SSSAJ Journal of Natural Resources and Life Sciences Education
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (44)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Jones, C.A.
Right arrow Articles by Inskeep, W.P.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Jones, C.A.
Right arrow Articles by Inskeep, W.P.
GeoRef
Right arrow GeoRef Citation
Agricola
Right arrow Articles by Jones, C.A.
Right arrow Articles by Inskeep, W.P.
Soil Science Society of America Journal 64:600-608 (2000)
© 2000 Soil Science Society of America

DIVISION S-2-SOIL CHEMISTRY

Rates of Microbially Mediated Arsenate Reduction and Solubilization

C.A. Jones, H.W. Langner, K. Anderson, T.R. McDermott and W.P. Inskeep

Dep. of Land Resources and Environ. Sci., Montana State Univ., Bozeman, MT 59717 USA

binskeep{at}montana.edu


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 Materials and methods
 Results and discussion
 Conclusions
 REFERENCES
 
Reduction of arsenate [As(V)] to arsenite[As(III)] influences the mobility and toxicity of arsenic (As), yet the mechanisms controlling the rate of reduction in soils and natural waters are poorly understood. The goal of this study was to determine processes affecting reduction rates of both aqueous and sorbed phase As(V). Reduction experiments were conducted anaerobically in serum bottles with a range of glucose and As(V) concentrations. Serum bottles were inoculated with microorganisms extracted directly from an agricultural soil having naturally elevated concentrations of As (unenriched population), or with a pure culture isolate obtained from the same soil after enrichment for As(V) reduction. At As(V) concentrations ranging from 6 to 600 µM, the rate of As(V) reduction by the soil isolate was first order with respect to both As(V) concentration and microbial biomass. Reduction rates of As(V) with the soil isolate were 2 to10 fold greater than in the unenriched population, suggesting As(V) reducers represented only a subset of the unenriched population. Compiled data indicated that the pure culture isolate was fermenting glucose, and potentially reducing As(V) as a detoxification mechanism. In a parallel study, reduction rates of As(V) with the unenriched population were evaluated in the presence of goethite or ferrihydrite. When redox potential decreased from 500 to near 0 mV, aqueous As concentrations decreased by approximately 30% in a goethite suspension with a high As surface coverage, yet increased by seven fold in a goethite suspension with a low As surface coverage. In a ferrihydrite suspension, aqueous As concentrations during reduction increased approximately 100 fold faster than in a goethite suspension at similar initial aqueous As(V) concentrations, corresponding to differences in Fe oxide surface areas and reductive dissolution rates. The results indicate that rates of As mobilization during reduction in soils are highly dependent on oxide surface area and As surface coverage.

Abbreviations: bp, base pair • DGGE, denaturing gradient gel electrophoresis • EC, electrical conductivity • PCR, polymerase chain reaction • redox potential, EH


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
Materials and methods
 Results and discussion
 Conclusions
 REFERENCES
 
CHANGES IN REDOX POTENTIAL (EH) can significantly affect arsenic (As) toxicity and mobilization in contaminated soils, sediments, and natural waters by altering As valence state. More specifically, several studies have shown that the reduction of arsenate [As(V)] to arsenite[As(III)] generally increases As toxicity and mobility (Cullen and Reimer, 1989; Korte and Fernando, 1991). Although As(III) is not thermodynamically stable under oxidized conditions (Sadiq, 1997), concentrations of As(III) have been found to exceed As(V) concentrations in surface waters containing detectable concentrations of O2, suggesting nonequilibrium conditions (Aurillo et al., 1994; Abdullah et al., 1995; Sohrin and Matsui, 1997). These findings indicate that the rate of As(V) reduction exceeded the rate of As(III) oxidation in these surface waters. Prevalence of As(III) was generally correlated with high phytoplankton abundance, suggesting that the nonequilibrium conditions were microbially mediated. These results emphasize the importance of understanding biologically mediated processes affecting As(V)/As(III) cycling in natural systems.

Microbial reduction of As(V) to As(III) may occur through a process thought to be a detoxification mechanism or via dissimilatory reduction (respiration). Arsenic detoxification has been documented in Escherichia coli, Staphylococcus aureus, and Staphylococcus xylosis, and is controlled by ars genes that encode for As(V) reduction via an As(V) reductase, followed by As(III) removal from the cell with an efflux pump (Tamaki and Frankenberger, 1992; Cervantes et al., 1994). Dissimilatory reduction has been documented in several environmental isolates capable of coupling C oxidation with As(V) reduction (Ahmann et al., 1994; Laverman et al., 1995; Macy et al., 1996; Newman et al., 1997; Blum et al., 1998). Dissimilatory As(V) reduction rates with these isolates have been found to be as high as 10 mM d-1 at an initial As(V) concentration of 5 mM (Laverman et al., 1995). However, little is known about rates of microbial reduction in natural systems where As(V) is generally not a dominant electron acceptor and where As(V) reducers may represent only a fraction of the microbial community. Furthermore, it is unknown whether detoxification mechanisms represent an important microbially mediated pathway for As(V) reduction in soil systems.

The bioavailability and behavior of As in natural systems is strongly influenced by sorption to solid surfaces such as oxides of Mn, Al, and Fe. Detailed spectroscopic studies (x-ray absorption fine structure spectroscopy) have shown that both As(V) and As(III) sorb to Fe oxides via inner-sphere surface complexation (Waychunas et al., 1995; Fendorf et al., 1997; Manning et al., 1998). Past work demonstrates that As(III) does not sorb as strongly as As(V) at low As surface coverage (Pierce and Moore, 1982), while more recent work shows that As(III) sorbs more strongly than As(V) at higher As surface coverage (Manning et al., 1998; Raven et al., 1998; Sun and Doner, 1998). Because of differences in relative sorption strengths of As(III) and As(V), changes in soil EH have been found to alter aqueous As concentrations. For example, decreased soil EH during flooding generally increases aqueous As concentrations initially (Masscheleyn et al., 1991a; Onken and Hossner, 1995; McGeehan et al., 1998), while extended flooding may decrease aqueous As concentrations because of sorption of As(III) or precipitation of As-sulfide solid phases (Rittle et al., 1995). Although it is expected that reductive dissolution of Fe oxide phases plays an important role in the solubilization of As under reducing conditions, it is unclear whether sorption of As(V) limits the rate of As(V) reduction, because of lower bioavailability of sorbed phase As. Substrate bioavailability is often limited by sorption in situations where rates of desorption-diffusion reactions are lower than rates of substrate utilization.

In summary, mechanisms controlling rates of microbially mediated As(V)/As(III) cycling in soils and natural waters are poorly defined. We have initiated studies on As redox cycling with the overall goal to enhance predictability of As behavior in nature. The specific objectives of the current study were to (i) assess the effects of microbial growth, As enrichment, and As(V) concentration on microbial reduction rates of aqueous As(V), (ii) determine the influence of Fe oxides on reduction rates of aqueous and sorbed phase As(V), and (iii) evaluate the effect of reducing conditions on aqueous As concentrations in the presence of Fe oxides.


   Materials and methods
TOP
 ABSTRACT
 INTRODUCTION
 Materials and methods
Results and discussion
 Conclusions
 REFERENCES
 
Microbial Samples and Molecular Characterization
All microorganisms used in this study were obtained from the upper 20 cm of a fine loamy, frigid Typic Calciaquoll, an agricultural soil (designated C3-NI) that contained naturally elevated concentrations of both aqueous (20 µM) and total As (840 µmol kg-1; Jones, 1998; Jones et al., 1999). Microorganisms were extracted from C3-NI soil by the methods described by Kingsley and Bohlool (1981). Briefly, the soil was resuspended in a buffer composed of 0.1 M (NH4)2HPO4 and 0.1% (w/v) hydrolyzed gelatin (9.5:1 buffer:soil ratio), and shaken (120 cycles min-1) for 30 min. The slurry was clarified by flocculation with CaCl2, and the resulting supernatant centrifuged at 5000 x g for 10 min to yield the unenriched microbial population (CN-0). As(V) reducers from the CN-0 population were enriched in a nutrient solution (hereafter referred to as AR medium) that was modified from that described by Angle et al. (1991) and consisted of the following: NH4Cl (6.0 mM), MgCl2 (2 mM), CaSO4 (1 mM), KH2PO4 (5 mM), NaOH (2.5 mM), H2SO4 (1 mM), FeCl2 (5 mM), Sigma Select yeast extract (0.2 g L-1), and KHCO3 (20 mM). All reagents used in this study were either certified by the ACS or Baker Analyzed. The medium was supplemented with 100 mL L-1 of micronutrient solution (Skerman, 1967), 300 mM Na2HAsO4, and glucose (5 mM). The glucose concentration (equivalent to 30 mM as C) was selected to achieve 100% reduction of As(V) based on preliminary experiments, and was similar to concentrations of dissolved organic carbon (DOC) measured in soil C3-NI (17 mM C). All incubations were carried out in serum bottles containing a N2(g)-purged AR liquid medium. To isolate the numerically dominant As(V) reducer in these enrichments (referred to as CN-8), the seventh transfer was diluted to extinction with 1:10 dilutions from the most-dilute suspension verified to reduce As(V). Aliquots were diluted, spread on AR agar, and incubated at 25°C in an anaerobic chamber. Based on colony morphology, these incubations produced a single organism. For further characterization, this bacterium was purified by repeated subculture, with the final culture verified to be an As(V) reducing organism.

Denaturing gradient gel electrophoresis (DGGE) was used to characterize microbial populations present in the original C3-NI soil, CN-0, and CN-8. Total sample DNA was extracted by mixing 0.4 g of autoclaved acid-washed glass beads and 0.8 mL of Na-phosphate buffer (120 mM, pH = 8.0) with either 0.4 g soil or 0.4 mL of a microbial cell suspension for 45 s with a Mini-Bead Beater (BioSpec Products, Bartlesville, OK). All other methods were as outlined by More et al. (1994). Briefly, after NH4-acetate precipitation, the DNA was further purified by isopropanol-precipitation (1 h at 4°C) and centrifugation (13 000 x g) for 15 min. The resulting DNA pellet was washed with 70% (v/v) ethanol (-20°C), dried (65°C), and then resuspended in TRIS-EDTA buffer. Polymerase chain reaction (PCR) amplification of the 1070- to 1392-bp region of the 16S rRNA gene (Escherichia coli nucleotide map numbers) was achieved with the following two primers (sequences): 1070 forward (5'-ATG GCT GTC GTC AGC T-3') and 1392 reverse with GC clamp (5'-CGC CCG CCG CGC CCC GCG CCC GGC CCG CCG CCC CCG CCC CAC GGG-3'). The reaction mix consisted of Assay Buffer B (Fisher Scientific, Pittsburgh, PA), 2.5 mM MgCl2, bovine serum albumin (Fisher Scientific), 0.2 mM of each dNTP, 0.5 mM of each primer, 1 mL DNA sample, 1.25 units Taq polymerase (Fisher Scientific), and deionized H2O to bring final volume to 50 mL. The amplification was conducted with a Perkin Elmer Applied Biosystems (Foster City, CA) Gene-Amp 9700 Thermal Cycler (one cycle of 45 s at 94°C; 35 cycles of 1 min at 94°C, 45 s at 55°C, and 45 s at 70°C; and one cycle at 72°C for 7 min).

A DCode Universal Mutation Detection System (Bio-Rad, Hercules, CA) was used to resolve the PCR-amplified 16S rDNA fragments in an 8% (w/v) acrylamide gel containing a 35 to 80% (v/v) gradient of a urea-formamide solution comprised of 7 M urea/40% (v/v) formamide. The DGGE conditions were 80 V for 16 h at 60°C, with a running buffer comprised of 40 mM Tris, 20 mM acetic acid, 2 mM EDTA at pH 8.5. The DGGE gels were photographed with UV transillumination. For CN-8 only, the 1070- to 1392-bp region of the 16S rDNA gene was sequenced and the results compared with sequences contained in both the RDP database (Maidak et al., 1997) and GenBank (Benson et al., 1997).

Reduction Rates of Aqueous As(V)
Experiments designed to determine rates of As(V) reduction in the presence of CN-0 or CN-8 were conducted in serum bottles (70 mL) containing 50 mL of AR broth. The bottles were capped with butyl rubber septa, crimp-sealed, and purged with N2(g) (10 min at approximately 60 mL min-1). The bottles were subsequently acidified with 0.15 mL of 6 M HCl to attain a pH of 6.5, then sterilized by autoclaving. The bottles were aseptically inoculated with CN-0 or CN-8 to attain an initial cell density of 106 cells mL-1, as determined by an empirically developed relationship between cell enumeration on YEPG (yeast extract, peptone, glucose) agar and optical density of cell suspensions at 500 nm (A500). Experiments were conducted over a range of initial As(V) concentrations (6 µM–5 mM as Na2HAsO4) at a constant concentration of total C (30 mM) added as glucose. In a separate experiment with 300 µM As(V), As(V) reduction rates were verified to not be limited by glucose concentrations ranging from 10 to 100 mM as C (results not shown). Serum bottles were agitated on a horizontal shaker (120 cycles min-1) at 25°C, and aseptically sampled as a function of time. At each sampling time, microbial numbers were estimated by measuring optical density (A500), and a duplicate sample was filtered (0.22 µm) for analysis of As(V), As(III), and dissolved organic C (described below). Reduction rates of aqueous As(V) were determined by performing linear regressions between As(V) concentration and time for the initial linear phase of As(V) reduction curves.

The fate of glucose-C in the 0.6, 2, and 5 mM As treatments with CN-8 was evaluated by spiking each serum bottle with 14C-glucose to an initial specific activity of 1.7 x 102 Bq mL-1. Both filtered (0.2 µm) and unfiltered samples were acidified with concentrated HCl (1% v/v) and purged of CO2 prior to 14C analysis. Oxidized C was calculated as the difference in C between an initial sample (t = 0) and each unfiltered, purged sample. The difference between filtered and unfiltered samples was assumed to equal biomass C.

Reduction Rates of As(V) in Presence of Fe Oxides
Microbial As(V) reduction rates in the presence of Fe oxides were evaluated and compared with reduction rates of aqueous As(V). A CN-0 population, described above, was used as an inoculum source in these experiments. Goethite and 2-line ferrihydrite were prepared by the methods of Schwertmann and Cornell (1991). The ferrihydrite suspension was washed with deionized H2O until NO-3 was not detectable (with ion chromatography); goethite was dialyzed until the electrical conductivity (EC) in the goethite suspension equaled the EC of deionized H2O. The suspensions were lyophilized, and ground with mortar and pestle to pass a 125-µm sieve. The identities of the solid phases were confirmed with x-ray diffraction, and surface areas were determined with three-point N2(g)-BET analysis yielding values of 25 m2 g-1 and 363 m2 g-1 for goethite and ferrihydrite, respectively. A batch sorption isotherm experiment with As(V) and As(III) was conducted with 2-line ferrihydrite to select an As(V) loading level for the reduction experiments. Suspensions of ferrihydrite (25 mM as Fe) were treated with a range of initial As(V) and As(III) concentrations (0, 0.03, 0.1, 0.3, 1, 3, 10, and 30 mM) made from Na2HAsO4·7H2O and NaAsO2 salts, respectively, in the following pH 6 nutrient solution: NH4NO3 (2.5 mM), MgCl2 (2 mM), CaSO4 (4 mM), KH2PO4 (5 mM), KOH (0.5 mM), NaOH (2.5 mM), HNO3 (3.5 mM), H2SO4 (1 mM), and micronutrients (as above). Suspensions were purged with N2(g), shaken for 96 h on a horizontal shaker (72 cycles min-1), and filtrates (0.2 µm) analyzed for As (described below). Equilibration time was determined by conducting a kinetic sorption study at 1 mM As(V).

As(V) reduction experiments were conducted for both goethite (at 100 mM As/2.5 mM Fe and 13.4 mM As/25 mM Fe) and ferrihydrite (300 mM As/25 mM Fe). Experiments were conducted in a controlled EH/pH chamber (2.4-L capacity) that was fitted with air-tight ports for combination pH and platinum (Pt) redox electrodes, a fritted glass nebulizer, and a gas outlet (modified from Patrick et al., 1973). During periods of active microbial growth, the controller (Model 05656-05, Cole Parmer, Vernon Hills, IL) could maintain the solution EH to within approximately ±25 mV by sparging the solution with 99% O2(g)/1% CO2(g) and/or 99% N2(g)/1% CO2(g) (certified by Air Liquide, La Porte, TX). CO2(g) was included to imitate a typical soil environment. The redox electrodes were initially tested in ferrous-ferric and quinhydrone reference solutions (ASTM, 1993), and recalibrated periodically during the experiments. The difference between the measured potential and the hydrogen electrode potential in quinhydrone reference solution (pH 6.86) was added to each measured potential to obtain EH values. Initially, the Fe oxide, As(V), and nutrient solution mixture (described above) was continually stirred and sparged with 99% O2(g)/1% CO2(g) for 2 to 4 d to obtain equilibrium with respect to solution and sorbed phase As concentrations. Each mixture was subsequently inoculated with a CN-0 cell suspension and sparged for an additional 24 h with the O2(g)/CO2(g) mixture prior to reducing EH by sparging with 99% N2 (g)/1% CO2 (g). Glucose was added as a non-specific electron donor at rates ranging from 2.3 to 5.2 mmol C d-1, adjusted periodically to attain the required experimental redox potential. The solution pH was held constant by delivering small aliquots of HCl (0.1–3 M) or KOH (0.1–1 M) either manually or with a Dosimat 655 pH controller (Brinkmann Instruments, Inc., Westbury, NY). Samples were filtered (0.2 µm) for determination of As(T), As(V), As(III), S(-II), Cl-, SO2-4, NO-3, PO3-4, alkalinity, EC, Fe, and cations (described below).

Analytical Techniques
Aqueous As was analyzed with continuous flow hydride generation atomic absorption spectrophotometry (HGAAS) by acidifying samples to 3 M HCl, pre-reducing any As(V) in 1% (w/v) KI, and mixing with 0.6% (w/v) NaBH4/0.5% (w/v) NaOH in a reaction coil. Flow rates were 7 and 1 mL min-1 for sample and NaBH4 reagent, respectively, and generated arsine was quantified in an air-acetylene flame at 193.7 nm. The average measured concentration of a 0.267 µM EPA certified standard (Spex Certiprep, Inc., Metuchen, NJ) was 0.268 ± 0.025 µM (n = 69). As(V) and As(III) were determined for the solution studies using the following method modified from Masscheleyn et al. (1991b). Five milliliters of filtered sample were added to 1 mL of 2 M TRIS buffer (pH 6.0). While sparging the buffered solution with N2(g), 1 mL of 3% NaBH4/0.1% NaOH was added incrementally (0.2 mL over 15 s, 3 min wait, 0.8 mL over 45 s, 3 min wait) to liberate arsine from any As(III) in the sample. Samples for As determination were preserved in 1% concentrated HCl until analyzed. The original sample, As(T), and the speciated sample, As(V), were both analyzed for As with HGAAS, and the As(III) concentration was calculated by difference. Analysis of standards made from NaAsO2 and Na2 HAsO4·7H2O salts yielded recoveries of 93 ± 10% (n = 7) and 110 ± 5% (n = 8) for As(III) and As(V), respectively. Quantification of As(III) and As(V) in the solid phase experiments was similar to the above procedure except 0.5 mL of both TRIS and NaBH4 were used, yielding As(III) and As(V) recoveries of 94 ± 14% (n = 17) and 110 ± 16% (n = 18), respectively.

Dissolved organic carbon (DOC) was analyzed with a DC-80 carbon analyzer (Tekmar-Dohrmann, Cincinnati, OH); cations and S by inductively coupled plasma spectrometry (ICP); and Cl-, NO-3, SO2-4, and PO3-4 by ion chromatography (IC). Total alkalinity was measured by HCl titration to an inflection point near pH 4.5. Titrated samples were then sparged with N2(g) for more than 15 min, titrated to the original pH with NaOH, and retitrated to the inflection pH with HCl in order to calculate carbonate alkalinity. Fe was analyzed with the phenanthroline method and S(-II) was determined by either iodometric titration (only for the high As:Fe goethite experiment) or the methylene blue method (APHA, 1989). Ion activities and saturation indices for Fe and As sulfide phases were calculated by the aqueous chemical equilibrium model, MINTEQ (Alison et al., 1991). Samples generated in experiments to determine the fate of 14C -glucose with CN-8 were analyzed for 14C using liquid scintillation (Model 2200CA, Packard Instr., Meriden, CT). Fermentation products were identified via GC-mass spectroscopy with a Model 5890 Series II+ gas chromatograph (Hewlett Packard Co., Santa Clara, CA), a Stable Wax column (Resteck Corp., Bellefonte, PA), and a Model VG70E double focusing mass spectrometer (VG Mass-lab, Thermo Instrument Systems, Hurst, TX). Aliphatic organic acids were quantified by ion chromatography (Dionex, Sunnyvale, CA; AS6-ICE column; 0.5 mL min-1 flow rate; mobile phase 80% 0.4 mM aqueous heptafluorobutyric acid and 20% acetonitrile; AMMS-2 suppressor) in combination with 14C-radioisotope detection (Radiomatic 500TR Series, Packard Instr.,) and suppressed electrical conductivity detection. Gas partial pressures in serum bottle headspaces were determined with a Carle (Tulsa, OK) Series 100 gas chromatograph and total headspace pressures were measured with a pressure transducer (Tensimeter, Soil Measurement Systems, Las Cruces, NM). Headspace gas volumes were calculated from the gas partial pressure, the total headspace pressure, and the total headspace volume. Moles of headspace gas were calculated with the ideal gas law.


   Results and discussion
TOP
 ABSTRACT
 INTRODUCTION
 Materials and methods
 Results and discussion
Conclusions
 REFERENCES
 
Effects of As Enrichment and Microbial Growth on As(V) Reduction
The effect of prior As enrichment on As(V) reduction was evaluated by inoculating solutions with either the unenriched population (CN-0) or the enriched soil isolate (CN-8). No reduction was observed in sterile controls, confirming that As(V) reduction was biologically mediated (Fig. 1) . Reduction of As(V) by CN-8 was considerably faster and more complete than by CN-0 (Fig. 1A, 1B). In addition, As(V) reduction occurred simultaneously with growth of CN-8 (as estimated by optical density), in sharp contrast to results with CN-0, where As(V) reduction occurred primarily after growth had apparently ceased ({approx}1 d).



View larger version (29K):
[in this window]
[in a new window]
  		Fig. 1 Extent of As(V) reduced (A, B) and growth curves (C, D) vs. time for a range of initial As(V) concentrations in serum bottles inoculated with either CN-0 or CN-8. Glucose concentration was 5 mM for each treatment. Experiment with CN-0 and an initial As(V) concentration of 600 µM was conducted in duplicate and error bars denote one standard deviation about the mean. Error bars smaller than symbols are not shown. As(V) reduced (%) = 100[1 - (AsV/AsT)]

 
Although initial growth of both CN-0 and CN-8 was somewhat delayed in 600 µM As treatments compared with lower As concentrations, growth rates and biomass yields were similar at all tested As(V) concentrations (Fig. 1C, 1D), including a 5 mM As(V) treatment (results not shown). The results of these experiments imply that As(V) was not used for anaerobic respiration in these incubations. If the coupling of As(V) reduction with C oxidation was the sole energy source for these organisms, we would instead expect a direct relationship between initial As(V) concentration and both growth rate and biomass yield.

Additional independent measurements demonstrated that CN-8 was fermenting glucose and was not oxidizing a significant amount of C via respiration. Analysis of C end products by ion chromatography and GC-MS found that the total dissolved organic C remaining after growth of CN-8 was approximately 21 mM C (=0.67 initial C) and was composed primarily of butyric acid (~6 mM as C), with lower concentrations of acetic acid, formic acid, butanol, and sec-butyl butyrate (1-methylpropyl butanoate). Furthermore, H2(g) was the only significant reduced gaseous product found in serum bottle headspaces (up to 0.11 atm [11 kPa]), and the amount of evolved H2(g) was found to equal approximately the amount of evolved CO2(g) on the basis of 14C analyses. These observations are consistent with the stoichiometry of a fermentation pathway yielding butyric acid and H2 as reduced species (Brock et al., 1994).

DGGE analysis was used to assess the complexity of the As(V)-reducing population in the enriched culture. The organism referred to above as CN-8 was isolated from a dilution extinction of the seventh serial transfer. This organism exhibited a single band in DGGE analysis (results not shown) and cultured as a single colony type on AR agar. The DGGE profile and colony morphology of CN-8 did not change after two subcultures. These observations are consistent with the conclusion that CN-8 was a pure culture. The 1070- to 1392-bp region of the 16S gene of this isolate was amplified by PCR and sequenced. Comparison of the CN-8 sequence against those contained in public databases suggested its closest relative (98% identity) to be Clostridium intestinalis (Benson et al., 1997; Maidak et al., 1997), a strict anaerobe known to ferment glucose. The genotypic data classifying CN-8 as Clostridium is in agreement with the phenotypic characterization that showed that CN-8 is a Gram positive, spore-forming obligate anaerobe, that ferments glucose to butyrate and H2. These are defining features for Clostridium (Willis, 1990).

Comparison of As(V) reduced by CN-0 and CN-8 as a function of optical density suggests that CN-8 was not a numerically dominant member of the total microbial community (Fig. 2) . For example, CN-0 growth had a substantially smaller effect on the extent of As(V) reduction than growth of CN-8. It is also noteworthy that the majority of As(V) reduced in the presence of CN-0 occurred after optical densities (A500) stabilized near 0.7. Conversely, the percentage of As(V) reduced by CN-8 was positively correlated (r2 = 0.95) to optical density (A500) when all data from 6 to 600 µM As were grouped (n = 36), and little As(V) reduction occurred after growth ceased. However, CN-8 growth rates were independent of initial As(V) concentration from 6 to 600 µM As, and as judged by final optical density measurements at As(V) concentrations up to 5 mM, As(V) concentration had little apparent effect on growth yield. These results show that As(V) was not required for CN-8 growth, nor did the presence of As(V) enhance the growth rate of CN-8. The fact that total CN-8 growth yield was independent of both the amount of initial As(V) and the total As(V) reduced suggests that As(V) reduction did not occur via respiration, and perhaps occurred via an alternate mechanism such as detoxification (Tamaki and Frankenberger, 1992; Cervantes et al., 1994).



View larger version (30K):
[in this window]
[in a new window]
  		Fig. 2 Extent of As(V) reduction as a function of optical density in serum bottles inoculated with either CN-0 or CN-8. Data for symbols connected by dashed lines were collected after the growth phase (Fig. 1). Experiments with initial As(V) concentrations of 600 µM were conducted in duplicate and error bars denote one standard deviation about the mean. Error bars smaller than symbols are not shown. As(V) reduced (%) = 100[1 - (AsV/AsT)]

 
Several organisms capable of fermentation (E. coli, S. aureus, S. xylosus) have been found to invoke As(V) resistance by reducing As(V) to As(III) with an As(V) reductase, and subsequently excreting As(III) from the cell with an efflux pump (Silver et al., 1993; Cervantes et al., 1994). Although the specific mechanism of As(V) reduction by CN-8 cannot be definitively elucidated with the data collected, our results have important implications for understanding As cycling in soils and natural water systems. The reduction of As(V) via detoxification occurs without utilization of As(V) as a terminal electron acceptor for respiration. If detoxification is an important mechanism of microbially mediated As(V) reduction in soils and natural waters, then the prediction of As valence, and thus As behavior, based solely on redox status may be problematic.

Significantly lower H2(g) production by CN-0 (0.015 atm [1.5 kPa]) than by the CN-8 isolate (0.11 atm [11 kPa]) suggests substantial qualitative differences in metabolic functions between cultures. This is not a surprising outcome given that CN-0 suspensions were mixed populations extracted from the C3-NI soil. Also, As(V) reduction kinetics were quite different between CN-0 and CN-8 (Fig. 1), with the majority of As(V) reduction in CN-0 occurring after net growth had ceased. The latter observation suggests that either a subset of the CN-0 population coupled As(V) reduction with organic acid oxidation following initial fermentation of glucose, or that substantial detoxification occurred during the stationary phase of growth in the CN-0 culture. We are currently continuing attempts to address As(V) reduction mechanisms in mixed cultures, and to determine the potential importance of CN-8, and similar organisms, in soils.

Effect of As(V) Concentration on As(V) Reduction Rate
The effects of initial As(V) concentration on rates of As(V) reduction (M d-1) were evaluated in more detail for both CN-0 and CN-8 (Fig. 3) . Rates of As(V) reduction with CN-0 were lower than CN-8 by approximately 2 fold at 6 µM As and 10 fold at 600 µM As, despite similar growth rates for CN-0 and CN-8 (Fig. 1). As discussed above, differences in reduction rates were consistent with differences in microbial populations between CN-0 and CN-8. Reduction rates for CN-8 reached a maximum of 1 mM d-1 at initial As(V) concentrations above 600 µM As(V); these rates are approximately 5 to 10 fold lower than rates of dissimilatory As(V) reduction previously reported for strains MIT-13 (Ahmann et al., 1994) and SES-3 (Laverman et al., 1995). The difference in As(V) reduction mechanisms between CN-8 (detoxification) and isolates MIT-13 and SES-3 (respiration) may explain the differences in observed rates of As(V) reduction.



View larger version (22K):
[in this window]
[in a new window]
  		Fig. 3 Effect of initial As(V) concentration on As(V) reduction rate in serum bottles inoculated with CN-0 or CN-8 at initial glucose concentrations of 5 mM. Rates of As(V) reduction for two environmental isolates from other studies are shown for comparison. Experiments with initial As(V) concentrations of 600 µM were replicated (triplicates on CN-8, duplicates on CN-0) and error bars denote one standard deviation about the mean. Error bar for CN-0 is smaller than symbol

 
The dependence of As(V) reduction rate on As(V) concentration (6–600 µM) and microbial biomass during the growth phase of CN-8 can be formalized with the following rate equation:

(1)
where X represents microbial biomass in units of millimolar C. Measured values of X (from correlation of optical density and biomass C) and As(V) as a function of time were used to obtain fitted values of k1 for experiments performed at 6 to 600 µM As(V). In all cases, this rate equation provided an excellent fit to the experimental data (r2 >= 0.97) and resulted in consistent values of k1 (0.029–0.039 h-1 mM C-1) across As concentrations spanning two orders of magnitude (Table 1) . Increases in biomass C were described with a growth model that was first order with respect to C, and fitted curves for biomass C and As(V) reduction agreed well with measured data (Fig. 4) . In summary, the As(V) reduction rate during the growth phase of CN-8 was first order with respect to both As(V) concentration and microbial biomass.


View this table:
[in this window]
[in a new window]
  		Table 1 Summary of optimized model parameters for growth{dagger} and As(V) reduction rate{ddagger} during growth phase of CN-8

 


View larger version (20K):
[in this window]
[in a new window]
  		Fig. 4 Microbial biomass and As(V) concentrations versus time for a representative As treatment (60 µM) inoculated with CN-8. The fitted microbial growth curve was generated assuming first order kinetics (dX/dt = kX) until a maximum of X = 2.7 mM C was reached. The solved expression for X was substituted into the As(V) reduction model (see Eq. [1]) and the As(V) reduction curve generated by optimizing k1

 
Reduction of As(V) in the Presence of Fe Oxides
Incubation studies using CN-0 were also conducted in the presence of Fe oxides in controlled EH/pH chambers to (i) determine if the presence of Fe(III), a ubiquitous electron acceptor, inhibited reduction of aqueous As(V), and (ii) evaluate rates of As solubilization from Fe oxides as a result of microbially induced reduction of As(V) and Fe oxides. Unlike CN-8, the CN-0 population likely contained organisms capable of respiration; and, therefore, reductive dissolution of Fe oxides may have been possible. In a goethite suspension with a high As:Fe molar ratio (0.04), approximately 25% of the aqueous As(V) was reduced to As(III) under oxygenated conditions (Day 2–4), suggesting a detoxification process (Fig. 5) . As Pt EH decreased from 450 to 200 mV (Day 4–5), As(V) was reduced at a peak rate of 110 µM d-1. Under these lower redox conditions, the higher As(V) reduction rate could have been due to both detoxification as well as As(V) respiration. Although the studies in the presence and absence of goethite were conducted under somewhat different conditions, the observed As(V) reduction rates were within the range attained by CN-0 in the absence of a solid phase (Fig. 3). The results of these experiments implied reduction of aqueous As(V) was not inhibited by the presence of goethite.



View larger version (24K):
[in this window]
[in a new window]
  		Fig. 5 Concentrations of aqueous As(T), As(V), and As(Ill) vs. time as Pt electrode EH decreased in a stirred goethite suspension with a high As:Fe molar ratio (0.04). Suspension was inoculated with a CN-0 population on day 3 and N2(g) replaced O2(g) on Day 4 to promote reduction. Sulfide levels ranged from 5 x 10-6 to 8 x 10-5 M

 
Total aqueous As concentrations gradually declined as As(V) was reduced (Fig. 5), likely because As(III) sorbs more strongly than As(V) to goethite at pH 7.5 and high As:Fe ratios (Sun and Doner, 1998). When the Pt EH decreased to 0 mV (Day 5–7), total aqueous As decreased sharply upon detection of S(-II), implying an As(III)-sulfide solid phase may have precipitated. Ion activity products (IAPs) were lower than Ksp values for amorphous As2S3 (Eary, 1992) and orpiment (Webster, 1990) by factors of at least 16.6 and 3.2, respectively. Although the results indicate the samples were undersaturated with respect to both solid phases, it can not be conclusively stated that As(III)-sulfide solid phases did not precipitate given discrepancies in literature equilibrium constants (Sadiq and Lindsay, 1979).

To evaluate As(V) reduction in a system dominated by sorbed, rather than aqueous As, an experiment was conducted in a goethite suspension with a low As:Fe ratio (5.3 x 10-4 mol mol-1). Initial aqueous As concentrations represented only 0.5% of total As, yet increased seven fold over 25 d as Pt EH decreased from 500 to -100 mV (Fig. 6) . The increase in aqueous As concentration (measured predominantly as As(III)) was likely due to weaker sorption of As(III) than As(V) on Fe oxides at low As:Fe molar ratios (Pierce and Moore, 1982). Rates of As solubilization averaged 0.01 mM d-1, considerably slower than the minimum observed reduction rate of aqueous As(V) (12 mM d-1). The significantly lower rates of As solubilization as compared with reduction rates of As(V) in solution suggest that reactions responsible for desorption of As(V) controlled rates of formation of aqueous As(III). Reductive dissolution of goethite could have accounted for a portion of the increase in aqueous As concentrations. However, aqueous Fe concentrations did not increase as redox potential decreased and were near the detection limit of 1 µM (~0.004% of total Fe) for most of the experiment. This finding is in agreement with evidence that low surface area and the formation of surface coatings significantly inhibit goethite reduction rates (Roden and Zachara, 1996). Therefore, the 7-fold increase in total aqueous As concentrations was likely due to differences in As sorption between As(V) and As(III), rather than reductive dissolution of goethite.



View larger version (25K):
[in this window]
[in a new window]
  		Fig. 6 Concentrations of aqueous As(T), As(V), and As(III) versus time as Pt electrode EH decreased in a stirred goethite suspension with a low As:Fe molar ratio (5.3 x 10-4). Suspension was inoculated with a CN-0 population on Day 2 and N2(g) replaced O2(g) on Day 3 to promote reduction

 
Initial rates of As solubilization during reduction were much greater in a 2-line ferrihydrite suspension than in a goethite suspension, due to differences in reductive dissolution rates between goethite and ferrihydrite. Total aqueous As concentrations increased by a factor of 20 in less than 2 d as Pt EH decreased from 475 to 300 mV (Fig. 7) . Further decreasing Pt EH to 200 mV resulted in a subsequent, though smaller, increase in As concentration. Both large spikes in aqueous As concentrations were transient, however, with final aqueous As concentrations representing a 2-fold increase over initial concentrations. Reductive dissolution of ferrihydrite occurred concurrently with increases in total aqueous As, as determined from measurable increases in aqueous Fe (Fig. 7B). Although aqueous Fe concentrations could be inflated by the presence of colloidal Fe solid phase in the filtrate, initial aqueous Fe concentrations represented less than 0.004% of the total Fe, suggesting that the synthesized ferrihydrite did not contain significant colloidal Fe. In addition, Fe(II) represented approximately 100% of total aqueous phase Fe in the majority of samples, supporting the conclusion that the observed increase in aqueous Fe was due to reductive dissolution rather than the presence of a Fe(III) solid phase in the filtrate.



View larger version (33K):
[in this window]
[in a new window]
  		Fig. 7 Effect of decreased Pt electrode EH (A) on As and Fe concentrations (B) in a stirred ferrihydrite suspension at an As:Fe molar ratio of 0.012. Suspension was inoculated with a CN-0 population on Day 4 and N2(g) replaced O2(g) on Day 5 to promote reduction

 
The average rate of As solubilization (0.9 mM d-1) during reduction in ferrihydrite suspensions was approximately 100 fold higher than the average rate in the goethite experiment (0.01 mM d-1) at similar aqueous As concentrations (Fig. 6). The difference in As solubilization rates was correlated with differences in reductive dissolution rates between goethite and ferrihydrite, consistent with published reports showing crystallinity and surface area are important factors in reductive dissolution rates of Fe oxides (Roden and Zachara, 1996). When the Pt electrode EH was stabilized at both 300 and 200 mV, aqueous As concentrations declined, apparently because of resorption of As. Attempts to quantify sorbed As(V) and As(III) by extracting ferrihydrite in 10 mM to 3 M phosphate buffers was unsuccessful because of inadequate As recoveries (<32 %). Currently, we are evaluating other techniques for the quantification of sorbed phase As(V) and As(III), in hopes of more accurately quantifying reduction rates of sorbed phase As(V).

The reduction of As(V) to As(III) in each of our studies could be caused by either microbial reduction or via abiotic reduction. For example, S(II), Fe(II), H2(g), and reduced organic acids could each reduce As(V) to As(III) based on thermodynamic equilibrium. However, abiotic As(V) reduction has been shown to be substantially slower than microbially mediated reduction (Ahmann et al., 1997; Newman et al., 1997). Therefore, the rates of As(V) reduction in our studies are believed to be controlled directly by microbial processes rather than by abiotic reduction.


   Conclusions
TOP
 ABSTRACT
 INTRODUCTION
 Materials and methods
 Results and discussion
 Conclusions
REFERENCES
 
The effects of prior enrichment in the presence of As, microbial number, initial As(V) concentration, and presence of Fe solid phases on As(V) reduction rates were evaluated with microorganisms obtained from an As-enriched agricultural soil. Reduction rates by an As(V)-enriched isolate (CN-8) were up to ten fold higher than by an unenriched population (CN-0). For CN-8, As(V) reduction rates were first-order in both As(V) concentration and microbial biomass when As(V) concentrations were <=600 µM. At initial As(V) concentrations greater than 600 µM, reduction rates of aqueous As(V) reached a maximum of approximately 1 mM d-1. We also showed that CN-8, whose closest RDP relative was Clostridium intestinalis, fermented glucose and produced butyric acid and H2(g) as the dominant reduced species. Microbial growth rates were independent of initial As(V) concentration or the mass of As(V) reduced, suggesting that CN-8 was not reducing As(V) via respiration, and may have employed a detoxification pathway similar to that described in studies using E. coli, S. aureus, and S. xylosus. We are continuing efforts to investigate the potential role of detoxification pathways in As cycling in soils and natural water systems.

In incubations containing CN-0, the effects of microbial growth on the solubilization of total As in the presence of Fe oxide phases depended in part on As surface coverage, and on the surface area or crystallinity of the Fe oxide phase. Microbial reduction rates of aqueous As(V) were not affected by goethite, but solubilization of As was highly dependent on the As:Fe ratio due to the relative dependence of As(V) and As(III) sorption strengths on As surface coverage. The net rate of As desorption from ferrihydrite during reduction was found to be approximately 100-fold higher than from goethite at similar aqueous As concentrations. This was due, at least in part, to differences in reductive dissolution rates between solid phases. Based on our findings, rates of As mobilization in reducing environments are controlled by rates of As desorption, or more directly, rates of reductive dissolution of the oxide phase.American Society for Testing and Materials 1993


   ACKNOWLEDGMENTS
 
Although the research described in this article has been funded in part by the United States Environmental Protection Agency through a grant (R825403-01-0) to W. Inskeep, it has not been subjected to the Agency's required peer and policy review and therefore does not necessarily reflect the views of the Agency, and no official endorsement should be inferred. This work was also supported with funds from the Montana Agricultural Experiment Station (104398). We thank Dr. P. Grossl for providing the goethite used in our studies, Dr. J. Robinson for assistance with DGGE analysis, and Mr. R. Macur for technical assistance.

Received for publication April 30, 1999.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 Materials and methods
 Results and discussion
 Conclusions
 REFERENCES
 




This article has been cited by other articles:


Home page
 Appl. Environ. Microbiol.Home page
S. D'Imperio, C. R. Lehr, M. Breary, and T. R. McDermott
Autecology of an Arsenite Chemolithotroph: Sulfide Constraints on Function and Distribution in a Geothermal Spring
Appl. Envir. Microbiol., November 1, 2007; 73(21): 7067 - 7074.
[Abstract] [Full Text] [PDF]

Home page
 J. Environ. Qual.Home page
D. Vetterlein, K. Szegedi, J. Ackermann, J. Mattusch, H.-U. Neue, H. Tanneberg, and R. Jahn
Competitive Mobilization of Phosphate and Arsenate Associated with Goethite by Root Activity
J. Environ. Qual., October 24, 2007; 36(6): 1811 - 1820.
[Abstract] [Full Text] [PDF]

Home page
 J. Environ. Qual.Home page
J. E. McLean, R. R. Dupont, and D. L. Sorensen
Iron and arsenic release from aquifer solids in response to biostimulation.
J. Environ. Qual., July 1, 2006; 35(4): 1193 - 1203.
[Abstract] [Full Text] [PDF]

Home page
 Appl. Environ. Microbiol.Home page
T. R. Kulp, S. E. Hoeft, and R. S. Oremland
Redox Transformations of Arsenic Oxyanions in Periphyton Communities
Appl. Envir. Microbiol., November 1, 2004; 70(11): 6428 - 6434.
[Abstract] [Full Text] [PDF]

Home page
 ScienceHome page
R. S. Oremland and J. F. Stolz
The Ecology of Arsenic
Science, May 9, 2003; 300(5621): 939 - 944.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (44)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Jones, C.A.
Right arrow Articles by Inskeep, W.P.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Jones, C.A.
Right arrow Articles by Inskeep, W.P.
GeoRef
Right arrow GeoRef Citation
Agricola
Right arrow Articles by Jones, C.A.
Right arrow Articles by Inskeep, W.P.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
The SCI Journals Agronomy Journal Crop Science
Journal of Natural Resources
and Life Sciences Education		 Vadose Zone Journal
Journal of Plant Registrations Journal of
Environmental Quality		 The Plant Genome