Thilo Behrends

What do acid-base titrations of live bacteria tell us? A preliminary assessment

Abstract.To gain insight into the non-equilibrium processes that affect the titration curves of bacteria, we performed pH stat experiments with suspensions of live cells of the Gram-negative bacterium Shewanella putrefaciens. The experiments lasted for 5 hours, during which acid or base addition was monitored. Periodically, the electrophoretic mobility of the cells, as well as the buffer capacity and the concentrations of cations and dissolved organic carbon (DOC) of the solution, were measured. At the end of the experiments, the viability of the cells was determined. In a limited number of cases, final solutions were screened for the presence of cell-wall constituents using gel electrophoresis. The results showed a very different behavior of the cell suspensions under acid and alkaline conditions. At pH 4, acid addition ceased after 20 minutes. The cells remained intact but were no longer viable at the end of the experiment, while little change in the buffer capacity of the solution was observed. The data at pH 4 were consistent with protonation of cell wall functional groups. At pH 8 and 10, base addition continued during the entire duration of the experiments. The cells remained viable, and the buffer capacity and DOC concentration of the solutions increased with time. Gel electrophoresis indicated that proteins and lipopolysaccharides had been released to solution at pH 10. In contrast to pH 4, the buffering capacity of the bacterial cells under alkaline conditions did not appear to be limited by the initial availability of ionisable functional groups in the cell wall. This preliminary study shows that a complex set of processes, including active metabolic responses, control the acid-base activity of live cell suspensions.

Biosorption of metals (Cu2+, Zn2+) and anions (F−, H2PO4−) by viable and autoclaved cells of the Gram-negative bacterium Shewanella putrefaciens

Microbial biomass represents a potentially cost-effective sorbent for water treatment applications. High sorption capacities for both cations and anions are demonstrated here for viable and autoclaved cell suspensions of the Gram-negative bacterium Shewanella putrefaciens. FTIR absorption spectra and pH-dependent zeta-potentials are similar for the viable and killed bacterial cells. Potentiometric titrations, however, reveal a two to three times higher OH(-) buffering capacity for the living cells. The Cu(2+) sorption capacity of the viable cells is also about twice that of the autoclaved cells. Sorption of fluoride and phosphate is not pH-dependent, although an initial addition of acid or base was needed to activate the anion binding sites. Uptake of fluoride is comparable for viable and killed cells. For the viable cells, the isotherms of Zn(2+) and Cu(2+) indicate the presence of at least two distinct populations of cell wall binding sites. In competitive sorption experiments, Cu(2+) completely inhibits the binding of Zn(2+) to the cells at aqueous concentrations above 150 mg L(-1). The release of dissolved organic compounds by the viable cells depends on the concentrations of metal cations or fluoride to which the cells are exposed. In particular, the presence of Cu(2+) nearly completely suppresses the release of protein-like substances, possibly reflecting Cu(2+) toxicity.

Systematic study of effects of pH and ionic strength on attachment of phage PRD1.

Objectives of this work are to investigate effects of pH and ionic strength (IS) on virus transport in saturated soil and to develop a quantitative relationship for these effects. A series of 50-cm column experiments with clean quartz sand under saturated conditions and with pH values of 5, 6, 7, 8, and IS values of 1, 10, and 20 mM were conducted. Bacteriophage PRD1 was used as a model virus. Applying a one-site kinetic model, attachment, detachment, and inactivation rate coefficients were determined from fitting breakthrough curves using the software package Hydrus-1D. Attachment rate coefficients increased with decreasing pH and increasing IS, in agreement with DLVO theory. Sticking efficiencies were calculated from the attachment rate coefficients and used to develop an empirical formula for sticking efficiency as a function of pH and IS. This relationship is applicable under unfavorable conditions for virus attachment. We compared sticking efficiencies predicted by the empirical formula with those from field and column experiments. Within the calibrated range of pH and IS, the predicted and observed sticking efficiencies are in reasonable agreement for bacteriophages PRD1 and MS2. However, the formula significantly overestimates sticking efficiencies for IS higher than 100 mM. In addition, it performs less well for viruses with different surface reactivity than PRD1 and MS2. Effects of pH and IS on detachment and inactivation rate coefficients were also investigated but the experimental results do not allow constraining these parameters with sufficient certainty.

Transformation of Hematite into Magnetite During Dissimilatory Iron Reduction—Conditions and Mechanisms

Magnetite formation during the reduction of nanoparticulate hematite by Shewanella putrefaciens 200R is investigated in media of variable composition, at circumneutral pH and with lactate as electron donor. The relative rates of production of dissolved Fe(II) and Fe(III), aqueous speciation, plus chemical gradients control whether or not magnetite forms in the experiments. High bicarbonate concentrations result in the precipitation of magnetite, presumably by enhancing the non-reductive dissolution of hematite, hence causing the simultaneous production of soluble Fe(III) and Fe(II) in the incubations. Magnetite formation is inhibited when hematite dissolution is slowed down by adsorption of oxyanions (phosphate and sulfate) at the mineral surface, when the reduction of soluble Fe(III) is enhanced by increasing the cell density or adding an electron shuttle (AQS), or when aqueous Fe(II) is complexed by ferrozine. In experiments where hematite suspensions with and without bacteria are separated by a dialysis membrane, magnetite formation occurs mainly in the cell-free portion of the reaction system. Most likely, precipitation of magnetite is favored in the cell-free suspension because of a higher soluble Fe(III) to Fe(II) ratio. The formation of magnetite in the absence of cells further implies that its nucleation is not catalyzed by the bacterial surfaces.

Oxygen dependency of neutrophilic Fe(II) oxidation by Leptothrix differs from abiotic reaction

Neutrophilic Fe(II) oxidizing microorganisms are found in many natural environments. It has been hypothesized that, at low oxygen concentrations, microbial iron oxidation is favored over abiotic oxidation. Here, we compare the kinetics of abiotic Fe(II) oxidation to oxidation in the presence of the bacterium Leptothrix cholodnii Appels isolated from a wetland sediment. Rates of Fe(II) oxidation were determined in batch experiments at 20°C, pH 7 and oxygen concentrations between 3 and 120 μmol/l. The reaction progress in experiments with and without cells exhibited two distinct phases. During the initial phase, the oxygen dependency of microbial Fe(II) oxidation followed a Michaelis-Menten rate expression (KM = 24.5 ± 10 μmol O2/l, vmax = 1.8 ± 0.2 μmol Fe(II)/(l min) for 108 cells/ml). In contrast, abiotic rates increased linearly with increasing oxygen concentrations. At similar oxygen concentrations, initial Fe(II) oxidation rates were faster in the experiments with bacteria. During the second phase, the accumulated iron oxides catalyzed further oxidative iron precipitation in both abiotic and microbial reaction systems. That is, abiotic oxidation also dominated the reaction progress in the presence of bacteria. In fact, in some experiments with bacteria, iron oxidation during the second phase proceeded slower than in the absence of bacteria, possibly due to an inhibitory effect of extracellular polymeric substances on the growth of Fe(III) oxides. Thus, our results suggest that the competitive advantage of microbial iron oxidation in low oxygen environments may be limited by the autocatalytic nature of abiotic Fe(III) oxide precipitation, unless the accumulation of Fe(III) oxides is prevented, for example, through a close coupling of Fe(II) oxidation and Fe(III) reduction.

Distribution and Diversity of Gallionella-Like Neutrophilic Iron Oxidizers in a Tidal Freshwater Marsh†‡

ABSTRACT Microbial iron oxidation is an integral part of the iron redox cycle in wetlands. Nonetheless, relatively little is known about the composition and ecology of iron-oxidizing communities in the soils and sediments of wetlands. In this study, sediment cores were collected across a freshwater tidal marsh in order to characterize the iron-oxidizing bacteria (FeOB) and to link their distributions to the geochemical properties of the sediments. We applied recently designed 16S rRNA primers targeting Gallionella-related FeOB by using a nested PCR-denaturing gradient gel electrophoresis (DGGE) approach combined with a novel quantitative PCR (qPCR) assay. Gallionella-related FeOB were detected in most of the samples. The diversity and abundance of the putative FeOB were generally higher in the upper 5 to 12 cm of sediment than in deeper sediment and higher in samples collected in April than in those collected in July and October. Oxygen supply by macrofauna appears to be a major force in controlling the spatial and temporal variations in FeOB communities. The higher abundance of Gallionella-related FeOB in April coincided with elevated concentrations of extractable Fe(III) in the sediments. Despite this coincidence, the distributions of FeOB did not exhibit a simple relationship to the redox zonation inferred from the geochemical depth profiles.

Effect of sediment properties on the sorption of C12-2-LAS in marine and estuarine sediments

Linear alkylbenzene sulfonates (LAS) are anionic high production volume surfactants used in the manufacture of cleaning products. Here, we have studied the effect of the characteristics of marine and estuarine sediments on the sorption of LAS. Sorption experiments were performed with single sediment materials (pure clays and sea sand), with sediments treated to reduce their organic carbon content, and with field marine and estuarine sediments. C12-2-LAS was used as a model compound. Sorption to the clays montmorillonite and kaolinite resulted in non-linear isotherms very similar for both clays. When reducing the organic content, sorption coefficients decreased proportionally to the fraction removed in fine grain sediments but this was not the case for the sandy sediment. The correlation of the sediment characteristics with the sorption coefficients at different surfactant concentrations showed that at concentrations below 10 microg C12-2-LAS/L, the clay content correlated better with sorption, while the organic fraction became more significant at higher concentrations.

Effect of dissolved calcium on the removal of bacteriophage PRD1 during soil passage: The role of double-layer interactions

The objective of this work was to investigate and obtain quantitative relations for the effects of Ca(2+) concentration on virus removal in saturated soil and to compare the experimental findings with predictions of the DLVO theory. In order to do so, a systematic study was performed with a range of calcium concentrations corresponding to natural field conditions. Experiments were conducted in a 50-cm column with clean quartz sand under saturated conditions. Inflow solutions were prepared by adding CaCl(2,) NaCl and NaHCO(3) to de-ionized water. Values of pH and ionic strength were fixed at 7 and 10mM, respectively. Bacteriophage PRD1 was used as a conservative model virus for virus removal. The samples were assayed using the plaque forming technique. Attachment, detachment and inactivation rate coefficients were determined from fitting breakthrough curves. Attachment rate coefficients were found to increase with increasing calcium concentration. Results were used to calculate sticking efficiency, for which an empirical formula as a function of Ca(2+) was developed. Numerical solutions of the Poisson-Boltzmann equation were obtained to evaluate the effect of Ca(2+) on the double-layer interactions between quartz and PRD1. Based on these results, the DLVO interaction energies were calculated. It turned out that the experimental findings cannot be explained with the distance profiles of the DLVO interaction. The discrepancy between theory and experiment can be attributed to underestimation of the van der Waals interactions, chemisorption of Ca(2+) onto the surfaces, or by factors affecting the double-layer interactions, which are not included in the Poisson-Boltzmann equation. When abruptly changing from inflow solution containing Ca(2+) to a Ca(2+)-free solution, pronounced mobilization of viruses was observed. This indicates virus removal is not irreversible and that chemical perturbations of the groundwater can cause a burst of released viruses.

Uranium mobility in subsurface aqueous systems: the influence of redox conditions

Abstract Uranium is a redox-sensitive element and can be immobilized by reduction from soluble U(VI) to insoluble U(IV). By performing flow-through experiments, uranium mobility was observed under different redox conditions. Inflow solutions with different electron acceptors, nitrate and sulphate, and a control inflow solution were used to obtain different sedimentary redox conditions. Uranium was about one order more mobile when nitrate was used than when sulphate or the control was used. The difference in uranium mobility is attributed to the reduction of uranium. Even though uranium mobility is heavily dependent on the redox state of uranium, sedimentary concentrations of organic matter argue that organic matter is the most important complexing agent and that this determines the retardation of uranium.

Bacteriophage PRD1 batch experiments to study attachment, detachment and inactivation processes

Knowledge of virus removal in subsurface environments is pivotal for assessing the risk of viral contamination of water resources and developing appropriate protection measures. Columns packed with sand are frequently used to quantify attachment, detachment and inactivation rates of viruses. Since column transport experiments are very laborious, a common alternative is to perform batch experiments where usually one or two measurements are done assuming equilibrium is reached. It is also possible to perform kinetic batch experiments. In that case, however, it is necessary to monitor changes in the concentration with time. This means that kinetic batch experiments will be almost as laborious as column experiments. Moreover, attachment and detachment rate coefficients derived from batch experiments may differ from those determined using column experiments. The aim of this study was to determine the utility of kinetic batch experiments and investigate the effects of different designs of the batch experiments on estimated attachment, detachment and inactivation rate coefficients. The experiments involved various combinations of container size, sand-water ratio, and mixing method (i.e., rolling or tumbling by pivoting the tubes around their horizontal or vertical axes, respectively). Batch experiments were conducted with clean quartz sand, water at pH 7 and ionic strength of 20 mM, and using the bacteriophage PRD1 as a model virus. Values of attachment, detachment and inactivation rate coefficients were found by fitting an analytical solution of the kinetic model equations to the data. Attachment rate coefficients were found to be systematically higher under tumbling than under rolling conditions because of better mixing and more efficient contact of phages with the surfaces of the sand grains. In both mixing methods, more sand in the container yielded higher attachment rate coefficients. A linear increase in the detachment rate coefficient was observed with increased solid-water ratio using tumbling method. Given the differences in the attachment rate coefficients, and assuming the same sticking efficiencies since chemical conditions of the batch and column experiments were the same, our results show that collision efficiencies of batch experiments are not the same as those of column experiments. Upscaling of the attachment rate from batch to column experiments hence requires proper understanding of the mixing conditions. Because batch experiments, in which the kinetics are monitored, are as laborious as column experiments, there seems to be no major advantage in performing batch instead of column experiments.