescence microscopy (BX51, Olympus) procedure using a combination of fluorochromes (SYTO 9 an propidium iodide) from the 3744

dx.doi.org/10.1021/es403358v | Environ. Sci. Technol. 2014, 48, 3742−3751

Environmental Science & Technology



sequential (Metrohm CO2 suppressor MCS) and chemical (Metrohm suppressor MSM II for chemical) suppression modules, and a conductivity detector. The separation was performed on a Metrosep ASupp5-250 column packed with poly(vinyl alcohol) particles functionalized with a quaternary ammonium group (5 μm particles diameter) and preceded by a guard column (Metrosep A supp 4/5 guard) and an RP 2Guard column to remove traces of organic compounds. The mobile phase consisted of a mixture of 3.2 mM Na2CO3 (SigmaAldrich, 99.5%) and 1 mM NaHCO3 (Sigma-Aldrich, 99.7%) in pure water. The flow rate was 0.7 mL min−1, and the sample loop volume was 20 μL. The detection limit and quantification for nitrite were respectively 0.1 μM and 0.3 μM. IC analysis throughout the incubation period did not reveal a quantifiable level of NO2−. This result indicates the oxidation of Fe(II) was not overestimated by an abiotic reaction between nitrite and Fe(II) during iron determination.41 Total (dissolved and particulate) Fe(II) and total Fe (Fe(II) + Fe(III)) (extractable with 2 M HCl) were analyzed after 1 h at 562 nm using the ferrozine method42 in an anoxic (N2/H2, 95:5) chamber (Coy Laboratory Product Inc., Grass Lake, MI, USA). This iron quantification protocol was compared to the revised ferrozine method of Klueglein and Kappler,41 and no significant changes were detected, which agrees closely with nitrite concentration