Determining the controls on metal bioavailability is crucial in order to understand the geomicrobiology of geologic systems with high metal concentrations, such as contaminated soils, and in order to optimize bioremediation strategies aimed at remediating those systems. Biotic ligand models of metal bioavailability characterize metal binding onto a wide range of organisms using a generic, unspecified metal-binding biotic ligand that does not account for the many complexities of metal adsorption reactions onto biological surfaces. These models often fail because of the heterogeneous, complex nature of cell surfaces.

In this study, we use our detailed understanding of the metal binding reactions on bacterial cell walls to create an advanced biotic ligand model that relates metal toxicity to the speciation and concentration of the metal adsorbed to the bacterial surface. We used batch measurements of Cd toxicty to Bacillus subtilis to validate and calibrate this approach. Bacterial growth was measured by optical density in the presence of a constant concentration of Cd at either 1 or 2 ppm. EDTA was added to control the bacterial adsorption of Cd, using EDTA:Cd molar ratios of 0, 0.25, 0.5, 1 and 2. A toxicity factor was calculated for each Cd-bearing experiment as the optical density after 24 hours of growth divided by the 24-hour optical density of a Cd-free control. In all cases, a minimal growth medium was used to limit variables, allowing the precise calculation of aqueous and surface speciation of Cd for each experimental condition. These calculations used previously determined site-specific binding constants, acidity constants and site concentrations for the B. subtilis cell wall functional groups.

Cd toxicity increased with decreasing EDTA concentration, and we found a strong correlation between the total concentration of adsorbed Cd and the measured toxicity factor. Our results suggest that biotic ligand models that incorporate more sophisticated models of the metal binding environment on cell walls will be more flexible and accurate than previous bioavailability models. Therefore, bacterial biotic ligand models that include mechanistically-sound metal binding reactions will dramatically improve our understanding and ability to predict metal bioavailability in complex geologic systems.