OUP user menu

Iron respiration by Acidiphilium cryptum at pH 5

Azize Azra Bilgin, JoAnn Silverstein, Joy D. Jenkins
DOI: http://dx.doi.org/10.1016/j.femsec.2003.08.018 137-143 First published online: 1 July 2004


The growth of acidophilic iron respiring bacteria at pH > 4.5 may be a key to the transition from acidic to circumneutral conditions that would occur during restoration of acid mine drainage sites. Flasks containing Acidiphilium cryptum ATCC 33463 were incubated initially under aerobic conditions in liquid medium containing Fe2(SO4)3 and glucose at an initial pH of 5. Significant iron respiration was observed after flasks were sealed to prevent oxygenation; at the same time, medium pH increased from 4.5 to 6. No soluble Fe(III) was detected throughout the experiments, consistent with pH conditions, indicating that bacteria were able to respire using precipitated ferric iron species. In addition, the concentration of soluble Fe2+ reached a plateau, even though iron respiration appeared to continue, possibly due to precipitation of mixed Fe (II)/Fe(III)-oxide as magnetite. Results suggest that A. cryptum has a wide range of pH tolerance, which may enable it to play a role in controlling acid generation by means of establishing growth conditions favorable to neutrophilic bacteria such as sulfate reduction.

  • Acid mine drainage
  • Iron respiration
  • Acidiphilium cryptum
  • Bioremediation

1 Introduction

There have been numerous efforts to remediate acid mine drainage (AMD) environments. Reducing the flow of oxygen and/or water to disturbed rock piles and seeps can be done with soil barriers or inundation of formations [1]. Other treatment methods include neutralizing the acidic water by addition of lime or soda ash or employing the neutralization capacity of wetlands [24]. These methods have often proved to be costly and produce wastes that require further treatment and disposal.

Autotrophic iron-oxidizing bacteria play a major role in catalyzing AMD at pH below 4 [5]. An alternative to treating acidified metal-laden drainage is in situ prevention of pyrite oxidation by enhancement of the growth of native non-iron oxidizing heterotrophic bacteria [6]. The growth of heterotrophic bacteria in waste rock piles is carbon limited in many AMD environments. If an external carbon source is supplied, heterotrophic bacteria will consume limited supplies of oxygen and produce anoxic conditions favorable to iron and sulfate reduction leading to neutralization of drainage water and metal immobilization [7,8]. Because sulfate reducing bacteria do not grow readily under acidic conditions, it was hypothesized that bacterial iron respiration may be key to the transition from pyrite oxidation and acid generation to sulfate reduction and metal ion precipitation that would characterize long-lasting AMD site restoration [9].

The implications of sustaining iron-reducing conditions in the waste rock environment are substantial [6,10]. First, ferric iron, which serves as the oxidant of pyritic minerals, is reduced to the less oxidized state, ferrous iron, halting further abiotic pyrite oxidation. Second, the increase in pH accompanying iron reduction of solid phase ferric iron could serve to precipitate many otherwise soluble metals, significantly lowering the metal loading to the environment. Finally, the combination of higher water pH, lower redox conditions, and biogenic organic carbon from heterotrophic growth could sustain other bacterial processes such as sulfate reduction [6,8,11,12].

Numerous strains of acidophilic, heterotrophic bacteria have been isolated from highly acidic environments (pH < 3.0) [12,13]. Relatively few of these bacteria have been fully characterized. Those that have been characterized belong mainly to the genus Acidiphilium, such as Acidiphilium cryptum[12], A. angustum, A. facilis and A. rubrum[14] and A. organovorum[15]. Members of the Acidiphilium genus are gram-negative, obligately acidophilic, rod-shaped bacteria, and they have also been reported to be obligate aerobes [12,13]. However, more recently researchers have identified a significant number of Acidiphilium and Acidiphilium-like bacteria that carry out dissimilatory reduction of ferric iron under anoxic and microaerophilic conditions [16,17].

Harrison et al. [12] isolated A. cryptum directly from coal refuse and also as a contaminant in cultures of A. ferrooxidans isolated from coal mine drainage and from a copper ore leaching site. Harrison [18] suggested that metabolites such as pyruvate produced by autotrophic iron oxidizing bacteria might provide sufficient carbon and energy for growth of A. cryptum. He observed that A. cryptum grew on organic bacterial by-products such as pyruvic acid, which is a selection advantage in acid rock drainage environments where organic carbon typically is scarce.

In a recent study Kusel et al. [19] compared the growth of A. cryptum JF-5 on several carbon sources. Glucose was consumed under anoxic conditions without apparent delay and stimulated the formation of Fe(II) and CO2, and the pH increased from 3.2 to 5.8. The pH of abiotic control microcosms changed only slightly from 3.2 to 3.6, indicating that the larger pH increase was associated with bacterial iron reduction. The pH tolerance of A. cryptum in the range of 2.1–5.8 was investigated in aerobic growth conditions, and an optimum pH for aerobic growth was reported to be 3.2.

This study was an investigation of the hypothesis that iron respiration is a key reaction for restoration of acid mine drainage generation sites by mediating drainage water pH increase to near neutral conditions. It has been reported that iron respiration by A. cryptum raised medium pH from 3.2 to 5.8, however, the initiation of anoxic growth of the acidophile at pH > 5 has not been studied. Of particular interest is the ability of iron respiring acidophiles to grow under anoxic conditions on solid phase ferric iron that would be the dominant Fe species at pH > 5.

2 Materials and methods

2.1 Bacteria cultures

Acidiphilium cryptum (ATCC 33463) was the heterotrophic iron reducing strain used in all experiments. Inocula were obtained from the American Type Culture Collection (ATCC, Manassas, VA). Cultures were maintained using the glucose minimal medium described by Wichlacz and Unz [13].

2.2 Shake flask experiments

Experiments were performed in 500 ml Erlenmeyer flasks mounted on a shaker table (New Brunswick Scientific Co., Inc., Edison, NJ) agitated at 200 rpm. The initial liquid volume of each flask was 300 ml, with 3 ml samples removed periodically for chemical analyses. Sterility was ensured by autoclaving the flasks and liquid solutions. At the beginning of the experiments inoculated flasks were covered with cotton plugs to prevent contamination of the culture while allowing oxygen transfer into the flask. After a period of aerobic growth, flasks were capped with butyl rubber septa. Sampling continued as the headspace oxygen was consumed and anoxic reactions occurred. Liquid samples were taken using a sterile 21-gauge needle and syringe. Ferric sulfate [Fe2(SO4)3· 5H2O] (Fisher Scientific, Pittsburg, PA) was used as the source of ferric iron and added in the amount of approximately 18 mM (1 g l−1) of Fe3+. pH adjustment prior to the beginning of each experiment was accomplished using 2 N NaOH. Glucose (C6H12O6) (Fisher Scientific) was the carbon/energy substrate for heterotrophic bacterial growth in all experiments. Flask media contained 1 g l−1 organic carbon (TOC) from glucose (equivalent to 14 mM glucose) at the beginning of the experiment. Additional glucose was supplied to the one set of flasks at 19th day in the amount of 1 g l−1 (14 mM glucose). All flasks also contained modified LHET2 medium [20]: 2 g of (NH4)2SO4 per liter, 0.5 g of K2HPO4 per liter, 0.5 g of MgSO4· 7H2O per liter, 0.1 g of KCl per liter, 0.002 g of NaCl per liter, 0.106 g of tryptic soy broth per liter and 0.1 g of yeast extract per liter; the pH was adjusted to 5 with 2 N NaOH in all flasks. A control flask containing sterile media was incubated under the same conditions as the bacterial growth flasks.

The DAPI (4′,6-diamidino-2-phenylindole) staining method was used to quantify the initial cell density of A. cryptum in all the inoculated growth flasks [21]. Efforts were made to insure a uniform inoculum cell density for each flask. The average initial cell concentration for the replicate flasks was 8.0 × 107± 1.6 × 107 cells ml−1 as measured by DAPI.

2.3 Determination of iron concentration and speciation

Both soluble and solid-phase iron were determined spectrophotometrically using 1,10-phenanthroline according to a modification of the procedure in Standard Methods [6,22]. Soluble ferrous iron chelates with 1,10-phenanthroline to form an orange complex that was measured at 510 nm on a UV160U spectrophotometer (Shimadzu Corporation, Kyoto, Japan). For the purpose of this study, soluble iron was defined as the fraction of iron that passed through a 0.2 μm filter; the same filter size used to remove bacteria from solution. Ferric iron in the filtrate was reduced to the ferrous form using the chemical reducing agent hydroxylamine hydrochloride (NH2OH · HCl) (Fisher Scientific) to measure total iron as described above. The soluble ferric iron concentration was calculated by subtracting ferrous from total iron. Unfiltered samples were digested at 150 °C for 15 min in 25% hydroxylamine hydrochloride to convert solid and soluble Fe3+ to Fe2+, which was measured as described above.

2.4 Total organic carbon pH, and dissolved oxygen measurement

Total soluble organic carbon (TOC) was used to monitor consumption of glucose during the experiments. The TOC of filtered samples was measured as the non-purgable organic carbon fraction using an autoanalyzer (TOC 5000, Shimadzu Corporation, Kyoto). Reactor pH was monitored using a calomel electrode and pH meter (Accumet AB15 Basic & Biobasic) (Fisher Scientific). Dissolved oxygen (DO) was measured using a DO probe and meter (Model 52 YSI Incorporated, Yellow Springs, OH).

3 Results and discussion

Approximately 9 mM (0.5 g l−1) of soluble Fe2+ accumulated in the seven days after aeration ended in flasks containing 18 mM (1 g l−1) Fe(III) inoculated with A. cryptum (at 17th day), as can be seen in Fig. 1(a). During the same time, dissolved oxygen decreased from 0.19 mM (6 mg l−1) to less than 0.03 mM (1 mg l−1), and pH increased from 4.5 to 6, as shown in Fig. 1(b) and (c), respectively. No soluble Fe2+ was observed in an abiotic control flask with sterile iron-glucose medium. pH in the control reactor decreased from 4.95 to 4.51.


(a) Soluble Fe2+ and Fe3+ concentrations in flask reactors during oxic (0–17 days) and anoxic (17–50 days) incubation of A. cryptum. Aeration was stopped after 17 days. Data points are averages of samples from three replicate flasks. Error bars are ±1 SD. (b) Dissolved oxygen concentration in flask reactors during anoxic (17–50 days) incubation of A. cryptum. Aeration was stopped after 16 days. Dissolved oxygen concentration during oxic incubation was not measured since the flasks were aerated. Data points are averages of samples from three replicate flasks. Error bars are ±1 SD. (c) pH change oxic (0–17 days) and anoxic (17–50 days) incubation of A. cryptum. Aeration was stopped after 17 days. Data points are averages of samples from three replicate flasks. Error bars are ±1 SD. (d) Glucose consumption measured as soluble TOC. Glucose consumption continued throughout the anoxic growth period although soluble Fe2+ accumulation stopped at 23 days. Error bars are ±1 SD.

Although the total iron initially measured in the reactors was 17 mM (0.96 g l−1), only a negligible amount was in soluble form, indicating that the bacteria were obtaining the Fe(III) electron acceptor from solid-phase species, which is consistent with pH > 5. Fig. 2(a) is a micrograph of DAPI-stained A. cryptum species growing in close association with Fe precipitates. Suspended cells were also observed in the flask medium, as shown in Fig. 2(b), of cells stained after the medium was passed through 5 μm membrane filters to remove most of the precipitated iron.


DAPI stained A. cryptum cells under anoxic conditions at pH 5: (a) associated with iron precipitates which can be seen as diffuse fluorescent patches at the bottom of graph indicating solid surfaces (a); and (b) after filtration of flask medium through 5 μm filters. Micrographs are 100× magnification.

After approximately 23 days, the soluble Fe2+ concentration reached a plateau. It was thought that this plateau could have been a result of Fe2+ precipitation. The equilibrium chemistry model MINTEQ A2 was used to identify the iron minerals that might be precipitating when pH was above 5. Using the flask media formulation, vivianite, melanterite and magnetite were identified as supersaturated minerals. Magnetite is an iron (II)/iron (III) precipitate[(Fe2+)2(Fe3+)O3](s) and its formation would be a sink for both Fe(II) and Fe(III) ions. There have been reports of magnetite precipitation above pH 5 in previous research. Lovley [23] reported magnetite as a common end product of ferric iron-reducing neutrophilic bacteria. Cummings et al. [24] observed magnetite precipitates in sediments at Lake Coeur d'Alene, Idaho (CDAR) with pH conditions between 5.2 and 6. It was suggested that Fe(III)-reducing microorganisms reduced hydrous ferric oxides to magnetite. Comparative 16S rRNA sequence analyses showed that the bacteria belonged to the Geobacteraceae family [24]. Production of magnetite by A. cryptum could have been a sink for ferrous iron produced during respiration, explaining the plateau in soluble Fe2+. Magnetite forms a black precipitate, which was not observed in the flask solids. However, X-ray diffraction analyses were conducted on the flask precipitates at the end of the experiments and the results suggest that the precipitates could contain magnetite (Fig. 3). However, since the concentrations are just above the detection limits these results are not conclusive.


X-ray diffraction profiles for iron precipitates in flask media. Top graph shows the precipitate peaks. Bottom graph shows the magnetite standard peak at 35.3°. A small peak (∼20 CPS) at 35.3° was measured in the flask precipitates (top profile).

Soluble TOC data also support an Fe(II) precipitation explanation. The theoretical stoichiometry for iron respiration of ferric hydroxide species, predicts that four moles of Fe2+ is produced per mole of glucose-TOC consumed, neglecting cell yield Embedded Image 1 Fig. 1(d) is a profile of TOC consumption during both aerobic and anoxic growth. In this experiment, approximately 14 mM glucose (1 g l−1 TOC) was consumed and 9 mM ferrous iron (Fe2+) was produced during iron respiration, which is equivalent to an Fe(II) produced: TOC consumed molar ratio of 0.11, indicating that bacteria were using significantly more TOC than would be predicted by Eq. (1). Linear regression analysis was used to evaluate the slopes of the average Fe2+ accumulation and TOC consumption profiles during anoxic growth after the Fe2+ plateau was reached. The slope of the Fe2+ concentration profile was 1.4 × 10−7 M (7.8 × 10−6 g)-Fe2+ l−1 h−1, which is not statistically different from zero (p= 0.82). In other words, there was no significant increase in soluble Fe2+. At the same time, the TOC concentration profile slope during the same period was −0.014 mM-glucose l−1 h−1 (−1 mg-TOC l−1 h−1) differing from zero slope (no TOC consumption) with p= 0.14 an indication that TOC consumption continued although no Fe2+ accumulated. In addition, the pH of the media in the inoculated flasks continued to increase after Fe2+ accumulation had stopped, from 5.84 to 6.02, as shown in Fig. 3. The slope of the regression line fitted to the pH data was significantly greater than zero (p= 0.02), consistent with bacteria using ferric hydroxide(s) and consuming protons during iron respiration (Eq. (1)). An alternate explanation is that diffusion of headspace oxygen into the flask media supported aerobic oxidation of the glucose substrate, as evidenced by the slow decline in dissolved oxygen during the anoxic period from 19 to 50 days shown in Fig. 1(b). However, this is not consistent with the pH increase observed.

After 50 days, flask media were centrifuged at 10,000g for 15 min and transferred to fresh ferric iron and glucose media. Re-inoculated flasks were sealed to prevent oxygen transfer although the headspace gas was not sparged. Initial pH was adjusted to 5.2. Profiles for soluble ferrous and ferric iron are shown in Fig. 4(a). As in the previous experiment, negligible dissolved ferric iron was detected. Soluble Fe2+ had begun to appear by the first sampling point, at six days of incubation, and dissolved oxygen had been mostly consumed by this time, as shown in Fig. 4(b). After 31 days of anoxic incubation (dissolved oxygen less than 0.03 mM (1 mg l−1)), the three-replicate flask average soluble Fe2+ was approximately 4 mM (223 mg l−1), significantly lower than the value after similar reaction time in the previous experiment, 8.6 mM (483 mg l−1) (p= 0.05). For this experiment, statistical analysis indicated that Fe2+ was continuing to accumulate. The slope of the soluble ferrous iron profile was positive and significantly greater than zero (p= 0.05). In addition, although pH increased from 4.59 to 5.29 as shown in Fig. 4(c), this was significantly less than the increase observed in the first experiment. Thus while the trends of Fe2+ production from solid-phase Fe(III) respiration accompanied by utilization of TOC and pH increase were similar between the first and second experiments, iron respiration by A. cryptum appeared to be slower in the second experiment.


(a) Soluble Fe2+ and Fe3+ concentrations in flask reactors during non-aerated incubation of A. cryptum. Data points are averages of samples from three replicate flasks. Error bars are ±1 SD. (b) Dissolved oxygen in flask reactors during non-aerated incubation of A. cryptum. Data points are averages of samples from three replicate flasks. Error bars are ±1 SD. (c) pH profiles during non-aerated incubation of A. cryptum. Data are averages of samples from three replicate flasks. Error bars are ±1 SD. (d) Soluble TOC profiles during non-aerated incubation of A. cryptum. Data points are averages of samples from three replicate flasks. Error bars are ±1 SD.

One explanation could be the difference between viable cell concentration that was transferred from the first flasks to fresh media in the second experiment. Lower number of active cells would account for the slower rate of iron reduction observed. Furthermore, there was no period of vigorous aeration in the second experiment. It would be expected that more cell growth would occur under aerobic conditions, resulting in a greater number of active cells at the start of the first experiment than the second.

Approximately 0.83 mM glucose (60 mg l−1 TOC) was consumed after dissolved oxygen had decreased to approximately 0. 03 mM O2 (1 mg l−1) and 4.5 mM (250 mg l−1) soluble Fe2+ accumulated (Fig. 4(d)). This is equivalent to an Fe2+:TOC molar ratio of 0.9:1, as in experiment 1, significantly lower than the stoichiometric ratio of 4:1.

4 Conclusions

An acidophilic heterotrophic bacterial strain, Acidiphilium cryptum, used ferric iron hydroxide precipitates as terminal electron acceptors during dissimilatory iron reduction in media above pH 5. Iron reduction by acidophiles at near-neutral pH support the feasibility of enhancing bacterial iron reduction by enrichment of heterotrophic acidophilic bacteria to neutralize acidic drainage water at the source. Furthermore, results suggest that the much of the iron reduced by bacteria may precipitate as pH rises preventing release of soluble iron into drainage water.


  1. [1]
  2. [2]
  3. [3]
  4. [4]
  5. [5]
  6. [6]
  7. [7]
  8. [8]
  9. [9]
  10. [10]
  11. [11]
  12. [12]
  13. [13]
  14. [14]
  15. [15]
  16. [16]
  17. [17]
  18. [18]
  19. [19]
  20. [20]
  21. [21]
  22. [22]
  23. [23]
  24. [24]
View Abstract