Kinetic Constant Variability in Bacterial Oxidation of Elemental Sulfur

Abstract

Wide ranges of growth yields on sulfur (from 2.4 × 10¹⁰ to 8.1 × 10¹¹ cells g⁻¹) and maximum sulfur oxidation rates (from 0.068 to 1.30 mmol liter⁻¹ h⁻¹) of an Acidithiobacillus ferrooxidans strain (CCM 4253) were observed in 73 batch cultures. No significant correlation between the constants was observed. Changes of the Michaelis constant for sulfur (from 0.46 to 15.5 mM) in resting cells were also noted.

Elemental sulfur is a fundamental substrate for sulfur-oxidizing bacteria. It can also serve as an intermediate in oxidation of metal sulfides. Acidithiobacillus ferrooxidans is an important acidophilic chemolithoautotrophic bacterium involved in formation of acid mine drainage and biohydrometallurgy (1, 8, 10, 11, 15). The growth yield on sulfur (Y_X/S) is related to metabolic utilization of the sulfur substrate. A higher Y_X/S corresponds to a higher number of sulfur electrons that are transferred via reverse electron transport (2) to form NADH for CO₂ reduction. We reported earlier that 3.5% of sulfur was used for reverse electron transport (3). This portion of sulfur served directly for cell growth and the remainder for ATP formation. The existing Y_X/S data (3, 4, 6, 12, 13), however, were obtained from only a few cultures, and thus, conclusions regarding the variability of the growth yields, the subject of this study, cannot be made.

Media with dimethyl sulfoxide and Tween 80 (16), acetone (7, 14), and ground flowers of sulfur suspended in water (7) were reported to increase elemental sulfur solubility (or wettability) in resting-cell suspensions. Elemental sulfur dissolved in acetone has previously been used to determine the apparent Michaelis constant for sulfur (K_m) (14).

The purpose of this study was to investigate variability in metabolism of elemental sulfur using Y_X/S, maximum sulfur oxidation rate (v), and K_m as indicators.

For these experiments, A. ferrooxidans strain CCM 4253 (Czech Collection of Microorganisms) was grown on elemental sulfur (Sulfur Extra Pure; Riedel-deHaën, Germany) at 28°C in three ways: in a 10-liter bioreactor and 500-ml shake flasks as described earlier (3) and in a bioreactor (Biostat B-DCU; B. Braun Biotech International, Germany) with a 5-liter polyetheretherketone glass vessel (FairMenTec, Germany). This strain was shown to be highly related (100% identity) to the type strain (ATCC 23270) of A. ferrooxidans by 16S rRNA gene sequencing.

Y_X/S was determined as the slope of a linear dependence of the cell count on sulfate concentration. v was determined from the linear rate of sulfate formation. The sulfate concentration and the number of bacteria were determined isotachophoretically (9) and turbidimetrically (3), respectively.

Bacteria for the resting-cell suspension experiments were harvested by centrifugation from the growing culture when the pH reached 1; they were washed and resuspended in a basal salt medium (3) to a final concentration of 7 × 10⁹ bacteria per ml. Commercial colloidal sulfur (colloidal sulfur powder; Riedel-deHaën, Germany) was used as the sulfur substrate at 8 g per 100 ml water to determine K_m. The sulfur oxidation rate was determined as the oxygen uptake rate (Q) at 28°C using an oxygen electrode of the Clark type (Electrofact, The Netherlands). All values were corrected for low endogenous oxidation rates. In experiments at different temperatures, the electrode calibration for dissolved oxygen was made for each temperature separately. K_m was determined as a slope in the dependence of Q on Q/S, where S is actual sulfur concentration in cell suspension.

Although the values of Y_X/S and v within the same experimental run remained unchanged (within experimental error), the two constants demonstrated significantly different values in different experiments (Fig. 1). Some of the differences exceeded 1 order of magnitude. Y_X/S was not affected by v even where v decreased due to low pH or low sulfur concentration (not shown in the summarized data in Fig. 1). The 95% confidence intervals of Y_X/S (cells g⁻¹) for A. ferrooxidans strain CCM 4253 of (2.7 ± 0.5) × 10¹¹ and (2.3 ± 1.2) × 10¹¹ (3) were comparable to 6.25 × 10¹¹ for another strain of A. ferrooxidans (13), where no experimental error was indicated. Similar yields for Acidithiobacillus thiooxidans, 4.9 × 10¹¹ (6), 2.1 × 10¹¹ (12), and 8.4 × 10¹¹ (4), have also been determined. All the above values have not been out of the wide range of Y_X/S values in Fig. 1. However, no variability within the strain has been observed because the above data have corresponded to only one or a very few replicates. In contrast, the data in Fig. 1 represent a much higher culture number, in part under different conditions or time periods. The significant changes in Y_X/S might indicate, e.g., a highly variable number of sulfur electrons that are available for reverse electron transport.

FIG. 1.

Relationship between growth yield on sulfur (Y_X/S) and maximum sulfur oxidation rate (v) determined in 73 batch cultures of the A. ferrooxidans strain CCM 4253. Cultures included 10-liter cultures in a bioreactor with initial sulfur concentrations of 5 to 20 g/liter (•); 5-liter cultures in a bioreactor (B. Braun Biotech International, Germany) with an initial sulfur concentration of 10 g/liter (×); and 100-ml cultures in shake flasks with initial sulfur concentrations of 1 to 20 g/liter (-), 10 g/liter at an initial pH of 1 to 2 (▵), and 10 g/liter at an initial sulfate concentration of 4 to 160 mM (○).

Figure 1 shows no correlation between Y_X/S and v for most of the data. In fact, lower Y_X/S values appeared at the highest v values, which could indicate a downward trend of Y_X/S; however, these data are insufficient to allow formation of additional conclusions.

Using the commercial colloidal sulfur in resting-cell suspensions, the sulfur oxidation rate was much higher than those for the above-mentioned sulfur substrate media (not shown). Data in Fig. 2 (inset) demonstrated a linear dependence of K_m on cell concentration, indicating a cell steric effect. The K_m expressed per single cell was 6.2 ± 0.5 fmol cell⁻¹ (95% confidence interval). As the saturation constant for sulfur is considered to be constant during growth (3), growth and nongrowth conditions reveal different impacts on the limiting substrate concentration. The duration of cell suspension storage in the absence of sulfur also affected K_m (Fig. 2). This might be related to changes over time in cell surface or enzyme structure conformation. No significant effect of pH (at a pH of 7, 4, 3, 2, or 1) on K_m was observed (P > 0.05, not shown), although a pH optimum for sulfur oxidation ranged between 3 and 6. In contrast, two pH optima at 3 and 6 were detected in the acetone medium (14), where K_ms were 10.0 and 5.5 mM at the respective pHs. However, no experimental error was indicated to demonstrate the significance in the K_m difference. Compared to 25°C, the significant temperature effect (at 15, 25, 30, 35, or 40°C [not shown]) on K_m was observed only at 40°C (but only at P < 0.01). Although Michaelis-Menten kinetics was applicable, K_m was significantly dependent on cell concentration and cell storage time. In contrast, pH and temperature did not represent factors influencing K_m.

FIG. 2.

Dependence of the apparent Michaelis constant for sulfur on culture storage time at 4°C and on cell concentration (in the inset, r = 0.999) in resting cells. The bars show the standard deviations.

The considerable variability of the growth constants Y_X/S and v detected in many experiments showed the complexity of sulfur oxidation even within a single strain of A. ferrooxidans. No relationship between the growth constants was observed. As a result of this, the simple comparison and prediction of the kinetic constants are uncertain. Quite wide variations in the data from repeated experiments even under identical conditions may be anticipated, but such results would provide much more information regarding the metabolic capacity of bacteria under particular culture conditions and better knowledge in order to apply the real constant values in kinetic models. It is unclear whether the kinetic constant variability observed here is related to the property of sulfur metabolism in the studied strain or whether this is the case with other (acidi)thiobacilli or other sulfur-oxidizing bacteria (5). Changes in K_m were related to physiological and experimental conditions. In addition to different forms of sulfur substrate, the factors which affected K_m should be taken into consideration to obtain a clear characterization of the strain and its ability to metabolize substrates.

Nucleotide sequence accession number.

The 16S rRNA gene sequence of A. ferrooxidans strain CCM 4253 was deposited in GenBank under accession no. EF465493.