Sulfur oxidation precedes ferrous iron oxidation with dispersed sulfur
We serendipitously discovered an alternative sulfur formulation, supplied by a commercial vendor as colloidal sulfur, which surprisingly dispersed readily and evenly upon addition to the media to create a homogenous mixture (Fig. S4). The new formulation, referred to as d-sulfur, enabled A. ferrooxidans to exhibit starkly different consumption patterns for iron and sulfur as compared to c-sulfur even when added at a low concentration of 1.0 g/L. In cultures with d-sulfur, the pH of the solution rapidly decreased to a minimum whereas little ferrous iron oxidation occurred. The bulk of the iron oxidation and the associated pH increase occurred after 36 h (Fig. 2A). When c-sulfur was used instead at 1.0 g/L, we observed the commonly recognized behavior of ferrous iron oxidation followed by sulfur oxidation, indicating that the inoculated cells did not all adsorb to the c-sulfur surface (Fig. 2A). These results suggested that the physical properties of d-sulfur were special in allowing A. ferrooxidans to utilize the sulfur quickly, indicating that the observed consumption order was influenced by the hydrophobicity of the sulfur particles.
At higher concentrations of d-sulfur, the time needed for the completion of the ferrous iron oxidation was increasingly prolonged, demonstrating that additional sulfur repressed iron oxidation (Fig. 2B). At the two highest concentrations tested of d-sulfur, 5.0 g/L and 10.0 g/L, ferrous iron concentration initially decreased and then reversed to increasing over the course of hours (Fig. 2B). The media pH continued to decrease well below the critical pH of 1.3, which is suggested to inhibit iron oxidation, reaching a final pH of 0.73 ± 0.01 and 0.63 ± 0.01 at 168 h respectively, for d-sulfur concentrations of 5.0 and 10.0 g/L (Fig. 2B). With the formulated d-sulfur, the reduction of ferric iron was observed quickly than previously reported as the consumption of dispersed sulfur led to the rapid acidification of the media under aerobic conditions.
The color and different physical properties of d-sulfur compared to c-sulfur indicated that it had been chemically processed prior to being sold to us by the vendor (Fig. S4). Based on its appearance and properties, we hypothesized that the d-sulfur was a water dispersible granule used in agricultural applications, produced by emulsifying sulfur with a lignosulfonate salt (Eller & Person, 1969). As the exact composition of the commercial d-sulfur was unknown, we studied several classes of surfactant compounds and developed a simple sulfur formulation to enable the preferential consumption of elemental sulfur prior to ferrous iron oxidation in A. ferrooxidans. Numerous dispersant candidates were tested, including calcium lignosulfonate, other types of sulfonated compounds, a glycolipid, Tween-20, andA. ferrooxidans cell lysate. All of these compounds dispersed the c-sulfur into water, and dry formulated powders were obtained (Fig. S4). The calcium lignosulfonate dispersed sulfur, referred to as lig-sulfur, dispersed into AFM1 medium similarly to d-sulfur, unlike c-sulfur which immediately phase separated after vigorous vortexing with the medium (Fig. S4).
Once these powders were redispersed into AFM1 medium, only a few of the formulations tested displayed altered sulfur utilization behaviors. Little to no growth was detected with sulfur dispersed with sodium naphthalene sulfonate, poly-(sodium 4-styrenesulfonate), and Tween-20 as no media clarification or color changes were observed when added to the media, suggesting that these chemicals were incompatible with A. ferrooxidans . The substrate utilization behavior at 1 g/L lig-sulfur in conjunction with iron closely mimicked that of d-sulfur (Fig. 2C). Furthermore, the growth of the cells on 10 g/L of lig-sulfur also exhibited the transition from ferrous iron oxidation to ferric iron reduction demonstrating that the formulation we developed induced a similar response from A. ferrooxidans as d-sulfur (Fig. 2C). When the glycolipid sucrose palmitate was used to disperse sulfur, improvements to sulfur oxidation were achieved, but the effects were less pronounced than with the lignosulfonate. Comparison of the 20 (wt %) sucrose palmitate to the 33.3 wt % sucrose palmitate dispersed sulfur cultures demonstrated that there is a critical point for dispersion where the dispersed sulfur becomes bioavailable for oxidation by the cells so that sulfur oxidation largely precedes ferrous iron oxidation. The final pH for the 33.3 wt % sucrose palmitate sulfur was lower at the end of ferrous iron oxidation as compared to the 20 wt %, suggesting that there was faster utilization of sulfur when the additional sucrose palmitate was incorporated into the formulation (Fig. S5). Interestingly, we also found that the components of A. ferrooxidans cell lysate were sufficient for dispersion and enabled elemental sulfur oxidation before ferrous iron oxidation (Fig. S5).
We further characterized the aerobic ferric iron reduction that was potentially detected with growth on high concentrations of dispersed sulfur. To minimize the effects of ferrous iron oxidation by the cells, the ferrous iron was replaced with ferric iron, so that the metabolic energy for growth would have to come from the oxidation of d-sulfur. In the abiotic controls, we found that the d-sulfur itself results in some reduction of ferric iron, and is a primary contributor of the ferrous iron measured initially (Fig. 3). In the cultures with A. ferrooxidans , some additional ferrous iron than the control is measured for the first 40 h as the cells oxidize the d-sulfur, but a major portion of ferric iron reduction is observed as the cultures became more acidic below the pH threshold of 1.3. Over a third of the initial ferric iron in solution was reduced by 168 h when the pH dropped to the acid tolerance limit of A. ferrooxidans (Fig. 3). These data emphasize the observation that ferric reduction behavior continues until cell growth is inhibited by the extremely low pH.