The effect of adsorbed lipid on pyrite oxidation under biotic conditions
© Hao et al; licensee BioMed Central Ltd. 2006
Received: 20 January 2006
Accepted: 25 July 2006
Published: 25 July 2006
Skip to main content
© Hao et al; licensee BioMed Central Ltd. 2006
Received: 20 January 2006
Accepted: 25 July 2006
Published: 25 July 2006
The chemolithoautotrophic bacterium, Acidithiobacillus ferrooxidans, commonly occurs in acid mine drainage (AMD) environments where it is responsible for catalyzing the oxidation of pyrite and concomitant development of acidic conditions. This investigation reports on the growth of this bacterial species on the pyrite surface and in the aqueous phase at a pH close to 2 as well as the role of adsorbed lipid in preventing pyrite dissolution. Both acid washed pyrite and acid-washed pyrite coated with lipids were used as substrates in the studies. The choice of lipid, 1,2-bis(10,12-tricosadiynoyl)-sn-Glycero-3-Phosphocholine lipid (23:2 Diyne PC), a phosphocholine lipid, was based on earlier work that showed that this lipid inhibits the abiotic oxidation rate of pyrite. Atomic force microscopy showed that under the experimental conditions used in this study, the lipid formed ~4–20 nm layers on the mineral surface. Surface-bound lipid greatly suppresses the oxidation process catalyzed by A. ferrooxidans. This suppression continued for the duration of the experiments (25 days maximum). Analysis of the bacterial population on the pyrite surface and in solution over the course of the experiments suggested that the pyrite oxidation was dependent in large part on the fraction of bacteria bound to the pyrite surface.
It is well-known that several species of prokaryotes are able to catalyze the oxidation of pyrite and play an important role in the development of Acid Mine Drainage (AMD), a severe environmental problem. Acidithiobacillus ferrooxidans, for example, utilize reduced Fe2+ in AMD environments as an electron donor for energy production at low pH . Due to the importance of microbes in the chemistry of such environments, a significant amount of research activity has been focused on understanding the role of microbes in the oxidation of sulfur-bearing minerals such as pyrite [2–11]. With regard to microbial-induced pyrite oxidation it is the general consensus that these microorganisms exert their impact on pyrite dissolution to a large degree by increasing the amount of available Fe3+ reactant [i.e., convert Fe2+ to Fe3+], which increases the rate of pyrite oxidation [3, 12, 13]. A goal of our research was to build on this prior research and to extend our understanding of microbial-accelerated pyrite oxidation to surfaces having adsorbed organic layers. With regard to this last point, the inhibition of pyrite oxidation using lipid having two hydrophobic tails (per polar head) has been studied as an AMD abatement strategy [14–16]. However, the lipid-induced inhibition of pyrite oxidation in the presence of microbes relevant to AMD has not been investigated. Toward this end, the experiments reported here were designed to investigate the effect of A. ferrooxidans microorganisms on pyrite oxidation in the presence and absence of adsorbed lipid layer at a solution pH near 2.
Our studies show that surface-bound lipid inhibits the oxidation of pyrite, even in the presence of bacteria. The lipid layers that suppress the oxidation limit the colonization of the bacteria on the pyrite, compared to lipid-free samples. This effect is observed throughout the duration of the experiment (about 25 days in all cases). Hence, in the presence of A. ferrooxidans the lipid coating remains in tact for at least 25 days and probably much longer. Furthermore, our study supports the notion that the rate of pyrite oxidation process is significantly controlled by the bacteria that colonize the pyrite surface.
Pyrite pretreatment followed protocols, described in earlier work,  to generate clean pyrite (~10 μm average diameter) powder with a surface area of 0.75 m2/g. In brief, pyrite samples were crushed and subsequently acid washed using 0.1 M HCl. Acid washing of the pyrite was carried out by flowing pH 2-deoxygenated water over the mineral powder .
A pure culture of A. ferrooxidans (ATTC 23270) was obtained from the American Type Culture Collection (Manassas, VA). The culture was grown and maintained in ATCC #2039 broth as suggested by the supplier. Briefly, the medium consisted of two separately prepared solutions: Solution A contained 0.8 g (NH4)2SO4, 2.0 g MgSO4·7H2O, 0.4 g K2HPO4, 5.0 mL Wolfe's Mineral Solution, and 800 mL distilled water, adjusted to pH 2.3 with H2SO4 and then filter-sterilized. Solution B contained 20.0 g FeSO4·7H2O and 200.0 ml distilled water, which was filter-sterilized and then mixed with Solution A (above) to produce the complete medium. Cultures (9 ml) for experiments were grown unshaken using 100 ml of the medium in 250 ml Erlenmeyer flasks (autoclaved at 126°C and 20 psi for 0.5 h prior to use) at 25°C. The cultures were stoppered with nylon wool wrapped with aluminum foil.
For the experiments, A. ferrooxidans was grown to the early stationary phase of growth and harvested by a series of filtrations. First, the bacteria were separated from particulate material, which are presumably iron hydroxide particles in the media, by filtration through a 5-μm pore size polycarbonate filter (Millipore™). Second, the bacteria in the filtrate were collected on a 0.2 μm polycarbonate filter by vacuum filtration and resuspended in sterile water [pH 2.3] by vortexing. The amount of nutrients in this cell suspension (carried over from the culture) was 1% of the original concentration. This low nutrient concentration was necessary to minimize the impact of aqueous phosphate on pyrite oxidation . Nevertheless, the concentration of nutrients present was sufficient to support growth of the bacteria in the presence of pyrite.
Lipid solutions were prepared by techniques described elsewhere . In the current study, 0.1 g of acid washed pyrite was added to lipid solution (20 ml). The lipid solution contained 7.5 μmol of 23:2 Diyne PC lipid (Avanti Polar Lipids), which has the following structure:
This particular concentration of lipid was chosen on the basis of prior research in our laboratory, and it was sufficient to cover the majority of the pyrite surface . After 1 hour of mixing, the pyrite slurry was allowed to settle and the supernatant, which contained free lipid, was subsequently removed by decanting. The remaining fraction of lipid-coated pyrite slurry was used in the experiments.
The experiment was performed on two separate dates (from here on referred to as Experiment 1 and 2). The experimental conditions, treatments and analyses for both experiments were identical with the exception of the starting bacterial densities (see below). The treatments consisted of: 1) pyrite, 2) lipid-coated pyrite, 3) pyrite + bacteria, 4) lipid-coated pyrite + bacteria. All the treatments were prepared, in duplicate, in distilled water (pH 2.0) for a total volume of 50 ml. The treatments with pyrite contained 0.1 g of pyrite. Approximately the same amount of pyrite was present in the pyrite/lipid treatments (see preparation of lipid-coated pyrite slurry described above). The final density of A. ferrooxidans in the treatments with bacteria was approximately 3.3 × 107cells/ml for Experiment 1 and 1.7 × 107cells/ml for Experiment 2. Each treatment mixture was stirred continuously (using a teflon coated stir bar) throughout the experiment, which lasted a maximum of 25 days. Aliquots (1 ml) were withdrawn every 2–3 days for cell count analysis and preserved in 3.7% formaldehyde. In addition, 0.1 ml aliquots were withdrawn to determine the iron concentration in solution, which was used as a reaction progress variable. Total iron solution concentration determinations (including ferrous and ferric iron) were determined spectrophotometrically by using the ferrozine technique (UV method 8008). Since this technique is intrinsically sensitive to the concentration of ferrous, ascorbic acid was added to the solution prior to analysis to reduce any ferric to ferrous. All the sample aliquots for the ferrozine test were filtered through a 0.2 μm polycarbonate filter before analysis. Solution pH measurements were carried out with a PHM82 standard meter with a resolution of 0.01 pH units.
Enumeration of bacterial cell densities in solution and on pyrite powder were performed by epifluorescence microscopy following staining with the fluorochrome dye, acridine orange (AO) [18, 19]. Briefly, preserved water samples were vacuum filtered onto 0.2-μm-pore-size blackened polycarbonate filters (Nuclepore) and the cells collected on the filter were stained with AO at a final concentration of 0.01% for 3 min. The volume of sample filtered for cell counts was chosen such that the layer of pyrite particles that accumulated on the filter would not obscure the bacteria. Cells were observed and counted with a Zeiss standard microscope equipped for epifluorescence microscopy. Filter sets used for AO fluorescence observations are as follows: a BP450-490 exciter filter, and FT510 chromatic beam splitter, and an LP520 barrier filter. For each sample, the number of bacteria on 10 to 15 randomly chosen fields distributed over the filter (or approximately 300 to 600 cells in total per filter) was counted. To determine the number of cells bound to pyrite, 0.1 ml aliquots (initial total volume was 50 ml) of the bacteria/pyrite solutions were collected on a 5 μm polycarbonate membrane filter (Millipore). The particles were subsequently washed so that any bacteria in solution would pass through the filter. The density of cells on the particulate material remaining on the filter was then determined using the staining method described above.
Surfaces for AFM were prepared from pyrite cubes (Navajun, Spain) obtained from Ward's Natural Science. Growth surfaces were cut from the cube and were subsequently polished. Prior to an experiment, the samples were exposed to pH 2 water (by HCl addition) to remove oxidized layers prior to exposing to lipid solution. To prepare the lipid/pyrite samples the pyrite platelet was submerged into the lipid solution in a vertical orientation (via a leading edge). The sample was then immediately removed, washed with pH 2 water to remove any loosely bound lipid, and then reintroduced to pH 2 water for imaging with a PicoSPM II (Molecular Imaging) microscope. The probes used in all the AFM measurements [NSC15, μMasch] had a nominal spring constant of 40 N/m and a resonant frequency of 325 kHz.
It is generally accepted that both dissolved molecular oxygen and aqueous Fe3+ play a role in the oxidation of pyrite [12, 20]. The overall composite reactions for pyrite oxidation in the presence of both these oxidizing agents can be expressed as follows:
FeS2 + (7/2)O2 + H2O → Fe2+ + 2SO4 2- + 2H+ (1)
FeS2 + 14Fe3+ + 8H2O → 15Fe2+ + 2 SO4 2- + 16H+ (2)
Prior studies of pyrite oxidation have shown that Fe(III) is a more aggressive oxidant of the pyrite surface than dissolved molecular oxygen [21, 22]. Under strictly abiotic conditions, however, the concentration of Fe3+ in solution is relatively low, due to the slow oxidation kinetics associated with the conversion of Fe2+ product [see Eqn. (1)] to Fe3+ by dissolved O2. Hence, the contribution of eqn. (2) to pyrite oxidation under abiotic conditions is limited by the rate of ferrous iron oxidation, which is low in pH 2 solutions . It is precisely this slow abiotic oxidation step that is catalyzed by microbes, such as A. ferrooxidans. The subsequent oxidation of the mineral by the aqueous Fe3+ is referred to as the "indirect" mechanism of pyrite oxidation . This mechanism is in contrast to the "direct" mechanism, that involves the oxidation of the pyrite by surface colonized bacteria (either by enzymatic or non-enzymatic means) [8, 20].
There are at least three important observations that can be made. First, the rate of oxidation is the highest for experiments in which pyrite without a lipid pretreatment is exposed to A. ferrooxidans. Second, if the pyrite is exposed to lipid, prior to the exposure to the bacteria, there is a significant decrease in the amount of pyrite oxidation. Specifically, over the last 10-day period (where the solution is supporting the steady growth of the bacteria; see below) of the experiment, there is a >4-fold decrease in the oxidation rate when pyrite has an adsorbed lipid layer. Third, while lipid suppresses the amount of bacterial-induced pyrite oxidation, the rate of oxidation in the presence of microbes is higher than in the absence of microbes. Hence, the presence of the lipid on the pyrite surface does not entirely suppress the influence of the bacteria. For example, over the last 10-day period, the amount of oxidation associated with the bacteria/lipid/pyrite system was a factor of three higher than the rate of oxidation associated with the abiotic lipid/pyrite system, and it was very similar (within experimental error) to the rate of oxidation associated with untreated pyrite in pH 2 water (i.e., no bacteria or lipid).
Summary of experimental observations for the amount of pyrite oxidation in the presence of lipid and bacteria.
Oxidation rate* (10-8M s-1m-2)
% Oxidation Suppression#
1.62 ± 0.1
0.30 ± 0.02
0.088 ± 0.005
0.21 ± 0.01
Figure 3 shows the distribution of A. ferrooxidans density between the surface bound and solution fraction as a function of time for the bacteria/lipid/pyrite system. Independent data sets for the two experimental runs are shown, and aside from differences in the initial cell densities at time-zero, the data from both experiments reveal similar trends. After a slight decrease in bacterial density in the bacteria/lipid/pyrite treatment over the initial 5 days (most evident in Figure 3b), there is an increase in the solution concentration of bacteria and a decrease in the surface-bound fraction (Figure 3a and 3b). We attribute the initial decrease of bacteria in solution to the lack of substrate (the treatments were prepared in DI water) required to support bacterial growth during the early stages of the experiment. Presumably, there is residual uncoated (i.e., lipid-free) pyrite present and its dissolution at the beginning of the experiment starts to provide nutrients for the bacteria so that a net growth can be achieved after the 5-day period. This experimental observation also is the reason why Table 1 compiles oxidation rate data for the last 10-day period of each experiment. In contrast, the surface bound fraction of bacteria shows an experimentally resolvable decrease between 5 and 20 days when lipid is present.
Figure 4a and 4b exhibit bacterial density measurements for the bacteria/pyrite system (no lipid). Again two independent experimental data sets are shown. While the solution density of bacteria in the pyrite and lipid/pyrite system is similar [compare Figure 4b (Exp. 1) to Figure 3b (Exp. 1)], the bacterial density on the surface of the lipid-free pyrite after 20 days is more than a factor of 10 greater than that of the lipid/pyrite system [compare Figure 4a (Exp. 1) to Figure 3a (Exp. 1)]. Additionally, the surface-bound fraction in the lipid-free circumstance continues to grow over the entire 25-day experiment, in contrast to the decrease associated with the lipid-coated pyrite case. Perhaps, the most revealing experimental observations concerned with the data presented in Figure 3 and 4 are as follows (N.B., values associated with Exp. 1 are used in the following discussion for convenience). First, the solution concentration of bacteria in the bacteria/pyrite system after the 20-day period is approximately 15% higher than in the bacteria/lipid/pyrite system (4.8 × 107compared to 4.1 × 107 cells/ml, respectively). Second, after this same time period, the attached cell density for the bacteria/pyrite system is more than an order of magnitude (20×) higher than for the bacteria/lipid/pyrite system (6.4 × 105 compared to 2.3 × 104 cells/cm2, respectively). Furthermore, the attached cell density for the lipid pyrite system actually decreases over the 20-day period (from 4.0 × 104 to 2.3 × 104 cells/cm2), while the surface cell density for the lipid-free system underwent more than a three-fold increase (from 1.9 × 105 to 6.4 × 105 cells/ml). These data should be considered in view of the rate data compiled in Table 1, which shows that the oxidation rate for the bacteria/pyrite system is more than a factor of 4 greater than the bacteria/lipid/pyrite system. Since, the solution concentration of bacteria is within about 15% for both these systems, it is difficult to attribute the difference in oxidation rate entirely to differences in the concentration of this fraction of bacteria. Instead, we argue that the difference in oxidation rate between the bacteria/lipid/pyrite and bacteria/pyrite systems must be largely due to differences in the concentrations of surface colonized bacteria for the two systems. Only the lipid-free pyrite shows a significant increase in the surface-bound fraction of bacteria, in contrast to the decreasing concentration of surface-bound bacteria on lipid/pyrite over the course of the experiment. Prior studies have in general shown the indirect mechanism to dominate, [3, 13] but during the early stages of the pyrite oxidation process the relative contribution of the direct mechanism can be significant . We suspect that the growth rates of the surface attached and solution phase bacteria achieved in our study (which are non-exponential growth rates) are consistent with an early stage of the oxidation process, where bacterial growth in solution is limited by the availability of aqueous Fe2+ (resulting from the dissolution of the mineral surface). In general, the importance of the surface-bound bacterial fraction for pyrite oxidation is well appreciated by prior research. It has already been shown in prior research, for example, that bacteria attached to the pyrite surface has significant effects on the oxidation process, due in part to the microbe's influence on the evolution of the mineral surface and contacting solution [7, 26, 27].
The mechanism by which the lipid suppresses the pyrite oxidation reaction appears to be connected to the formation of a surface coating with a hydrophobic pocket. The AFM analysis of this coating shows a relatively thin coating ranging from 7 to 16 nm, within the experimental error of our measurement. Prior results in our laboratory suggested that another phosphocholine lipid, L-α-Phosphatidylcholine, Hydrogenated (Egg, Chicken), assembles into a bilayer structure on the pyrite surface. Such a structure would consist of a phosphate group in a phosphocholine headgroup binding to the mineral, and a second phosphate group extending into the aqueous phase, leaving an hydrophobic interior. This prior model was first deduced from vibrational studies of the lipid/mineral interaction,  but more recent atomic force microscopy (AFM) results lend support to this model . In particular, imaging of the lipid/pyrite (platelet sample) surface at similar conditions to those used in the prior study (i.e., pyrite surface area to lipid concentration ratio) find that the layer height of the lipid is consistent with what would be expected for a bilayer, although these structures occupy the surface along with thicker or mulitilayer structures . Our imaging results of the 23:2 Diyne lipid in the present study are consistent the presence of bilayer structures. The chain length of the particular lipid used in this study is approximately 3.5 nm. Hence, a bilayer might of been expected to be about 7 nm, or two stacked bilayers would be ~14 nm, similar to some of our experimentally determined heights. Certainly, a more detailed study of the surface is needed to make any conclusive arguments concerning the existence of a bilayer structure. Perhaps, more importantly, we infer from our AFM results that the pyrite powder used in our study would likewise be covered by a lipid coating under similar conditions. We further argue that this coating both limits the amount of microbial adhesion and the reaction of oxidants, such as Fe3+, and water with the mineral surface. Finally, the results from the present study also show that the 23:2 Diyne PC structure is stable in the presence of A. ferrooxidans under our experimental conditions for at least a period of 25 days.
The amount of pyrite oxidation in the presence of bacteria and with an adsorbed phosphocholine layer has been investigated. The presence of A. ferrooxidans greatly increases the amount of pyrite oxidation relative to the bacteria-free system. Pretreatment of pyrite with the phosphocholine lipid reduces its oxidation rate compared to untreated pyrite. By measuring the density of surface bound and solution bacteria it was determined that under our experimental conditions the amount of pyrite oxidation is more a function of the surface-colonized bacteria concentration than the solution bacterial fraction. Our results also show that the lipid oxidation barrier layers are stable in the presence of A. ferrooxidans for at least 25 days under our experimental conditions. A separate study will be conducted to evaluate the fate of the surface-bound lipids in the presence of heterotrophic bacteria.
D.R.S. and M.A.S. greatly appreciate support from the Department of Energy, Basic Energy Sciences from grants DEFG029ER14644 and DEFG0296ER14633, respectively. This research also was supported by the Center for Environmental Molecular Science (CEMS) at Stony Brook that is funded by the National Science Foundation (CHE-0221934). We also acknowledge the assistance of Mark A. Randa (Temple University) for help with culturing A. ferrooxidans, and Sudeep Debnath (Temple University) for help in obtaining the AFM images.
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.