Nitrilotriacetate Stimulation of Anaerobic Fe(III) Respiration by Mobilization of Humic Materials in Soil

Abstract

An enrichment culture capable of naphthalene mineralization reduced Fe(III) oxides without direct contact in anaerobic soil microcosms when the Fe(III) was placed in dialysis membranes or entrapped within alginate beads. Both techniques demonstrated that a component in soil, possibly humic materials, facilitated Fe(III) reduction when direct contact between cells and Fe(III) was not possible. The addition of the synthetic Fe(III) chelator, nitrilotriacetic acid (NTA), to soil enhanced Fe(III) reduction across the dialysis membrane and alginate beads, with the medium changing from clear to a dark brown color. An NTA-soil extract was more effective in Fe(III) reduction than the extracted soil itself. Characteristics of the NTA extract were consistent with that of humic substances. The results indicate that NTA improved Fe(III) reduction not by Fe(III) solubilization but by extraction of humic substances from soil into the aqueous medium. This is the first study in which stimulation of Fe(III) reduction through the addition of chemical chelators is shown to be due to the extraction of electron-shuttling compounds from the soil and not to solubilization of the Fe(III) and indicates that mobilization of humic materials could be an important component of anaerobic biostimulation.

Fe(III) reduction is an important process in the intrinsic biodegradation of natural and xenobiotic organic compounds in anaerobic environments (13, 19). Although Fe(III) is one of the most abundant terminal electron acceptors in subsurface soils, it occurs in the form of poorly bioavailable, insoluble oxides at neutral pH. The mechanisms of microbial electron transport to insoluble electron acceptors are poorly understood. The microorganisms may accomplish this by either (i) attaching to the iron substrate and directly transferring electrons to it, (ii) using available electron-shuttling compounds or producing their own, or (iii) solubilizing the iron.

Iron reduction does not occur when some pure cultures are separated from Fe(III) oxides using dialysis membranes (3, 12, 33) or by entrapping the Fe(III) oxides in alginate beads (24). The assumption is that any electron-shuttling or Fe(III)-solubilizing compounds present would diffuse through the membrane or the alginate beads to allow Fe(III) reduction. Some cultures of dissimilatory iron-reducing bacteria have been shown to transfer electrons to insoluble iron via outer membrane proteins (5, 22, 23) when the cells are in direct contact with the oxides. Periplasmic proteins may also play a role by transferring electrons to Fe(III) reductases present in the outer membrane (11). In most subsurface environments, the mineral surface area available for direct contact with bacteria would preclude high rates of electron transfer. Soluble humic substances are more readily bioavailable than insoluble Fe(III) oxides (18) and can access Fe(III) in pore spaces too small for microorganisms to enter.

Humic substances, ubiquitous in aquatic and terrestrial environments, may play an important role in stimulating Fe(III) reduction by acting as electron shuttles (18). The quinone moieties in humic substances are thought to act as electron acceptors (15, 18, 30). The ability of humic-reducing microorganisms to reduce anthraquinone-2,6-disulfonate (AQDS), a humic acid analogue (8, 9, 20), demonstrates that extracellular quinones can serve as electron acceptors. In this model, microorganisms transfer electrons to the quinones, generating hydroquinones, which in turn abiotically transfer electrons to Fe(III) oxides. This electron shuttling regenerates the oxidized form of the humic colloids, so that it may continually undergo cycles of reduction and oxidation. Recent evidence suggests that the role of humic substances as electron shuttles may be a key mechanism for Fe(III) reduction in nature (25).

Another way that the bioavailability of insoluble Fe(III) may be enhanced is to increase its solubility with biological Fe(III) chelators, such as naturally produced siderophores (28) or chemical chelators such as nitrilotriacetic acid (NTA) (3, 16, 17). In one study, the accelerated rate of Fe(III) reduction and the degradation of aromatic hydrocarbons in sediments from a petroleum-contaminated aquifer were attributed to the Fe(III)-chelating ability of NTA (16).

In the present study, the need for direct contact between cells and insoluble Fe(III) oxide was evaluated in the presence of soil and NTA. A sediment-derived culture was grown in anaerobic soil microcosms in which Fe(III) oxide was retained within a dialysis membrane or entrapped within alginate beads. Soil was added to better simulate field conditions and to learn more about the role of soil components in Fe(III) reduction in aquatic sediments. Stimulation of Fe(III) reduction by an electron-shuttling agent, AQDS, and a Fe(III) chelator, NTA, was also examined.

MATERIALS AND METHODS

Inoculum and growth conditions.

The inoculum used for the experiments was obtained from an anaerobic enrichment culture grown with naphthalene as the electron donor and amorphous Fe(III) oxide as the electron acceptor. This enrichment culture was derived from coal tar-contaminated sediments collected from Lake Ontario, Canada. Although this consortium was previously shown to mineralize naphthalene and anthracene under similar Fe(III)-reducing conditions (27), it may also use the ethanol in which the naphthalene was dissolved and/or organic matter from the soil as sources of carbon and energy.

Cells were grown under strict anaerobic conditions in medium containing (g/liter): NaHCO₃, 8.75; NH₄Cl, 0.875; NaH₂PO₄ H₂O, 2.1; KCl, 0.35; as well as vitamins and trace elements (13). Amorphous Fe(III) oxyhydroxide was synthesized as described previously (13) by neutralizing a 0.4 M FeCl₃ solution to pH 7 with NaOH. Amorphous iron was added as the terminal electron acceptor, and 20 ppm of naphthalene dissolved in ethanol was added as the carbon source. Uncontaminated garden soil (10% dry weight/vol) and a 10% (vol/vol) inoculum from a naphthalene-enrichment culture were added to each microcosm. Based on the Unified Soil Classification System, the garden soil was 14% silt, 54% fine sand, 29.5% medium sand, and 2.5% coarse sand. Using methods A and C of ASTM D2974-87 (1), the soil was found to have 2.35% ± 0.03% moisture and 2.03% ± 0.49% organic matter, respectively, and had been free of pesticides and fertilizers for several years. Soil pH, measured as described by Forster (10), was 6.1.

Medium was dispensed into anaerobic culture tubes (25 ml) or serum bottles (100 ml), capped with Teflon-lined, butyl rubber stoppers, and sealed with aluminum crimps. To remove oxygen, each bottle was purged with high-purity nitrogen passed through heated copper filings (7). All further manipulations, including inoculation, were carried out in an anaerobic glove box (Nexus One; Vacuum Atmospheres Company, Hawthorn, Calif.). To study the effect of electron-shuttling agents, AQDS (final concentration, 50 μM) was added to the medium. To study the effect of synthetic chelators, 75 mmol of NTA/liter was added. To prepare abiotic controls, experiments were set up as described above and, after inoculation, cells were killed by the addition of 3.89% (wt/vol) HgCl₂ (4).

Fe(III) in dialysis membrane or immobilized in alginate beads.

Cells were separated from Fe(III) oxides using either dialysis membranes or alginate beads. For membrane separation, Fe(III) oxide was placed in Spectra/Por dialysis tubing (Spectrum Laboratories, Rancho Dominguez, Calif.) with a molecular weight cutoff of 8,000. The dialysis tubing was rinsed and added to anaerobic culture tubes (25 ml). To entrap the Fe(III) in alginate beads, a solution of 500 mmol of Fe(III) oxide/liter in 2% (wt/vol) alginate was prepared. The alginate beads (diameter, ca. 5 mm) were formed as previously described (26) and added to serum bottles (100 ml) to a final concentration of 75 mmol of Fe(III)/liter. The dialysis membrane experiments were carried out in duplicate, and the alginate bead experiments were carried out in triplicate.

NTA-Fe(III) solubilization.

The ability of NTA and the soil to solubilize Fe(III) without added inoculum was examined. Samples containing mineral salts medium and amorphous Fe(III) (75 mmol/liter) in the presence and absence of soil (10% [wt/vol]) and NTA (75 mmol/liter) were prepared. After 20 h, the amount of dissolved Fe(III) was determined by centrifuging the mixture at 5,580 × g for 10 min in a Beckmann Allegra 21R centrifuge and measuring the Fe(III) in the supernatant. These experiments were carried out in triplicate.

NTA-soil extraction.

Uncontaminated garden soil (2.5 g) was added to 25 ml of mineral salts medium containing 75 mmol of NTA/liter. The mixture was vortexed for ca. 10 min until the liquid became a dark brown color. The mixture was then centrifuged at 5,580 × g for 10 min, and the dark brown supernatant (referred to as NTA-soil extract) was recovered.

The above-described extraction and centrifugation procedure was repeated with the same soil sample several times until no further brown color was extracted into the liquid medium. The brown supernatant was then decanted, and the extracted soil was used to set up microcosms. All of these experiments were carried out in triplicate, and Fe(II) concentrations were measured over time.

Controls were designed to ensure that the NTA and the NTA-soil extract could not be used as substrates for growth by the cells. Microcosms were prepared in triplicate using the NTA or the NTA-soil extract and no added carbon source.

Infrared (IR) analysis.

Humic substances were extracted from the soil using an alkali extraction method adapted from the work of Stevenson (31) and Schnitzer (29). A 25-g sample of sieved, uncontaminated garden soil was washed with 250 ml of 0.1 M HCl for 1 h with shaking. After the soil was allowed to settle, the aqueous-phase material was decanted and discarded. Approximately 25 ml of water was added to the soil, and the slurry was allowed to sit for 30 min. The pH was then adjusted to 7.0 with the addition of 1 M NaOH. Under an atmosphere of nitrogen gas, the total volume of the slurry was brought to 250 ml with the addition of 0.1 M NaOH, and the mixture was stirred for 24 h. The mixture was then centrifuged, and the particle-free supernatant was adjusted to pH 1.0 by adding 6 M HCl with constant stirring. This resulted in the formation of a dark brown precipitate. The suspension was allowed to stand for 12 h and centrifuged to recover the precipitate. The precipitate was suspended in a solution of 0.1 M HCl and 0.3 M HF in a plastic container overnight to remove mineral impurities and then washed with distilled water to remove chlorides. The sample was considered to be free of chlorides when the addition of 0.1 M AgNO₃ to the supernatant did not produce a precipitate. The sample was then freeze-dried (Savant SuperModulyo).

The NTA-soil extract for IR analysis was prepared in a manner similar to that described for the NTA-soil extract used in the microcosm experiments except that the 75 mM NTA (trisodium salt) solution did not contain mineral salts and the mixture was stirred for 24 h (not vortexed for 10 min) under an atmosphere of nitrogen gas. Once the supernatant was recovered by centrifugation, a precipitate was obtained and processed as in the alkali extraction method described above.

Pellets of the alkali-extracted humic materials, NTA-soil extract, and a commercial sodium salt humic acid (Sigma Chemical) were prepared by mixing approximately 1 mg of the sample with 100 mg of KBr. Infrared spectra were recorded online using an IR spectrophotometer (MB-120 FT-IR; Bomem, Quebec, Canada) with an attached computer running GRAMS/AI software.

Fe(II) and Fe(III) measurements.

Fe(II) accumulation was measured as described previously (13) by digesting 0.1 ml of the sample in 5 ml of 0.5 M HCl. After 1 h at room temperature, 0.1 ml of the extract was added to 5 ml of ferrozine (1 g/liter; monosodium salt of 3-(2-pyridyl)-5,6-bis(4-phenylsulfonic acid)-1,2,4-triazine) (Spectrum Laboratories, Gardena, Calif.) in 50 mM HEPES. The Fe(II) concentration was determined by measuring the absorbance at 562 nm. Standards of Fe(II) were prepared using ferrous ethylene-diammonium sulfate.

The Fe(III) concentration was measured by digesting 0.1 ml of the sample in 0.25 M hydroxylamine hydrochloride in 0.25 M HCl (14). Under these acidic conditions, the hydroxylamine converts all of the Fe(III) to Fe(II). After 1 h at room temperature, 0.1 ml of the extract was added to 5 ml of ferrozine (1 g/liter) in 50 mM HEPES, and the absorbance was measured at 562 nm. The amount of Fe(III) was calculated as the difference between the Fe(II) measured in the hydroxylamine and HCl extractions.

RESULTS

Fe(III) reduction in abiotic controls.

Fe(III) reduction was measured as the accumulation of HCl-extractable Fe(II) over time in the liquid phase outside the dialysis membrane or alginate beads or from a homogenous suspension when the cells were in direct contact with the Fe(III) oxide. Little to no Fe(II) accumulated in abiotic controls when Fe(III) was separated from killed cells (data not shown) or in direct contact with them (Fig. 1C and D), indicating that Fe(III) reduction was mostly mediated by a biological process.

FIG. 1.

Change in free Fe(II) concentration when a naphthalene enrichment culture was grown with and without soil and amorphous Fe(III) oxide as the terminal electron acceptor. Results with error bars are means for triplicate microcosms, and results without error bars are averages for duplicate microcosms. (□), no barrier; (▴), Fe(III) in alginate; (•), Fe(III) in dialysis membrane; ×, no barrier, abiotic.

Fe(III) reduction without soil and without NTA.

When the culture was grown in direct contact with amorphous Fe(III) oxide without soil and without NTA, Fe(III) reduction occurred at a rate of about 0.6 mmol/liter/day, achieving a plateau at 34.7 ± 2.1 mmol/liter on day 53 (Fig. 1A). When the Fe(III) oxide was retained within a dialysis membrane, there was no Fe(III) reduction as measured by the lack of Fe(II) accumulation in the medium outside the dialysis membrane. This finding was further substantiated by the observation that the Fe(III) oxide in the membrane kept its original reddish-brown color after more than 60 days of incubation. When the Fe(III) oxide was entrapped in alginate beads, a small amount of Fe(III) reduction occurred, with the Fe(II) concentration increasing from 2.3 ± 0.4 mmol/liter at day zero to 4.8 ± 1.4 mmol/liter by day 61.

Fe(III) reduction with NTA, no soil.

When 75 mmol of NTA/liter (but no soil) was present in the microcosms, there was no Fe(III) reduction when the microorganisms were separated by either a dialysis membrane or by entrapment in alginate beads (Fig. 1B). A separate experiment showed that when the microorganisms were in direct contact with the Fe(III) oxide without soil, NTA addition up to a 7.5 mM concentration did not affect the rate or extent of Fe(III) reduction (data not shown). However, at higher NTA concentrations, Fe(III) reduction decreased, with 40% as much Fe(III) reduction occurring at 75 mM NTA by day 70 as with a control lacking NTA. In contrast, when soil was present, the rate and extent of Fe(III) reduction increased as the NTA concentration increased, even at low NTA concentrations (0.75 to 75 mM), and were always higher than without soil. For example, by day 70 with soil at 75 mM NTA, Fe(III) reduction was 50% higher than without NTA and was 125% higher than without both soil and NTA.

Fe(III) reduction with soil, no NTA.

When the culture was grown in direct contact with amorphous Fe(III) oxide in the presence of soil but no NTA, Fe(III) reduction was faster, and a slightly larger amount of Fe(III) reduction was achieved (Fig. 1C) than when both soil and NTA were absent (Fig. 1A) in the same time period. When the Fe(III) oxide was retained in a dialysis membrane or entrapped in alginate beads, slight Fe(III) reduction occurred, with a higher rate occurring within the alginate beads (6.2 ± 0.8 mmol/liter in 42 days) than through the dialysis membrane (2 mmol/liter in 43 days). About 15% of the Fe(III) reduction led to the formation of soluble Fe(II) measured in the aqueous-phase material outside the dialysis tubing, while another 15% was measured in the solids recovered from inside the dialysis tubing (Table 1).

TABLE 1.

Concentrations and amounts of Fe(II) and Fe(III) after 70 days in microcosms grown in soil ± NTA with Fe(III) retained within a dialysis membrane or in direct contact with cellsa^a

Fe(III) oxide	Concn, mM (amt, mmol)				Total Fe recovered (mmol)	% Fe(II) of total Fe recovered
	Outside membrane		Inside membrane
		Fe(III)	Fe(II)	Fe(III)			Fe(II)
In direct contact with cells
Abiotic	46.8 (1)	0.6 (0.01)	NAb	NA	1.01	1.0
Biotic
Soil	0.32 (0.01)	36.8 (0.8)	NA	NA	0.81	99
Soil and NTA	0.8 (0.02)	55.0 (1.23)	NA	NA	1.25	98
Separated by dialysis membrane
Abiotic	1.4 (0.03)	0.9 (0.02)	333.6 (0.62)	9.6 (0.02)	0.69	5.8
Biotic
Soil	0 (0)	4.6 (0.10)	259.6 (0.49)	52.9 (0.10)	0.69	30
Soil and NTA	0 (0)	25.9 (0.58)	44.9 (0.08)	53.2 (0.10)	0.76	90

^aThe average initial Fe(II) concentrations in all microcosms outside the dialysis membrane at t=0 days was measured to be 1.2 ± 0.3 mM. The error on the analysis for triplicate samples is ±6.0%.

^bNA, not applicable.

Fe(III) reduction with soil and NTA.

When the cells were grown in the presence of soil and NTA in direct contact with the Fe(III) oxide, the rate of Fe(III) reduction was similar to that of the culture grown in the presence of soil alone, both achieving initial rates of 0.8 mmol/liter/day (Fig. 1C and D). However, in the presence of soil and NTA, the rate and extent of Fe(III) reduction when no barrier was present (31.0 ± 0.5 mmol/liter in 41 days) or across the dialysis membrane (29 mmol/liter in 42 days) or alginate beads (38.0 ± 5.2 mmol/liter in 42 days) were statistically (α = 0.05) higher than with dialysis and alginate microcosms containing soil alone (less than 6.3 ± 0.8 mmol/liter in 42 days) or NTA alone (less than 3.3 ± 0.2 mmol/liter in 42 days).

Fe(III) reduction inside dialysis membrane and alginate beads.

During the course of the experiments, it was observed that the Fe(III) oxide within the membrane and beads changed from a reddish-brown color to a black color in the biotic but not the abiotic microcosms containing soil alone or soil and NTA. There was no attraction between the black Fe(II) minerals and a magnet, indicating that they were not magnetite. At the end of the experiments, measurements of the insoluble iron inside the membrane showed that the Fe(II) content was at least five times higher in the biotic microcosms than for the abiotic control (Table 1). Efforts to measure the Fe(III) and Fe(II) contents inside the alginate beads at the end of the experiment were unsuccessful due to difficulties in extracting all of the iron from within the beads.

In a separate experiment with soil alone, after the dialysis membrane was ruptured at day 85, the concentration of Fe(II) measured in the medium increased immediately from 3 to 12 mmol/liter due to the release of the Fe(II) minerals from the dialysis membrane. Following rupture, reduction of the Fe(III) occurred more rapidly, from 12 to 35 mmol/liter, in 20 days. When the membrane was not ruptured, there was little increase in Fe(III) reduction.

Fe(III) reduction with AQDS with and without soil.

AQDS (50 μM), in the absence of soil, enhanced the rate and extent of Fe(III) reduction when the cells were either grown in direct contact with the insoluble Fe(III) oxides or separated by immobilization of the Fe(III) in the alginate beads (Fig. 1A and 2A). In contrast, AQDS did not improve the rate of Fe(III) reduction through the dialysis membrane even when soil was present (Fig. 1C and 2B). However, the medium did turn an orange color indicative of the accumulation of 2,6-anthrahydroquinone disulfonate, the reduced form of AQDS. Soil and AQDS enhanced the rate of Fe(III) reduction (Fig. 2B) over that with soil alone (Fig. 1C) or AQDS alone (Fig. 2A), particularly when the Fe(III) was immobilized in alginate beads.

FIG. 2.

Change in Fe(II) concentration when a naphthalene enrichment culture was grown with and without soil with AQDS and with amorphous Fe(III) oxide as the terminal electron acceptor. Results with error bars are means for triplicate microcosms, and results without error bars are averages for duplicate microcosms. (□), no barrier; (▴), Fe(III) in alginate; (•), Fe(III) in dialysis membrane; ×, no barrier, abiotic.

Abiotic Fe(III) solubilization with NTA.

To determine whether enhanced Fe(III) reduction in the presence of NTA and soil was due primarily to Fe(III) solubilization by NTA, the effect of soil and NTA on the solubilization of the Fe(III) oxide without an inoculum was examined (Fig. 3). Without soil or NTA, about 0.24 ± 0.06 mmol of dissolved Fe(III)/liter was obtained. With NTA alone, there was a slight increase in the soluble Fe(III) to 0.29 ± 0.09 mmol/liter. With soil alone, the dissolved Fe(III) concentration was lowest. The highest amount of dissolved Fe(III) (0.43 ± 0.08 mmol/liter) was obtained in the medium containing both NTA and soil. However, this is about two orders of magnitude less than the 30 to 40 mmol of Fe(III)/liter which was reduced in the experiments in Fig. 1D. Without inoculum or soil, there was only a small increase in soluble Fe(III), from 0.20 ± 0.04 to 0.52 ± 0.08 mmol/liter as the concentration of NTA was increased from 0 to 75 mmol/liter.

FIG. 3.

The effect of soil and NTA on the concentration of dissolved Fe(III) in the supernatant after centrifugation, with no inoculum added. In all samples, 75 mmol of Fe(III)/liter was added. After 20 h, any undissolved Fe(III) was removed via centrifugation prior to Fe(III) measurement. In samples where indicated, 10% (wt/vol) soil and 75 mM NTA were added. The results are the means for triplicate samples.

Fe(III) reduction with NTA-soil extract.

When NTA was added to any microcosm containing soil, the medium changed from colorless to dark brown. The brown color was not due to Fe(III) solubilization, since the color change occurred even when Fe(III) was omitted and the liquid phase remained colorless in the presence of both NTA and Fe(III) without soil.

The stimulatory effect of the NTA-soil extract was evaluated using the dark brown supernatant obtained by mixing 75 mM NTA in mineral salts medium with soil and using the NTA-extracted soil, which was obtained by repeated extraction until no further color could be removed. The NTA-soil extract was much more effective in improving Fe(III) reduction than the NTA-extracted soil, where some Fe(III) reduction was observed only in the later stage of the culture (Fig. 4). The rate and extent of Fe(III) reduction was highest with unextracted soil, and no Fe(III) reduction occurred with NTA alone. Microcosms containing NTA or the NTA-soil extract and no added carbon source showed no Fe(III) reduction (results not shown), indicating that neither the NTA nor the NTA-extracted material was used as an electron donor.

FIG. 4.

Fe(II) concentration when the naphthalene enrichment culture was grown with 75 mM NTA, 75 mmol of Fe(III)/liter, and with either soil, no soil, NTA-extracted soil, or the brown NTA-soil extract. The results are the means for triplicate samples. (•), plus soil and plus NTA; (▵), NTA-soil extract and NTA; (□), extracted soil and NTA; ×, minus soil and minus NTA; (○), abiotic.

IR analyses.

IR spectra of the NTA-soil extract (Fig. 5A), the alkali-extracted humic substances (Fig. 5B), and commercial sodium salt humic acid (Fig. 5C) were similar to that of soil humic materials previously reported (21) (Fig. 5D). The strong and broad absorption band in the 3,400-cm⁻¹ region in all three spectra is characteristic of humic substances and is attributed to H-bonded OH groups, including those of COOH (32). Also typical of humic substances are bands at 2,920 and 2,860 cm⁻¹. These bands, present in all three samples analyzed, are usually superimposed on the shoulder of the broad OH band (3,400 cm⁻¹) and are attributed to aliphatic C—H stretching (21).

FIG. 5.

IR spectra of NTA-soil extract (A), alkali-extracted humic acid (B), commercial sodium salt of humic acid (C), and a typical commercial humic acid (taken from the work of MacCarthy and Rice [21]; reprinted by permission of John Wiley & Sons) (D).

The 1,720-cm⁻¹ band in humic materials is attributed to C=O stretching by COOH groups. This band largely disappears when the humic substances are converted to the salt forms, in which case new bands appear near 1,575 and 1,390 cm⁻¹ (32), which are due to the carboxylate ion (COO⁻). This observation agrees with the IR results of this study. IR spectra of the NTA-soil extract (Fig. 5A) and the alkali-extracted humic materials (Fig. 5B) exhibited bands at 1,717 and 1,718 cm⁻¹, respectively. However, the spectrum of the commercial humic acid, which was in the form of a sodium salt, showed no band at 1,720 cm⁻¹ but did have bands at 1,617 and 1,395 cm⁻¹. These two bands likely correspond to the 1,575- and 1,390-cm⁻¹ bands reported in the literature for the carboxylate ion (32).

DISCUSSION

At a concentration of 4 mM or less, NTA has been shown to stimulate Fe(III) reduction by acting as a chelator (17). In this study, concentrations of up to 7.5 mM did not stimulate iron reduction in the absence of soil. However, a concentration of 75 mM NTA in our microcosms inhibited iron reduction, probably due to inhibitory effects common to chelating agents at high concentrations (6). This effect was not observed when soil was present, suggesting that a chemical interaction between the soil and the NTA removed this inhibition.

The extent and rate of Fe(III) reduction were significantly higher through dialysis membrane or alginate beads only when both NTA and soil were present. This may be due to (i) the solubilization of the Fe(III) oxide in a process requiring both NTA and soil and/or (ii) the activity of an electron-shuttling compound extracted by NTA from soil.

The largest amount of Fe(III) was solubilized when both NTA and soil were present (Fig. 3), but the increase was slight, with Fe(III) rising from 0.24 ± 0.06 to about 0.43 ± 0.08 mmol/liter. It is doubtful that such a small improvement in soluble Fe(III) could account for the actual increase in Fe(III) reduction observed. Without NTA but in the presence of soil, less than 0.17 ± 0.06 mmol of Fe(III)/liter was solubilized, indicating that the quantity of humic substances extracted from soil by water alone solubilized too little Fe(III) to stimulate Fe(III) reduction. Therefore, it is highly unlikely that the stimulatory effect of soil and NTA on Fe(III) reduction is due to Fe(III) chelation by either NTA or humic substances.

Although NTA alone did not solubilize a significant amount of Fe(III), there was a visible change in the color of the medium, going from clear to dark brown, when NTA was added to the soil-containing microcosms. This color change was not a result of Fe(III) being solubilized, since the brown color also appeared when Fe(III) was omitted. However, there was no color change with NTA and Fe(III) without soil. This demonstrates that NTA was extracting some component of the soil, possibly humic substances. The improved Fe(III) reduction observed with NTA and soil indicated that these extracted materials may be able to diffuse through the alginate beads or dialysis membrane to shuttle electrons between Fe(III) and the cells.

The NTA-soil extract was far more effective in improving Fe(III) reduction than NTA-extracted soil plus NTA, showing that the key soil component was mostly extracted by NTA. These results suggest that the enhanced naphthalene and benzene degradation under Fe(III) reducing conditions reported by Anderson and Lovley (2) after addition of NTA to soils may be due not only to an increase in Fe(III) solubility but also to an NTA-extracted electron-shuttling compound from soil, possibly humic materials. As the humic content of a soil increases, the latter mechanism would play a more important role than the former.

The major bands in the IR spectra of the NTA-soil extract, alkali-extracted humic materials, and commercial humic acid all had similar positions, indicating that all three samples contain similar functional groups. The NTA-soil extract and the alkali-extracted humic materials showed the greatest similarity in spectra. This would be expected if NTA was extracting humic materials, since both samples were prepared using soil from the same source. The ability of the NTA-soil extract to stimulate Fe(III) reduction and its precipitation from solution at pH 1 are properties consistent with humic substances.

Although extraction of humic materials from soil is commonly done using dilute alkali (29), significant auto-oxidation is known to occur (31). Salts of mineral or organic acids, such as sodium pyrophosphate (Na₄P₂O₇) or ammonium oxalate, are frequently used to avoid these chemical changes even though recovery is less (31). These salts are thought to interact with the polyvalent cations that are bound to the humic materials. In many soils, these cations keep the humic substances insoluble. The salts inactivate these cations by forming insoluble precipitates or soluble coordination complexes, leading to the solubilization of the humic material. Extraction by Na₄P₂O₇ occurs as follows (29, 31): R(COO)₄Ca₂ + Na₄P₂O₇ → R(COONa)₄ + Ca₂P₂O₇.

It is likely that the sodium salt of NTA solubilizes humic materials through a similar mechanism. This is supported by the observation that extraction of the soil component was possible by the sodium salt of NTA but not by NTA itself.

Our results suggest that the NTA-soil extract, NTA with soil, or soil alone has the capacity to shuttle electrons to Fe(III). This is, for the most part, consistent with results obtained with a known electron-shuttling compound, AQDS. However, AQDS did not improve the rate of Fe(III) reduction across the dialysis membrane even when soil was present (Fig. 1C and 2B), indicating that Fe(III) reduction across the membrane was due entirely to the effect of soil. However, the culture did reduce AQDS to 2,6-anthrahydroquinone disulfonate as the medium became orange. These results are consistent with a previous finding that AQDS (molecular mass, 412.3 Da) did not promote Fe(III) reduction across dialysis membranes, even with a molecular mass cutoff of 300,000 Da (24).

The mechanism by which NTA (molecular mass, 257.1 Da) in soil enhances Fe(III) reduction across the dialysis membrane and alginate beads is proposed (Fig. 6). NTA solubilizes humic materials from soil by forming complexes with the polyvalent cations bound to the humics. Sodium ions of the NTA salt replace these cations through ion exchange. Fe(III)-reducing bacteria degrade organic carbon by transferring electrons to the extracted humic materials. The extracted humic substances diffuse across the dialysis membrane or alginate bead and shuttle electrons to the Fe(III) within. Some Fe(II) is soluble and diffuses into the aqueous medium; some Fe(II) forms insoluble precipitates which remain within the dialysis membrane or alginate bead. The solubilized humic materials are able to access and reduce more of the insoluble Fe(III), thus indirectly improving Fe(III) bioavailability.

FIG. 6.

Model for NTA stimulation of Fe(III) reduction through dialysis membrane and alginate beads in the presence of soil. Solid lines denote reactions, and dotted lines denote migration of species.

The identity of the black Fe(II) precipitate is not known. Possible products of Fe(III) reduction include magnetite, siderite, and vivianite. While the latter is white, the black precipitate had no magnetic properties. It is possible that some of the precipitate was FeS due to the activity of sulfate-reducing bacteria, since 0.17 mM sulfate was added with the trace element solution. Measurements of the initial aqueous sulfate concentration were at a similar order of magnitude, indicating that little soluble sulfate was associated with the soil or the inoculum. Of the 75 mmol of Fe(III) added per liter, up to 35 to 40 mmol of Fe(II)/liter was formed. This is a far greater amount of Fe(III) reduction than that which could be solely attributed to sulfate-reducing bacteria in this system.

Techniques to accelerate dissimilatory Fe(III) reduction may allow accelerated contaminant degradation when significant amounts of Fe(III) oxide lie in pore spaces inaccessible to iron-reducing bacteria. The investigations presented here provide further insights into the mechanisms by which NTA can enhance Fe(III) reduction. This is the first study in which stimulation of Fe(III) reduction through the addition of chemical chelators in soil is shown to be due to the extraction of electron-shuttling compounds from the soil and not due to Fe(III) solubilization or Fe(II) complexation. These results present potential new avenues of using humic materials and NTA or other chemicals which solubilize humic materials to stimulate Fe(III) reduction for in situ bioremediation. These findings may also contribute to understanding the role that natural organic matter plays in Fe(III) reduction in soils.