Correspondence: Fang-Bai Li, Guangdong Key Laboratory of Agricultural Environment Pollution Integrated Control, Guangdong Institute of Eco-Environmental and Soil Sciences, No. 808, Tianyuan Road, Guangzhou 510650, China. Tel.:+86 20 8702 4721; fax: +86 20 8702 4123; e-mail: firstname.lastname@example.org
This work studied the ability of Comamonas koreensis CY01 to reduce Fe(III) (hydr)oxides by coupling the oxidation of electron donors and the enhanced biodegradation of 2,4-dichlorophenoxyacetic acid (2,4-D) by the presence of Fe(III) (hydr)oxides. The experimental results suggested that strain CY01 can utilize ferrihydrite, goethite, lepidocrocite or hematite as the terminal electron acceptor and citrate, glycerol, glucose or sucrose as the electron donor. Strain CY01 could transform 2,4-D to 4-chlorophenol through reductive side-chain removal and dechlorination. Under the anaerobic conditions, Fe(III) reduction and 2,4-D biodegradation by strain CY01 occurred simultaneously. The presence of Fe(III) (hydr)oxides would significantly enhance 2,4-D biodegradation, probably due to the fact that the reactive mineral-bound Fe(II) species generated from Fe(III) reduction can abiotically reduce 2,4-D. This is the first report of a strain of C. koreensis capable of reducing Fe(III) (hydr)oxides and 2,4-D, which extends the diversity of iron-reducing bacteria associated with dechlorination.
Diverse phylogenetic microorganisms are capable of coupling the oxidation of organic substrates with the reduction of Fe(III) minerals (Lovley et al., 2004; Scala et al., 2006; O'Loughlin, 2008). These iron-reducing bacteria are thought to play an important role in the cycle of iron and carbon, and also in the transformation processes of organic contaminants (Lovley, 1987, 1991; Fredrickson & Gorby, 1996) and metals (Lovley, 1993; Caccavo et al., 1994; Lovley et al., 2004) in anaerobic soil and sedimentary environments. Through the microbial metabolism of iron-reducing bacteria, organic contaminants may be biodegraded either as an electron donor or as an electron acceptor. For example, Geobacter metallireducens can oxidize toluene, phenol or p-cresol to carbon dioxide with Fe(III) as the electron acceptor in the aquatic sediments (Lovley & Anderson, 2000). Krumholz et al. (1996) found that a strain of Desulfuromonas capable of reducing Fe(III) can reductively dechlorinate tetrachloroethylene to cis-dichloroethylene.
Fe(III) is often present in the environments with chlorinated organic contaminants; thus, there occurs competition between Fe(III) and chlorinated organic contaminants as terminal electron acceptors. Kazumi et al. (1995a, b) first studied the biodegradation of haloaromatic compounds under Fe(III)-reducing conditions and found that haloaromatic compounds were biodegraded via a process coupling with Fe(III) reduction.
2,4-Dichlorophenoxyacetic acid (2,4-D) has been the most commonly used acidic phenoxy herbicide throughout the world for >60 years. The widespread usage of 2,4-D has resulted in serious ecological issues such as toxic effects on birds, beneficial insects, soil annelids, and nontarget plants and interferences with the growth of fish, amphibians and algae, particularly in their juvenile stages (Cox, 1999; Willemsen & Hailey, 2001; Chinalia & Killham, 2006). Previous studies mainly focused on the anaerobic biotransformation of 2,4-D by mixed cultures (Boyle et al., 1999; Zipper et al., 1999; Berestovskaya et al., 2000) and the factors influencing 2,4-D biodegradation (Boyle et al., 1999). Recently, it was found that the reduction rates of the herbicide were significantly increased by the addition of quinoid redox mediators, which were enzymatically reduced to the corresponding hydroquinone that can chemically reduce 2,4-D (Wang et al., 2009). However, the effect of Fe(III) on 2,4-D biotransformation under anaerobic conditions is not well understood, especially using a pure culture.
Comamonas koreensis CY01 was previously isolated from the subterranean forest sediment (Sihui, China). It is highly versatile in utilizing terminal electron acceptors, including humic substances, ferric citrate and 2,4-D (Wang et al., 2009), which makes it a promising model organism for studying the potential interaction between Fe(III) and 2,4-D. With the aim to understand how a microorganism adapts in the presence of both Fe(III) (hydr)oxide and 2,4-D, batch experiments were conducted to characterize Fe(III) reduction and the effect of Fe(III) reduction on 2,4-D degradation by strain CY01. The specific objectives were to (1) explore the Fe(III) reduction by strain CY01; (2) determine the anaerobic degradation pathway of 2,4-D by CY01; and (3) investigate the interaction between Fe(III) reduction and 2,4-D reductive dechlorination by CY01 under anaerobic conditions.
Materials and methods
Ferrihydrite was prepared using a standard method adapted from Roden & Urrutia (2002). Goethite (α-FeOOH) was synthesized according to the procedures of Li et al. (2008). Preparation of lepidocrocite (γ-FeOOH) and hematite (α-Fe2O3) followed the method described by Li et al. (2007). Methanol and acetic acid of HPLC grade were purchased from Shanghai Reagent Co. (China). 2,4-D, phenol, 2,4-dichlorophenol (2,4-DCP) and 4-chlorophenol (4-CP) of analytical grade were obtained from Sigma-Aldrich (Tokyo, Japan). All of the chemicals were used as received, without further purification. All stock solutions were prepared using deionized water.
Organism, media and cultivation
Comamonas koreensis CY01 (Wang et al., 2009) was isolated from the subterranean forest sediment in Sihui City, China. It was aerobically cultivated in Luria–Bertani (LB) medium at 30 °C. The cells were harvested at the late log phase by centrifugation (8000 g at 4 °C for 10 min), washed twice and resuspended with a sterilized mineral salts medium (MSM).
The MSM was used for all the Fe(III) reduction and 2,4-D transformation experiments. The medium contained the following components (amounts per liter of deionized water): 2.5 g NaHCO3, 0.25 g NH4Cl, 0.60 g NaH2PO4, 0.10 g KCl, 10 mL vitamin stock solution and 10 mL mineral stock solution (Zachara et al., 1998).
Strict anaerobic techniques and sterile conditions (Li et al., 2009) were used throughout the Fe(III) reduction and 2,4-D transformation test. The medium was dispensed into serum bottles (25.2 mL) and sterilized by autoclaving at 115 °C for 20 min. Other components (i.e. electron donors, 2,4-D and cell suspension) were added from the stock solutions after the medium was cooled to the ambient temperature. Then the bottles were purged with O2-free N2/CO2 (80/20 v/v) for 15 min, sealed with butyl-rubber stoppers and crimped with aluminum caps. All bottles were incubated in the dark at 30 °C, which was optimal for the microbial growth of strain CY01.
Fe(III) reduction by strain CY01
In the tests for using Fe(III) (hydr)oxides as alternative electron acceptors, each bottle contained 1 × 107 cells mL−1, 20 mL MSM with 5 mmol L−1 of glucose (electron donor) and one of the following Fe(III) (hydr)oxides as an electron acceptor (25 mmol L−1): ferrihydrite, α-FeOOH, γ-FeOOH and α-Fe2O3. For the experiments testing alternative electron donors, each bottle included 1 × 107 cells mL−1, 20 mL MSM with 25 mmol L−1 ferrihydrite (electron acceptor) and one of the following substrates as an electron donor (5 mmol L−1): formate, acetate, propionate, lactate, citrate, ethanol, glycerol, glucose and sucrose. In the treatment using ferrihydrite and glucose, the microbial growth was monitored.
2,4-D transformation experiments
In order to study 2,4-D transformation, a modified chloride-free medium was prepared, which contained the same compositions as the above MSM, except that all chloride salts were replaced with the corresponding sulfate salts. For 2,4-D transformation by strain CY01, each bottle was inoculated with 1 × 107 cells mL−1, 20 mL chloride-free medium with 180 μmol L−1 2,4-D and 5 mmol L−1 glucose. In the test studying the effect of Fe(III) (hydr)oxide on 2,4-D transformation by the isolate, 25 mmol L−1 of α-FeOOH was introduced into the same system described above.
For both Fe(III) reduction and 2,4-D transformation by strain CY01, all the treatments were conducted in triplicate. For each test, three control arrays were performed under the same conditions: one set without cells, one set with heat-killed cells (dead cells) and a biotic set without an electron donor/acceptor.
Analyses and calculations
During a 25-day incubation period, triplicate bottles were used for chemical analysis at an interval of 5 days. The reduction of Fe(III) was monitored by measuring the concentration of total Fe(II) and dissolved Fe(II). Total Fe(II) was determined by extraction with 0.5 mol L−1 HCl for 1.5 h (Fredrickson & Gorby, 1996), followed by colorimetric analysis with 1,10-phenanthroline (Lovley, 1997). Dissolved Fe(II) was measured by filtering a 1-mL suspension through a 0.22-μm syringe filter directly into 1,10-phenanthroline and reading the A510 nm immediately (Roden & Zachara, 1996). The cell numbers were determined by direct plate counting on aerobic LB agar medium.
Before the analysis of 2,4-D, 2,4-DCP, 4-CP, chloride and glucose, the samples were centrifuged at 4000 g for 10 min and filtered (Polyvinylidene fluoride, 0.22 μm; Millipore Inc.). The concentrations of 2,4-D and its potential degradation intermediates were analyzed by HPLC (Waters 1527/2487). HPLC used a reverse-phase C18 column (4.6 × 250 mm) and the conditions were as follows: UV detector wavelength, 285 nm; the isocratic mobile phase, methanol : ultrapure water : acetic acid, 60 : 38 : 2 (v/v/v); a flow rate of mobile phase, 1 mL min−1; and injector volume, 10 μL. The detection limit of each compound was about 3 μmol L−1. Chloride released from 2,4-D transformation was measured using ion chromatography (Dionex ICS-90) with an ion column (IonPac AS14A 4 × 250 mm). The mobile phase consisted of Na2CO3 (8 mmol L−1) and NaHCO3 (1 mmol L−1) solution, operated at a flow rate of 1 mL min−1. Glucose was determined spectrophotometrically using the method described by Miller (1959).
The Fe(III) reduction ratio was calculated as follows: Rre (%)=C/C0× 100, where C0 is the initial Fe(III) concentration and C is the final concentration of Fe(III) after 25 days.
Fe(III) reduction by strain CY01
The Fe(III)-reducing activity of strain CY01 was explored with four types of Fe(III) (hydr)oxides: ferrihydrite, α-FeOOH, γ-FeOOH and α-Fe2O3. After 25 days (Fig. 1), no Fe(II) was formed in the biotic control without glucose and <0.5 mmol L−1 of Fe(II) was observed in the abiotic control. In contrast, the total Fe(II) concentration in the active tests using ferrihydrite, α-FeOOH, γ-FeOOH or α-Fe2O3 as the electron acceptor reached 2.5, 1.6, 1.4 and 0.8 mmol L−1, respectively. The results suggested that (1) Fe(III) reduction by CY01 was a biological process because it required both active cells and glucose; (2) the four types of Fe(III) (hydr)oxides can be reduced by strain CY01 with glucose as the electron donor and (3) the reduction ratio of ferrihydrite was the highest (10%), followed by α-FeOOH (6.2%), γ-FeOOH (5.7%) and α-Fe2O3 (3.0%).
Nine types of organic substrates were tested as the alternative electron donors for Fe(III) reduction. As shown in Fig. 2, the production of Fe(II) increased with time in the incubations with citrate, glycerol, glucose and sucrose, while no Fe(II) was detected in the incubations with formate, acetate, propionate, lactate and ethanol (data not shown). The concentration of produced Fe(II) ranked in the order of sucrose>glucose>glycerol>citrate. In the abiotic control, nearly no Fe(II) was detected, except that about 0.5 and 0.1 mmol L−1 of Fe(II) were, respectively, observed in the sets with glucose or sucrose, probably due to the chemical redox reaction between Fe(III) and glucose/sucrose. These results indicated that sucrose, glucose, glycerol and citrate could serve as favorable electron donors for Fe(III) reduction by strain CY01, and the reduction ratio depended on the type of organic substrates. Organic substrates such as formate, acetate, propionate, lactate and ethanol were not available electron donors with respect to Fe(III) reduction.
In the active test using glucose as the sole electron donor and ferrihydrite as the electron acceptor, the cell numbers of C. koreensis CY01 increased over time over a 25-day incubation period (Fig. 3a). Relative to insignificant microbial growth in the control sets, the cell numbers in the active test reached a maximum density of 6.5 × 107 cells mL−1, which was approximately seven times greater than the initial density. As indicated in Fig. 3b, as C. koreensis CY01 grew, the total Fe(II) concentration increased at the cost of glucose.
Transformation of 2,4-D by strain CY01
To determine the pathway of 2,4-D transformation by strain CY01 under anaerobic conditions, changes in the concentrations of 2,4-D, intermediates (2,4-DCP, 4-CP and phenol) and chloride were measured (Fig. 4). Over a 25-day incubation period, 2,4-D decreased from an initial concentration of 180 to 150 μmol L−1 (Fig. 4a), accompanied by the consumption of glucose (data not shown). At the same time, intermediates of 2,4-DCP and 4-CP were observed (Fig. 4b), but no phenol was detected. In the control sets, the concentration of 2,4-D remained unchanged and no intermediate was detected, suggesting that 2,4-D reduction was a microbial process. The chlorine concentrations of 2,4-D, 2,4-DCP, 4-CP and anion chloride in the 2,4-D/glucose/cells system were calculated to establish the chlorine mass balance (Fig. 4c). Within 25 days, the 2,4-D-Cl concentration decreased from 359 to 300 μmol L−1. Concurrently, the concentration of 2,4-DCP-Cl reached a maximum of 20 μmol L−1, followed by a gradual decline to <5 μmol L−1. At the same time, 4-CP-Cl and anion chloride accumulated over time, and their amounts approached stability at the end of the incubation. Throughout the process of 2,4-D transformation, the chlorine mass recovery, the ratio of the sum of chloride compounds including 2,4-D, 2,4-DCP, 4-CP and anion chloride to the initial chlorine in 2,4-D, expressed in percentage, ranged from 101% to 105%. The nearly 100% mass recovery of chloride suggested that no intermediates other than 2,4-DCP and 4-CP were produced from 2,4-D degradation.
Effects of Fe(III) on 2,4-D transformation by strain CY01
α-FeOOH was added to the system of 2,4-D transformation by strain CY01 to investigate the effect of Fe(III) (hydr)oxide on 2,4-D anaerobic degradation. Figure 5a shows the changes in the concentrations of 2,4-D, 4-CP and Fe(II) over the 25-day incubation period in the treatment of 2,4-D/glucose/cells/α-FeOOH. As 2,4-D was degraded, the concentrations of 4-CP and total Fe(II) increased, indicating that 2,4-D dechlorination and Fe(III) reduction by CY01 can occur simultaneously. Figure 5b showed the amounts of 2,4-D and 4-CP in the treatments of 2,4-D/glucose/cells with and without α-FeOOH. In the treatments with and without α-FeOOH, approximately 27% and 17% 2,4-D were degraded, yielding 25% and 11% 4-CP, respectively. The results indicated that 2,4-D biotransformation can be enhanced by the presence of α-FeOOH.
The strain CY01 was able to reduce Fe(III) (hydr)oxides using different organic substrates. Among the factors influencing the microbial reduction of Fe(III) (hydr)oxide, the physical and chemical characteristics of Fe(III) are considered to be one of the most important effects (Liu et al., 2001; Roden, 2003). Zachara et al. (1998) found that the reduction of synthetic Fe(III) (hydr)oxides by Shewanella putrefaciens strain CN32 was consistent with their surface area and free energy, ranking as ferrihydrite>α-FeOOH>α-Fe2O3. Consistent with their findings, the extent of Fe(III) reduction in our study followed an order of ferrihydrite>α-FeOOH>γ-FeOOH>α-Fe2O3. Among these Fe(III) (hydr)oxides, ferrihydrite underwent reduction most easily, probably due to its poor crystalline property, while α-Fe2O3 reduction proceeded with more difficulty than others, likely a consequence of its smallest surface area (29.4 m2 g−1). On the other hand, the microbial reduction of Fe(III) also depended on the type of electron donors. The experimental results showed the extent to which the ferrihydrite reduction varied as a sequence of sucrose>glucose>glycerol>citrate, while the other alternatives such as formate, acetate, propionate, lactate and ethanol could not be utilized for Fe(III) reduction by strain CY01.
Under anaerobic conditions, 2,4-D was reduced to 4-CP by strain CY01. Previous studies (Boyle et al., 1999; Walters, 1999; Berestovskaya et al., 2000) suggested that 2,4-DCP and 4-CP are the most commonly detected metabolites from 2,4-D anaerobic degradation. In this study, the degradation pathway was proposed as an initial transformation of 2,4-D to 2,4-DCP via hydrolysis, which was followed by 2,4-DCP dechlorination to 4-CP (Fig. 6a). Microorganisms have a preference in transforming ortho- to para-chlorine (Kohring et al., 1989), which explained the detection of 4-CP rather than 2-CP. This was consistent with the findings reported by Chang et al. (1998), who observed an accumulation of 4-CP from 2,4-D biodegradation by soil organisms. According to previous literatures, 4-CP, as a product of 2,4-D degradation, can be further mineralized or dechlorinated to phenol (Kohring et al., 1989; Häggblom & Young, 1995; Häggblom, 1998; Robles-González et al., 2006). For instance, Robles-González et al. (2006) detected phenol as one of the intermediate metabolites of 2,4-D in the sulfate-reducing slurry bioreactor after a 30-day incubation period. Similarly, Kohring et al. (1989) observed that the production of 4-CP from 2,4-DCP degradation ceased after a 25-day incubation period, and approximately all 4-CP decomposed to phenol in the subsequent 50 days. However, no further dechlorination of 4-CP was observed in this study. The transformation of 4-CP to phenol may be expected if the incubation period is extended.
As shown in Fig. 5a, the maximum decline in the 2,4-D concentration took place at the time point when the Fe(II) concentration decreased, suggesting a possibility of an interaction between 2,4-D and Fe(II). At the same time, the data in Fig. 5b suggested an enhancement of 2,4-D transformation in the presence of α-FeOOH. In order to clarify the mechanism for enhanced transformation, a supplementary experiment was conducted. The system of glucose/cells/α-FeOOH was first operated under the same conditions as previous tests, autoclaved after a 15-day incubation period, after which 2,4-D was introduced into the system and incubated for another 10 days. As can be seen in Fig. 7, over the latter 10-day incubation period, the concentration of 2,4-D was decreased by 25% and 17% chloride was recovered; meanwhile, the sorbed-Fe(II) produced from the first incubation decreased from 1.19 to 0.57 mmol L−1. Previous experiments have demonstrated that the heat-killed cells and medium itself cannot react with 2,4-D (Fig. 4a), and so the 2,4-D transformation in the latter incubation seemed to be a consequence of the chemical reductive activity of biogenic sorbed Fe(II) as the dechlorination activity of strain CY01 was inhibited by autoclaving. This mechanism was similar to that found in previous studies (Amonette et al., 2000; Tobler et al., 2007), which indicated that the enhancement by Fe(III) oxides could be mainly attributed to the mineral-bound Fe(II) species. The mineral-bound Fe(II) species are found to be able to enhance the transformation rates of many reducible pollutants significantly. For instance, Amonette et al. (2000) reported that under anoxic conditions, carbon tetrachloride could be dechlorinated to chloroform by Fe(II) that was sorbed on the surface of α-FeOOH, while no reaction occurred when Fe(II) was present and α-FeOOH was absent. A recent study reported that the mineral-bound Fe(II) species, which resulted from Fe(III) reduction by G. metallireducens, was capable of the abiotic reduction of 4-nitroacetophenone (Tobler et al., 2007).
Therefore, it can be concluded that an Fe(III)–Fe(II) redox cycle was involved in 2,4-D transformation, enabling the degradation of 2,4-D by CY01 strain in the presence of α-FeOOH to proceed in two steps (Fig. 6b): (1) electrons were transferred from glucose to Fe(III) by microbial respiration, yielding Fe(II) and energy; (2) the electrons were then transferred to 2,4-D by mineral-bound Fe(II) species, leading to 2,4-D reduction and reoxidization of Fe(II) to Fe(III). In conclusion, the mechanism of Fe(III)-enhanced 2,4-D anaerobic biotransformation included the microbial reduction of Fe(III) to Fe(II), followed by the abiotic reduction of 2,4-D to 4-CP by the biogenetic mineral-bound Fe(II) species.
To the best of our knowledge, this study constitutes the first report for the anaerobic biodegradation of 2,4-D linked to Fe(III) reduction by a pure strain. The ability of strain CY01 to enhance dechlorination by Fe(III) (hydr)oxide makes it one of the few isolates reported to be able to do so. Further studies should be conducted to further confirm the interactive mechanism among strain CY01, Fe(III) (hydr)oxides and 2,4-D from a new perspective at the molecular level and determine whether structurally analogous pesticides can undergo a similar transformation by CY01 strain.
This research was supported by the National Natural Science Foundation of China (no. 40601043, 20777013 and 40801119).