Implicating the Glutathione-Gated Potassium Efflux System as a Cause of Electrophile-Induced Activated Sludge Deflocculation

Abstract

The glutathione-gated K⁺ efflux (GGKE) system represents a protective microbial stress response that is activated by electrophilic or thiol-reactive stressors. It was hypothesized that efflux of cytoplasmic K⁺ occurs in activated sludge communities in response to shock loads of industrially relevant electrophilic chemicals and results in significant deflocculation. Novosphingobium capsulatum, a bacterium consistent with others found in activated sludge treatment systems, responded to electrophilic thiol reactants with rapid efflux of up to 80% of its cytoplasmic K⁺ pool. Furthermore, N. capsulatum and activated sludge cultures exhibited dynamic efflux-uptake-efflux responses very similar to those observed by others in Escherichia coli K-12 exposed to the electrophilic stressors N-ethylmaleimide and 1-chloro-2,4-dinitrobenzene and the reducing agent dithiothreitol. Fluorescent LIVE/DEAD stains were used to show that cell lysis was not the cause of electrophile-induced K⁺ efflux. Nigericin was used to artificially stimulate K⁺ efflux from N. capsulatum and activated sludge cultures as a comparison to electrophile-induced K⁺ efflux and showed that cytoplasmic K⁺ efflux by both means corresponded with activated sludge deflocculation. These results parallel those of previous studies with pure cultures in which GGKE was shown to cause cytoplasmic K⁺ efflux and implicate the GGKE system as a probable causal mechanism for electrophile-induced, activated sludge deflocculation. Calculations support the notion that shock loads of electrophilic chemicals result in very high K⁺ concentrations within the activated sludge floc structure, and these K⁺ levels are comparable to that which caused deflocculation by external (nonphysiological) KCl addition.

Chemically induced upset of full-scale biological treatment systems occurs at an unacceptably high rate in the United States (18). Unfortunately, few studies have been conducted under well-controlled experimental conditions with indigenous microbial communities to show how certain wastewater conditions (sources) impact the performance of activated sludge treatment processes (effects). A range of microscopic- and molecular-level causal mechanisms exist that allow cells to adapt when exposed to chemical perturbations (26). The focus of our work is on rapid stress responses that are activated immediately or within a few seconds of perturbation events. The degree to which these rapid molecular-level mechanisms are involved with causing adverse process upsets in full-scale treatment systems is unknown, but we contend that they play an important role in defining how complex microbial communities respond to chemical perturbations. Understanding source-cause-effect relationships could lead to the development of upset early-warning technologies and mitigation strategies that will help to avoid significant environmental problems that are often associated with activated sludge process upset. This study represents an effort to further implicate a specific stress mechanism, which is activated by shock loads of toxic electrophilic chemicals, as a cause of deflocculation in activated sludge biological treatment systems.

The protective microbial stress response mechanism that is evaluated here and is believed to play a role in activated sludge system performance is the glutathione-gated K⁺ efflux (GGKE) system (5). GGKE is activated when thiol-reactive electrophilic chemicals are sequestered by cysteine-containing glutathione, thereby forming a glutathione-S conjugate or oxidized glutathione. It was determined that the Escherichia coli GGKE system includes two independent membrane-bound potassium efflux proteins (KefB and KefC) that are gated by reduced glutathione and activated by glutathione-S conjugates and possibly oxidized glutathione (5). A reaction between glutathione and the electrophilic toxin activates K⁺ efflux from the cytoplasm. GGKE was originally studied primarily in E. coli K-12 by Booth and coworkers (5). However, it has also been found to occur in a wide range of other gram-negative heterotrophic bacteria (9, 12). Ferguson et al. (11-13) determined that a chemical challenge by toxic electrophilic compounds (both mild thiol reactants and strong oxidants) elicits a protective mechanism in E. coli K-12 that prevents macromolecular damage by rapid efflux of intracellular K⁺ and concurrent cytoplasmic acidification.

In early work on the GGKE system, N-ethylmaleimide (NEM)-induced efflux of intracellular K⁺ was found to be reversible in E. coli K-12. Meury et al. (19) and Bakker and Mangerich (3) determined that adding a membrane-permeable thiol reductant, such as β-mercaptoethanol or dithiothreitol (DTT), after NEM-induced K⁺ efflux resulted in recovery of cytoplasmic K⁺, albeit at a slower rate than the observed efflux. Adding excess NEM after K⁺ recovery reactivated the GGKE system and caused efflux of the previously recovered K⁺ (efflux-uptake-efflux) (3, 19). Additionally, it was determined that both the efflux and recovery of intracellular K⁺ were mediated without de novo protein synthesis; however, uptake of K⁺ after an efflux event required a suitable exogenous energy source (10, 19, 20). The C-S-C bond of N-ethylsuccinimido-S-glutathione (glutathione-NEM conjugate) is expected to be chemically stable, even in the presence of excess DTT. However, Elmore et al. (10) found an 80% decrease in the cytoplasmic N-ethylsuccinimido-S-glutathione concentration after addition of DTT and suggested that removal of the GGKE activator was responsible for K⁺ uptake. Beutler et al. (4) found that the C-S-C bond of N-ethylsuccinimido-S-glutathione is stable under acidic and basic conditions but may undergo slow nonenzymatic hydrolysis at physiological pH and in the absence of a reductant such as DTT. Tötemeyer et al. (27) demonstrated that mid-log- and stationary-phase cultures of Pseudomonas putida could take up and detoxify NEM at low concentrations (up to 30 μM) within 5 h by removing it from the cell, but mineralization was not demonstrated.

Unlike NEM-induced K⁺ efflux, 1-chloro-2,4-dinitrobenzene (CDNB)-induced K⁺ efflux was not reversible in E. coli K-12 by addition of either β-mercaptoethanol or DTT (5). The aromatic thioether in glutathione-S-dinitrobenzene (glutathione-CDNB conjugate) is very stable and is not labile in the presence of excess reducing thiols. These differences in the stability of the C-S-C thioether bond may explain the differences in reversibility between NEM-induced and CDNB-induced K⁺ efflux (5). Regardless of the mechanism leading to recovery of cytoplasmic K⁺, it has been documented in E. coli K-12 that NEM-induced K⁺ efflux can be reversed by the addition of DTT and CDNB-induced K⁺ efflux cannot be reversed. If a similar pattern is observed in activated sludge, it will implicate the GGKE mechanism as being active and significant.

It is hypothesized that efflux of cytoplasmic K⁺ causes activated sludge deflocculation (effect) due to shock loads of toxic electrophilic chemicals (source). Furthermore, it is proposed that the GGKE stress response is the mechanism most likely to be responsible for regulating electrophile-activated cytoplasmic K⁺ efflux from activated sludge. The occurrence of rapid K⁺ efflux from activated sludge bacterial flocs that were exposed to sublethal perturbations of electrophilic chemicals was recently reported (6). Significant activated sludge deflocculation was observed, and the degree of deflocculation correlated with the degree of K⁺ efflux from the floc or biomass material to the bulk liquid. It is proposed that the sudden increase in monovalent cation levels in the extracellular polymeric substances (EPS) that make up a significant fraction of the flocs results in a rapid increase in the localized monovalent-to-divalent cation (M/D) ratio. High M/D ratios (>2 on an equivalence basis) have previously been shown to correlate with reduced floc strength and turbid effluents from laboratory and full-scale activated sludge systems (14, 15, 21, 22). Similarly, activated sludge floc exposure to electrophilic toxins and subsequent K⁺ efflux would cause an increase in the intrafloc M/D ratio, resulting in deflocculation.

In this work, batch experiments were conducted to determine whether GGKE-activated K⁺ efflux and influx patterns demonstrated previously in E. coli K-12 with the electrophiles NEM and CDNB could be repeated with N. capsulatum (formerly Sphingomonas capsulata) and activated sludge cultures obtained from a full-scale domestic wastewater treatment plant. By using cloning and 16S rRNA gene sequence analysis, Snaidr et al. (25) determined that N. capsulatum was present in a domestic activated sludge culture at approximately 3% of the total bacterial cell count, which classifies that microorganism as being prevalent in relation to the broadly diverse populations typically found in activated sludge cultures. Using N. capsulatum also provided a model for conducting experiments in the absence of flocs, thereby eliminating complications associated with electrophile or K⁺ diffusion into and out of the floc structure. DTT was used as the reductant to test the reversibility of the efflux process and further implicate GGKE. The antibiotic nigericin was used as a non-GGKE K⁺ efflux activator to compare with NEM and as a positive control for demonstrating the impact of cytoplasmic K⁺ efflux on deflocculation. The degree to which cell lysis contributed to K⁺ efflux was also investigated by using membrane-permeable and -impermeable fluorescent nucleic acid stains. Finally, the K⁺ concentration within the floc structure immediately following electrophile-induced K⁺ efflux was estimated and compared to previous reports that indicate the amount of exogenously added K⁺ needed to cause activated sludge deflocculation.

MATERIALS AND METHODS

Cultures and growth conditions.

N. capsulatum (ATCC 14666) was obtained from the American Type Culture Collection, and Luria-Bertani broth (ATCC medium 1065) was used as a maintenance growth medium. For all of the experiments presented here, N. capsulatum was grown to late log or early stationary phase (approximately 3 × 10⁹ to 8 × 10⁹ CFU/ml) in a semidefined medium (Glc/Cas) containing glucose and Casamino Acids (12.8 mM glucose, 0.5 g of Casamino Acids per liter, 40.1 mM Na₂HPO₄, 11.7 mM NaH₂PO₄, 1.0 mM KCl, 15.5 mM NH₄Cl, 0.72 mM MgSO₄, 4.7 μM MnSO₄ · H₂O, 0.12 μM Na₂MoO₄ · 2H₂O, 0.69 mM CaCl₂ · 2H₂O, 92 μM FeCl₃, 0.021 μM CoCl₂ · 6H₂O, 7.3 μM ZnCl₂, 0.76 μM CuCl₂ · 2H₂O, 1.3 μM H₃BO₃). N. capsulatum cell counts were performed by serially diluting samples in a medium containing 145 mM NaCl, 2.2 mM KH₂PO₄, 4.2 mM Na₂HPO₄, and 0.5 g of nutrient broth (adjusted to pH 7.3 with NaOH) per liter and plating on Luria-Bertani agar. Activated sludge experiments were initiated with fresh mixed liquor from the Blacksburg-Virginia Tech Wastewater Treatment Plant (Blacksburg, Va.). Fresh primary effluent was obtained at the same time as mixed liquor and used as the feed for all of the activated sludge experiments.

N. capsulatum experiments.

N. capsulatum was grown in batch cultures as described above. Immediately before each batch experiment was initiated, cells were harvested by centrifugation at 9,500 × g for 10 min and resuspended in sufficient Glc/Cas medium to concentrate the cells 3.33-fold. The concentrated culture (90 ml) was added to 150-ml beakers (in duplicate) and mixed vigorously with magnetic stir bars. The first potassium sample was collected within 5 to 10 min after resuspension in the Glc/Cas medium, and this was defined as time zero. For the experiments that also used ion-selective electrodes (ISEs) for K⁺ monitoring, the probes were inserted and allowed 5 to 10 min to stabilize prior to collection of the first liquid aliquot for K⁺ analysis and initiation of data acquisition.

CDNB was predissolved in 500 μl of acetone and added at 0.50 mM in a single spiked addition. The same volume of acetone was also added to the control. DTT reacts with CDNB and NEM on a 1:1 molar basis. For the CDNB experiment, DTT was added at 1.5 mM directly from a preweighed quantity of dry powder approximately 30 min after CDNB was added (three times the stoichiometric amount needed to react with the previously added CDNB) and 3.0 mM CDNB was added approximately 30 min later in 1,200 μl of acetone (two times the stoichiometric amount needed to react with the previously added DTT). Samples were collected throughout the experiment for potassium analysis. NEM was added from a concentrated stock solution in distilled, deionized H₂O at 0.4 mM in a single spiked addition. As with the CDNB experiment, DTT was added at 1.2 mM approximately 30 min following the initial NEM dose and 3.6 mM NEM was added approximately 30 min later. Nigericin was added at a range of concentrations (1 to 100 μM) from a 33.5 mM stock solution in ethanol.

Activated sludge experiments.

Activated sludge sequencing batch reactor (SBR) experiments were conducted with either 0.25- or 4-liter beakers. The small reactors had a working volume of 0.20 liter and were used both to simplify operation of K⁺ ISEs and so that the required amounts of nigericin and DTT could be minimized because of the expense. Both the large and small SBR systems were operated for only one 6-h cycle since the focus of this study was short-term adaptive responses. Aeration of both systems was done with aquarium air pumps connected to air stones. Mixing was provided by magnetic stir bars for the small reactors and 100-rpm electric motors with single-blade paddles for the large reactors. Experiments were started by adding fresh mixed liquor (equal to the working volume) to the beakers, stirring and aerating it for approximately 10 min, and then allowing it to settle for 30 min. One-fourth of the total reactor working volume was decanted as supernatant (the biomass settled to below this volume), and aerators and mixers were restarted. Fresh primary effluent was added to replace the decanted volume over 3 min. The first potassium sample was collected within 5 to 10 min after addition of the primary effluent, and this was defined as time zero.

For chemically stressed SBRs, CDNB or NEM was added at various concentrations (0.50 to 3.0 mM for CDNB, 0.40 to 7.2 mM for NEM) either immediately after the time zero sample or within the first 60 to 90 min of the reaction cycle via a single spiked addition. An unshocked control was operated in parallel for all experiments. The lower concentration of each stressor was previously determined to be that required to reduce the oxygen uptake (respiration) rate of the activated-sludge culture by 50%. A concentrated NEM stock solution was prepared in distilled, deionized H₂O, and the CDNB stock solution was made in acetone (the same volume of acetone as that added to the corresponding control SBR). For experiments in which DTT was used, it was added directly from a preweighed quantity of dry powder in a single spike addition approximately 30 to 45 min after addition of CDNB or NEM. DTT was added by using a molar concentration three times the stoichiometric amount needed to react with all of the NEM or CDNB. Subsequent CDNB or NEM addition occurred 30 to 45 min later with a molar concentration equivalent to two or three times the stoichiometric amount needed to react with all of the DTT added. Nigericin experiments were conducted by adding various amounts of a 33.5 mM stock solution in ethanol, and an equivalent volume of ethanol associated with the highest nigericin dose was added to the control reactor.

It was determined that most of the soluble organic matter added with the primary effluent was consumed in the SBRs within about 30 min (data not shown). Therefore, it was necessary to amend the primary effluent feed with a readily degradable energy source so that an exogenous substrate was present during DTT-activated K⁺ recovery periods. Primary effluent was amended with additional organic carbon by adding 400 mg of Bacto Peptone per liter, 510 mg of Na-acetate per liter, and 375 mg of glucose per liter directly to the fresh primary effluent.

Analytical procedures.

Soluble and cell- or floc-associated cation samples were collected for both N. capsulatum and activated sludge experiments and, depending on the experiment, analyzed by either ion chromatography (IC), atomic absorption spectrophotometry (AAS), or inductively coupled plasma emission spectrometry (ICP-ES) as described previously (6). Briefly, samples were prepared by centrifugation with an inert, insoluble silicone fluid that is slightly more dense than water but less dense than the biomass material. The silicone fluid rapidly separates the liquid and biomass phases, with the soluble phase isolated above and the cell- or floc-associated material isolated below the silicone fluid. The liquid phase was aspirated off for measurement of the soluble or bulk liquid cation content. After removal of the silicone fluid, the biomass pellet was acid-heat digested, filtered, and analyzed for cations. For the experiments described here, soluble potassium is defined as a bulk liquid K⁺ measurement, and cell or floc association is defined as the concentration of potassium associated with the biomass (either pure culture or activated sludge) normalized to the original sample volume.

For experiments that assessed the degree of deflocculation, activated sludge settling performance was evaluated by monitoring effluent total suspended solids and effluent volatile suspended solids (VSS) in accordance with Standard Methods for the Examination of Water and Wastewater, 20th edition (8). The 200-ml SBRs did not provide enough settled decant to perform analysis of effluent suspended solids; therefore, nephelometric turbidity measurements were used as a surrogate to quantify the degree of deflocculation.

In several representative experiments, K⁺ ISEs were used to monitor changes in the soluble bulk liquid K⁺ concentration in real time by computer data acquisition. The system consisted of two Orion 9719BN (Orion, Beverly, Mass.) combination K⁺ ISEs, two Accumet (Fisher Scientific, Atlanta, Ga.) automatic temperature compensation probes, and a two-channel Accumet AR25A pH-millivolt-ISE meter with serial-port data acquisition from the meter with a laptop computer running LabVIEW 6.0 software (National Instruments, Austin, Tex.). A data acquisition program was written to record K⁺, temperature, and millivolt data in a spreadsheet file every 6 s for the duration of each experiment. Smoothing of the ISE data was performed by a five-point moving average, and manual temperature corrections were made where appropriate. The ISEs were calibrated by using 0.026, 0.26, 1.28, and 2.56 mM KCl-K⁺ standards that contained NaCl to adjust the ionic strength of the standards to appropriate levels.

Bulk liquid NEM concentrations were measured with a Hewlett Packard 1090 high-pressure liquid chromatograph (HPLC) with a UV-visible light diode array detector and an Alltech Econosphere C₁₈ analytical column and a guard column. NEM samples were processed by filtering activated sludge or N. capsulatum supernatant with 0.2-μm-pore-size Gelman (Ann Arbor, Mich.) SUPOR-200 membrane filters and acidified with 50% (vol/vol) H₂SO₄ to approximately pH 2. Samples were stored in the dark at 4°C until analyzed. The HPLC eluent consisted of 90% (by volume) 13 mM H₃PO₄ in 18 MΩ H₂O and 10% (by volume) HPLC grade acetonitrile at a flow rate of 1.5 ml/min. Analysis was performed with an injection volume of 50 μl. NEM standards were diluted from SigmaUltra (Sigma-Aldrich, St. Louis, Mo.) crystals in 13 mM H₃PO₄ in 18 MΩ H₂O to concentrations ranging from 8 to 800 μM. Data acquisition and peak integration were performed with HP ChemStation software.

In order to evaluate cell lysis in both N. capsulatum and activated sludge cultures exposed to both NEM and nigericin, the LIVE/DEAD BacLight bacterial viability system (Molecular Probes, Inc., Eugene, Oreg.) was used to assess membrane integrity. For activated sludge, the stains were used as recommended by Molecular Probes, Inc., and as previously documented by Ramirez et al. (24). Syto 9 (green stain, indicates viable cells) and propidium iodide (red stain, indicates membrane-compromised cells) were added directly from stocks premade in dimethyl sulfoxide to mixed-liquor samples providing final concentrations of 5 and 30 μM, respectively. Staining was performed in the dark for 15 min before observation with a Zeiss Axiovert S1D0TV epifluorescence microscope (Carl Zeiss, Inc., Thornwood, N.Y.) at a magnification of ×400 or ×630 with a 500 ± 10-nm excitation filter, a 515-nm long-pass dichroic mirror, and a 520-nm emission filter. This fluorescence filter set allowed simultaneous observation of green and red fluorescence. For N. capsulatum samples, the cells were washed and resuspended in a bicarbonate buffer solution containing 1 mM KCl, 50 mM NaCl, and 50 mM NaHCO₃ adjusted to pH 7.0 with HCl. For best differentiation between green and red cells, the stain concentrations were adjusted to 0.5 and 30 μM for Syto 9 and propidium iodide, respectively. For both activated sludge and N. capsulatum, positive controls for lysed cells were obtained by adding 5% (vol/vol) isopropanol and incubating the mixture for 1 h with periodic mixing. The isopropanol was removed by washing with bicarbonate buffer solution before staining.

For activated sludge stained with LIVE/DEAD, it was possible to observe the live and dead regions within an individual floc. By phase-contrast microscopy in association with the LIVE/DEAD staining system, it was determined that only propidium iodide stained the extracellular polymeric matrix of flocs, thereby providing a means to determine the total viable (green) versus the nonviable (red) floc volume. The average volume of the nonviable extracellular fraction of activated sludge flocs (EPS and dead cell volume) was needed to calculate the K⁺ concentration in the floc structure following electrophile addition. By epifluorescence microscopy with filter sets that allowed either green or red fluorescence to be observed independently, it was possible to quantify the percentage of the area associated with green fluorescence relative to the percentage of the area associated with red fluorescence in individual flocs. This was accomplished with filter sets that contained a 480 ± 20-nm excitation filter, a 505-nm long-pass dichroic mirror, and a 535 ± 25-nm emission filter (green dye) and a 545 ± 15-nm excitation filter, a 570-nm long-pass dichroic mirror, and a 610 ± 37-nm emission filter (red dye). Black-and-white images of a series of activated sludge flocs observed under green- or red-only fluorescence were obtained with a computer-controlled charge-coupled device. The area of green or red fluorescence was quantified with PC-based Scion Image release beta 3b (Scion, Frederick, Md.). The red and green integrated densities were then summed to obtain the percentage of green or red volume for an individual floc. Eukaryotic organisms within the activated sludge matrix did not appear to stain either red or green and were only visible by standard phase-contrast microscopy.

RESULTS

N. capsulatum dynamic efflux-uptake experiments.

As shown in Fig. 1, N. capsulatum demonstrated significant and rapid K⁺ efflux in response to 0.4 mM NEM. The soluble-K⁺ concentration increased by approximately 0.30 mM within 5 min of NEM addition and approximately 1.0 to 1.1 mM over the course of the experiment. The magnitude of the NEM-mediated increase in soluble K⁺ shown in Fig. 1 was consistent with several independent experiments conducted under similar conditions (data not shown). The increase in the soluble-K⁺ concentration in the NEM-stressed culture occurred concurrently with a comparable decrease in the cell-associated phase (Fig. 1A), thereby confirming transport of K⁺ from the cells to the bulk liquid. At the same time, soluble and cell-associated potassium levels remained constant in the unstressed control. In all cases, electrophilic stress did not cause significant changes in either bulk liquid or cell-associated concentrations of other cations, including Fe²⁺, Fe³⁺, Al³⁺, Ca²⁺, Mg²⁺, and Na⁺ (data not shown).

FIG. 1.

(A) N. capsulatum dynamic NEM stress experiment with soluble-K⁺ and cell-associated (CA) K⁺ measurements. N. capsulatum was concentrated to 2.88 × 10¹⁰ ± 0.68 × 10¹⁰ CFU/ml. NEM was added at 0.40 mM to the stressed reactor, followed by DTT addition at 1.2 mM to both the NEM-stressed and control reactors and a final addition of 3.6 mM NEM to both reactors. Datum points represent the average and error bars represent the range of duplicate samples. (B) ISE soluble-K⁺ measurements versus time.

Addition of DTT caused significant and rapid K⁺ uptake to occur (Fig. 1B). The unstressed control showed a negligible amount of K⁺ uptake (approximately 0.03 mM) immediately after DTT was added but remained constant thereafter. The cell-associated data showed that potassium levels increased within the cells in the NEM-stressed culture but did not change significantly in the unstressed control culture (Fig. 1A). Almost 30 min after DTT addition, NEM addition to both the stressed (NEM-DTT-NEM) and previously unstressed control (control-DTT-NEM) reactors resulted in immediate efflux of K⁺ from both cultures, and the final soluble-K⁺ and cell-associated K⁺ levels were the same in both.

ISE data, shown in Fig. 1B, confirm the trends that were observed through grab sample measurements of the soluble fraction (Fig. 1A). In addition, the ISE data show that efflux did not occur until after NEM was added, and uptake did not occur to a significant degree until after DTT was added. Furthermore, the ISE data show that K⁺ efflux was immediate and rapid after addition of NEM, whereas uptake occurred at a slower rate. The data in Fig. 1B show that changes did not occur until after chemicals were added and clarify the trends that are shown in Fig. 1A by grab sample data.

For several representative N. capsulatum and activated sludge experiments, K⁺ ISE data are presented to provide an indication of the very rapid rate of K⁺ efflux and to corroborate the other potassium analyses. In most cases, the ISE data demonstrated the expected trends, but the electrode response tended to be erratic and drifted in a manner that was not consistent with K⁺ analysis by IC, ICP-ES, or AAS and not associated with temperature changes. Furthermore, some of the chemicals added during the batch experiments seemed to interfere with the function of the electrode polymer membrane. Although the ISEs revealed trends similar to those shown by other bulk liquid K⁺ concentration measurements, the ISE and IC measurements were typically lower than bulk liquid soluble-K⁺ analyses by ICP-ES and AAS. As expected, nigericin caused dramatic electrode interference, and these data are not reported. The current state of K⁺ ISE technology may not be appropriate for real-time evaluation of the bacterial K⁺ efflux response in complex environmental matrices, in the presence of electrophilic organic compounds, or in full-scale wastewater treatment systems. Nevertheless, selected ISE data are shown to clarify the real-time K⁺ efflux and uptake trends observed in the bulk liquid, which would have been impossible to show with grab samples alone.

As shown in Fig. 2, addition of 0.5 mM CDNB to N. capsulatum resulted in an increase in the bulk liquid soluble-K⁺ concentration of 1.1 mM over the course of the experiment. The increase in the soluble-K⁺ concentration in the CDNB-stressed culture occurred concurrently with a comparable decrease in the cell-associated phase, thereby confirming transport of K⁺ from the cells to the bulk liquid. At the same time, soluble-potassium and cell-associated potassium levels remained constant in the unstressed control. In contrast to results observed with NEM, addition of DTT caused no apparent reversal of the K⁺ efflux response. Approximately 40 min after DTT addition, CDNB was added to both the stressed (CDNB-DTT-CDNB) and previously unstressed control (control-DTT-CDNB) reactors. There was no additional increase in the bulk liquid K⁺ concentration in the stressed reactor, and the control released K⁺ from the cell-associated phase into the bulk liquid to a concentration comparable to that observed in the stressed reactor approximately 40 min after CDNB addition.

FIG. 2.

N. capsulatum dynamic CDNB stress experiment with soluble-K⁺ and cell-associated (CA) K⁺ measurements. N. capsulatum was concentrated to 1.10 × 10¹⁰ ± 0.15 × 10¹⁰ CFU/ml. CDNB was added to the stressed reactor at 0.50 mM, followed by DTT addition to both the CDNB-stressed and control reactors at 1.5 mM and a final addition of 3.0 mM CDNB to both reactors. Datum points represent the average and error bars represent the range of duplicate samples.

The results of the dynamic efflux-uptake experiment suggest that CDNB-induced efflux is irreversible and NEM-induced efflux is reversible in N. capsulatum. This gram-negative organism exhibits cytoplasmic K⁺ activity that parallels that previously observed for E. coli K-12 and for which GGKE was shown to be the causal mechanism (2, 10, 19).

Activated sludge dynamic efflux-uptake experiments.

K⁺ efflux was also observed in the activated sludge culture exposed to 0.4 mM NEM (Fig. 3), with approximately 0.1 mM K⁺ released. However, the rate of the K⁺ efflux response was considerably slower in activated sludge than in N. capsulatum. ISE data indicated that maximal N. capsulatum K⁺ efflux to the bulk liquid typically occurred within 5 to 10 min (Fig. 1B), whereas maximal activated sludge culture efflux to the bulk liquid required at least 30 min (Fig. 3B). It is suspected that this difference in efflux rate is due to diffusion limitations in activated sludge flocs, both for the stressor diffusing into the interior of the floc and for residual K⁺ diffusing from the floc to the bulk liquid. Data confirming K⁺ transport from the floc-associated phase to the bulk liquid in response to NEM in activated sludge have been reported elsewhere (6). Addition of DTT resulted in slow but noticeable uptake of K⁺ from the bulk liquid (approximately 0.02 mM K⁺). Again, the rate of K⁺ uptake was slower for the flocculant activated sludge culture than for the N. capsulatum culture. Addition of NEM to both the stressed and control cultures resulted in comparable final concentrations of K⁺, suggesting that there was a maximum level of K⁺ efflux that could be achieved in the activated sludge culture when it was exposed to NEM. These results are consistent with the results shown here for N. capsulatum and elsewhere for E. coli (3, 10, 19).

FIG. 3.

(A) Activated sludge dynamic NEM stress experiment conducted with SBRs operated in parallel with soluble-K⁺ (bulk liquid) measurements. NEM was added to the stressed reactor at 0.40 mM, followed by DTT addition to both the NEM-stressed and control reactors at 1.2 mM and a final addition of 3.6 mM NEM to both reactors. Datum points represent the average and error bars represent the range of duplicate samples. (B) ISE soluble-K⁺ measurements versus time.

The pattern of the dynamic response of activated sludge cultures to CDNB shock is different from that of the response to NEM shock, which is consistent with previous findings for E. coli (10). As shown in Fig. 4, CDNB-induced K⁺ efflux was not reversed by DTT addition at a molar concentration three times the level of CDNB added. Subsequent addition of 3 mM CDNB did not increase the rate or extent of K⁺ efflux in the experimental reactor but did cause a comparable level of K⁺ efflux from the control. This result shows that CDNB-initiated K⁺ efflux cannot be reversed in activated sludge cultures by the strong reductant DTT, whereas NEM-initiated K⁺ efflux can. This observation is consistent with results reported here for N. capsulatum and for E. coli by others (5).

FIG. 4.

Activated sludge dynamic CDNB stress experiment conducted with SBRs operated in parallel with soluble-K⁺ (bulk liquid) measurements. CDNB was added to the stressed reactor at 0.50 mM, followed by DTT addition to both the CDNB-stressed and control reactors at 1.5 mM and a final addition of 3.0 mM CDNB to both reactors. Datum points represent the average and error bars represent the range of duplicate samples.

Results from previous SBR experiments, in which the system was operated for several days after chemical shock loading, indicated that one NEM shock event was sufficient to weaken the floc structure for an extended period of time, even after the stressor had washed out of the system and the bulk liquid K⁺ concentration returned to control levels (6). Additional experiments were conducted with activated sludge to determine if addition of DTT could reverse a deflocculation event. As shown in Fig. 5, the effluent VSS concentrations were significantly higher in the SBRs receiving NEM than in the unstressed control. The damage within the floc structure caused by the original NEM-induced K⁺ efflux was not remedied by the slight uptake of K⁺ that occurred after DTT addition. NEM levels were measured whenever K⁺ samples were collected during this experiment; these data indicate that NEM concentrations were at target levels at time zero and that limited biodegradation occurred over the time of the experiment (Table 1). We conclude that the initial efflux response, which is observed immediately after electrophile addition, is responsible for floc deterioration and that uptake of K⁺ does not reverse this effect.

FIG. 5.

Effect of dynamic activated-sludge efflux on deflocculation as shown by soluble-K⁺ (bulk liquid) and effluent VSS measurements. This experiment was conducted with four 3.5-liter SBRs operated in parallel. One SBR served as the control, one received only 0.80 mM NEM (NEM), one received 0.80 mM NEM followed by 2.4 mM DTT (NEM/DTT), and one received 0.80 mM NEM, 2.4 mM DTT, and 4.8 mM NEM (NEM/DTT/NEM). Effluent VSS was measured at the end of the 5.25-h reaction period following 30 min of quiescent settling. Datum points represent the average and error bars represent the range of duplicate samples.

TABLE 1.

Measured and predicted NEM concentrations for the dynamic activated sludge efflux-uptake-efflux experiment conducted with 3.5-liter SBRs^a

Sample time (min)	Soluble NEM concn (mg/liter)
	Control		NEM-DTT-NEM		NEM-DTT		NEM
	Measured	Expected	Measured	Expected	Measured	Expected	Measured	Expected
0	NPDb	0	NPD	0	NPD	0	NPD	0
28	NPD	0	99.3 ± 0.21	100	99.7 ± 0.12	100	101 ± 0.11	100
49	NPD	0	NPD	0	NPD	0	97.3 ± 0.08	100
105	NPD	0	NPD	0	NPD	0	90.9 ± 1.8	100
127	NPD	0	390 ± 7.0	400	NPD	0	88.9 ± 0.05	100
300	NPD	0	397 ± 0.31	400	NPD	0	73.0 ± 4.2	100

^aNEM concentrations were measured by HPLC in the mixed-liquor supernatant. Expected NEM concentrations are reported assuming complete and rapid equimolar reaction between NEM and DTT and no other sorptive or reactive losses of NEM. The reported values represent the average of duplicate analyses ± 1 standard deviation.

^bNPD, no NEM peak detected.

K⁺ efflux due to cell lysis.

Several observations suggest that cell lysis was not responsible for electrophile-induced K⁺ efflux. We previously showed that maximal K⁺ efflux and deflocculation occurred in activated sludge cultures at electrophile concentrations significantly less than that required to reduce the oxygen uptake rate by 50% (6). We also showed that electrophilic stress to activated sludge cultures resulted in significant K⁺ efflux but no significant efflux of other cations, including Fe²⁺, Fe³⁺, Al³⁺, Ca²⁺, Mg²⁺, and Na⁺. Electrophile addition, even at a very high concentration (NEM at 16 mM or CDNB at 1 mM) caused negligible release of alkaline phosphatase, a strictly periplasmic enzyme in bacteria, from activated sludge cultures compared to a measurement of alkaline phosphatase activities in an activated sludge cell extract (6). In this work, the LIVE/DEAD BacLight system was used to assess membrane integrity after addition of NEM to N. capsulatum and activated-sludge cultures. Epifluorescence microscopic observations revealed that NEM stress did not yield more red cells than those observed in controls for N. capsulatum (data not shown). For activated sludge, multiple flocs were observed from both control and NEM-stressed mixed liquor. As expected, some red zones were observed within individual flocs for both stressed and control mixed liquors, but most of the floc area stained green. After NEM stress at 0.4 mM for 60 min, there was no discernible increase in the floc area that stained red relative to the control (data not shown). These results indicate that the K⁺ efflux observed is not due to significant cell death and lysis in response to the chemical stressor.

Effect of nigericin on K⁺ efflux.

Previously, researchers demonstrated that external addition of K⁺ (as KCl) to activated sludge at concentrations of 10 to 23 mM K⁺ caused release of Ca²⁺ and/or Mg²⁺ from the floc structure, presumably as a result of ion-exchange mechanisms. The divalent cation release was followed by a rapid deterioration of settling and dewatering properties (7, 22). The level of K⁺ added externally as KCl (10 to 23 mM) was significantly higher than the maximal increase in the bulk liquid K⁺ due to NEM or CDNB shock loading (typically, 0.15 to 0.25 mM, depending on the concentration of mixed-liquor suspended solids [MLSS]). Considering the effects of excess monovalent cation on floc strength, the excretion of K⁺ into the floc structure from bacterial cytoplasms due to electrophilic stress is likely a very different phenomenon than addition of K⁺ externally. As expected, external addition of KCl to an SBR, at a level comparable to that measured in the bulk phase at maximal electrophile-induced K⁺ efflux (0.26 mM KCl), caused no change in effluent VSS or settling properties relative to those of a control reactor (6). However, we previously hypothesized that electrophile-induced cytoplasmic K⁺ efflux activity results in very high localized concentrations of K⁺ within the floc structure immediately after electrophilic stress (6). Furthermore, it is believed that a fraction of the high levels of K⁺ (located outside of cells but still within the floc structure) diffuses slowly to the bulk liquid.

In order to evaluate the effect of K⁺ efflux on floc integrity, the antibiotic nigericin was used to induce K⁺ release independently of the KefBC-like GGKE system and without the addition of electrophilic toxins. The purpose of the nigericin experiments was to evaluate the K⁺ efflux response in both activated sludge and pure cultures compared to the response to electrophilic stress. Second, these experiments were intended to provide an indication that the increase in bulk liquid K⁺ in activated sludge was derived from the cytoplasm (i.e., transport of K⁺ from the cytoplasm of activated sludge microorganisms to the floc structure and then to the bulk liquid). Finally, if the hypothesis that the deflocculation response is caused by cytoplasmic K⁺ efflux is correct, a deflocculation response should occur in response to nigericin-induced efflux.

Nigericin is a cyclic polyether ionophore that is closed by a carboxylate group hydrogen bonded to a hydroxyl substituent (16). This molecule is absorbed into the cytoplasmic membrane and very selectively transports K⁺ down a concentration gradient. Nigericin actually serves as an artificial K⁺/H⁺ antiporter and thus does not require a chemical uncoupler for rapid and significant cytoplasmic K⁺ efflux (16). In bacteria, nigericin eliminates both the catabolic ΔpH gradient and the significant outwardly directed K⁺ concentration gradient by essentially making the membrane permeable to only H⁺ and K⁺. The benefit of using this antibiotic is to simulate the action of cytoplasmic K⁺ efflux without adding an electrophilic stressor. Nigericin has been used in both eukaryotic (17) and prokaryotic transport experiments to equilibrate the intracellular K⁺ level with that in the bulk liquid (1, 23).

Nigericin addition (50 μM) to N. capsulatum resulted in significant K⁺ efflux that was comparable to the efflux derived when 0.80 mM NEM was added (Fig. 6). Again, measurable K⁺ efflux occurred immediately after chemical addition. More importantly, DTT caused recovery from NEM-induced K⁺ efflux but not from nigericin-promoted K⁺ release. This was as expected since nigericin-induced efflux should not depend on the redox state of the cytoplasm or the presence of excess reductant. Following nigericin addition, subsequent NEM addition (7.2 mM) resulted in negligible K⁺ release, suggesting that the maximum K⁺ release possible for that level of nigericin addition had already been achieved.

FIG. 6.

N. capsulatum dynamic stress experiment at a cell density of (1.8 ± 1.1) × 10¹⁰ CFU/ml with either nigericin (50 μM) or NEM (0.80 mM) addition. After the initial addition of nigericin or NEM, both systems received DTT at a concentration of 2.4 mM and a subsequent addition of NEM at 7.2 mM. Soluble-K⁺ (bulk liquid) concentrations are plotted versus time. Datum points represent the average and error bars represent the range of duplicate samples.

Nigericin caused rapid and significant K⁺ efflux from activated sludge to the bulk liquid at a seemingly faster rate than 0.4 mM NEM addition (Fig. 7). Increasing concentrations of nigericin caused enhanced K⁺ release into the bulk liquid. On the basis of the average floc-associated K⁺ concentration observed in the activated sludge culture used during this study (0.13 mmol of K⁺/g of mixed-liquor VSS), the degree of efflux obtained for 100 μM nigericin approximately represents the point of complete K⁺ release from the floc structure to the bulk liquid (0.14 mmol of K⁺/g of mixed-liquor VSS).

FIG. 7.

Nigericin (Nig) was added to five 200-ml activated sludge SBRs at concentrations ranging from 1.0 to 100 μM. NEM was added at 0.40 mM to a sixth SBR as a positive control for deflocculation and K⁺ efflux, and a seventh SBR served as the control. Soluble-K⁺ (bulk liquid) concentrations are plotted versus time. Datum points represent the average and error bars represent the range of duplicate samples.

Nigericin caused activated sludge deflocculation, and the degree of deflocculation increased as the concentration of nigericin increased (Fig. 8). These results show that release of cytoplasmic K⁺ into the floc leads to deflocculation of activated sludge cultures and supports the notion that the GGKE mechanism can also cause deflocculation. The degree of deflocculation obtained with 0.4 mM NEM was similar to that obtained by addition of 50 to 100 μM nigericin; however, the K⁺ efflux response for these nigericin concentrations was 46 and 61% greater (Fig. 7). Previous results indicate that the degree of deflocculation is well correlated with the amount of K⁺ efflux for NEM (6). The explanation for the discrepancy between electrophile-induced potassium efflux activity (presumed to be associated with GGKE) and nigericin addition may be the differences in the rate of the K⁺ efflux response, and this deserves further attention. The nigericin concentrations added to the activated sludge cultures and N. capsulatum represent relatively high doses, although LIVE/DEAD staining of the activated sludge cultures revealed limited death due to nigericin at the highest concentration over the time of the study (data not shown). Retention of cell viability during the nigericin study allowed this investigation to remain focused on the impact of short-term cytoplasmic K⁺ efflux responses on floc structure. Although nigericin can cause cell de-energization and eventual cell death, we did not reach that point during this study. This investigation focused on short-term responses in both activated sludge and pure cultures, and thus this potential outcome is probably not critical in the consideration of these data. The previous analysis of cell lysis in response to electrophilic stressors was purely focused on addressing the question of whether the observed K⁺ efflux response could be due to lysis as opposed to GGKE activity. Over long-term experiments, the consequence of cell killing would likely be a dramatic process upset and floc deterioration. However, in these short-term experiments, complete de-energization does not explain the rapid deflocculation response observed.

FIG. 8.

Effluent turbidities measured after 30 min of quiescent settling following the addition of either nigericin (Nig) or NEM and 5.25 h of SBR reaction time (same experiments as reported in Fig. 7). Error bars represent 1 standard deviation of quadruple turbidity measurements. NTU, nephelometric turbidity units.

Changes in pH and other cations.

Experiments were performed with both activated sludge and N. capsulatum cultures to determine if significant changes in bulk liquid or floc-associated pH and other cationic species (including Na⁺) could contradict our research hypothesis, and experiments were performed with both activated sludge and pure cultures to better understand these effects. For both activated sludge and N. capsulata cultures, no significant changes occurred in bulk liquid or floc- or cell-associated levels of other cations, including Fe²⁺, Fe³⁺, Al³⁺, Ca²⁺, Mg²⁺, and NH₄⁺ (bulk liquid measurements only). For activated sludge, this was discussed in detail previously (6).

Since cytoplasmic acidification is coincident with K⁺ efflux, a concomitant increase in bulk liquid or intrafloc structure pH might be observed. However, several factors make detection of extracellular changes in pH during cytoplasmic K⁺ efflux events difficult. First, futile cycling of H⁺, Na⁺, and NH₄⁺ can occur through other membrane-bound ionophores, resulting in a much less than equimolar exchange of K⁺ with H⁺ by the GGKE system (12, 13). Therefore, it is difficult to predict the amount of H⁺ transported for a given increase in the bulk liquid K⁺ concentration. It is also difficult to speculate on the pH changes within the floc structure as a result of GGKE activity, especially considering requirements for electroneutrality and the impermeability of the bacterial outer membrane to protons. We have conducted several studies on the impact of pH changes on deflocculation and have found that small changes in pH (±1 to 2 U) have little impact on the flocculation of activated sludge (data not shown).

For all pure-culture and activated sludge experiments, pH was monitored either periodically or online for the entire experiment by computerized data acquisition. In pure-culture experiments, we could not detect a significant change in the bulk liquid pH because of electrophilic stress or nigericin when it was measured in phosphate-buffered growth medium. When N. capsulatum was resuspended in an unbuffered medium (at similar ionic strength), the pH did change approximately 0.1 to 0.2 U, but both increasing and decreasing pH trends were observed. The same pattern was observed in experiments in which activated sludge was exposed to NEM and CDNB. When nigericin was added to activated sludge cultures, the pH increased (as expected) 0.3 to 0.4 U, but the increase was not consistent with the degree of K⁺ efflux or the nigericin dose added. In conclusion, we did not identify changes in bulk liquid pH consistent with the K⁺ efflux patterns observed in either pure or activated sludge cultures exposed to NEM or CDNB. This result does not discount the notion that cytoplasmic K⁺ efflux occurred, since other cations that do not impact pH could be taken up to maintain the charge balance. However, it does highlight the difficulty associated with experiments of this nature that involve flocculant cultures.

K⁺ efflux calculations.

As discussed above, it was previously demonstrated that when K⁺ was added externally as KCl to activated sludge at concentrations of 10 to 23 mM, deflocculation of the mixed liquor occurred (7, 22). Therefore, to further implicate GGKE-derived K⁺ efflux as a possible cause of activated-sludge deflocculation, the K⁺ concentration that would be expected in the floc structure (i.e., outside of the cells but within the floc structure) immediately following electrophilic-stressor addition was calculated.

(i) Calculations based on measured N. capsulatum efflux.

A series of eight activated sludge flocs were observed after LIVE/DEAD staining under both green-only and red-only fluorescence. Independent quantification of the red and green signals indicated an average of 70.2% green (f_cyto) and 29.8% red (f_EPS) with a standard deviation of 9.0%. For these calculations, it was assumed that the f_cyto and f_EPS fractions represent the volume fractions of activated sludge floc that are composed of viable cytoplasm (capable of concentrating K⁺) or nonviable cytoplasm and EPS, respectively. This is represented symbolically as follows:

(1)

where

(2)

and

(3)

V_floc is the total floc volume (milliliters of floc per liter of mixed liquor), V_cyto is the floc volume that is composed of viable cytoplasm, and V_EPS is the floc volume that represents dead cells or EPS.

In order to determine V_floc, the MLSS concentration and the floc density (ρ_floc) were used. An average MLSS concentration of 2,014 ± 505 mg/liter was measured from 10 independent SBR activated sludge experiments in which NEM was added at 0.4 mM. The activated sludge floc density is usually only slightly more than 1.0 g/ml for any given mixed liquor, and the overall calculation was found to be relatively insensitive to ρ_floc. It was previously determined that the floc density in the same domestic mixed liquor varied only from 1.008 to 1.024 g/ml under a variety of experimental conditions (15, 21). It is also recognized that the floc density was never less than 1.019 g/ml on the basis of the silicone fluid separation method used to rapidly separate floc material from the bulk liquid. A floc density of 1.021 g/ml was used to calculate V_floc as follows:

(4)

With this value, the measured f_cyto and f_EPS values, and equations 11 to 3, V_cyto and V_EPS were calculated to be 1.38 ml of cytoplasm/liter of mixed liquor and 0.588 ml of EPS/liter of mixed liquor, respectively.

From the N. capsulatum experiment described in Fig. 1, it was observed that the cell-associated K⁺ concentration decreased by 80% from 0.61 to 0.13 mM at a cell concentration of 2.88 ± 0.68 × 10¹⁰ CFU/ml. Therefore, N. capsulatum cells contained approximately 2.11 × 10⁻¹⁴ mmol of K⁺/cell before efflux and about 0.44 × 10⁻¹⁴ mmol of K⁺/cell after electrophile addition. For calculation purposes, it was assumed that the V_cyto of the activated sludge was composed entirely of cells of a size comparable to that of N. capsulatum. By using the measured diameter of N. capsulatum, it was possible to compute the K⁺ concentration in the cytoplasm of either N. capsulatum or activated sludge cells (C_cyto) both before and after NEM addition. It was determined that the overall calculation is most sensitive to the cell diameter, simply because the cell volume is a function of the radius cubed. As a result, the average diameter of late-log-phase and early-stationary-phase N. capsulatum cells was carefully determined by quantifying green-only LIVE fluorescence after exposure to 0.4 mM NEM. There was no detectable increase in the number of red cells due to NEM exposure at this concentration. The average cell diameter was determined by quantifying the green area for 8,388 cells in six spatially calibrated digital images. The average cell diameter was 0.682 ± 0.047 μm, which gives an average cell volume of 0.166 μm³. Therefore, C_cyto is 0.13 and 0.03 M K⁺ before and after NEM addition, respectively (a 77% decrease). This corresponds to a release of about 0.1 M K⁺ (C′_cyto). Assuming that the K⁺ is released from the bacterial cytoplasm into the floc structure EPS, the decrease in C_cyto represents the increase in the K⁺ concentration in the EPS (C_EPS). C_EPS can thus be calculated as follows:

(5)

Therefore, the concentration of K⁺ in the floc structure EPS (C_EPS) was calculated to be approximately 240 mM after electrophile addition. This concentration is an order of magnitude greater than that known to induce floc deterioration as a result of external KCl addition. If the errors associated with each of the estimated parameters are applied (i.e., term ± 1 standard deviation) such that C_EPS is minimized, the lowest estimate of C_EPS following NEM addition is 83 mM K⁺. This concentration is still sufficiently large to cause deflocculation.

The assumption that the diameter of late-log- and early-stationary-phase N. capsulatum cells is comparable to that of mixed-liquor cells may introduce significant error into this estimate approach. Therefore, a second estimate approach was used.

(ii) Calculations based on measured activated-sludge efflux.

Ten activated sludge experiments were conducted with 0.4 mM NEM added, and the average increase in the bulk liquid K⁺ concentration was 0.19 ± 0.064 mM after 5 h of contact with NEM. As expected, it was determined that the amount of K⁺ efflux derived from 0.4 mM NEM varied slightly over time but most critically depended on the MLSS concentration (R² = 0.75 and P < 0.01). An estimate of C_EPS was derived by using the average activated sludge K⁺ efflux value, assuming that all of the K⁺ that ended up in bulk liquid (0.19 mM) was originally from floc-immobilized bacterial cytoplasms. It also must be assumed that the K⁺ release into the floc structure EPS was instantaneous and was followed by diffusion of all of the C_EPS to the bulk liquid. By using the V_EPS value calculated as described above for activated sludge, the average C_EPS is 320 mM (minimum C_EPS of 162 mM if the error associated with each parameter is applied). Again, this concentration is well above that known to cause deflocculation when K⁺ is added externally.

DISCUSSION

The results presented here provide new evidence in support of the hypothesis that electrophile-induced cytoplasmic K⁺ efflux was responsible for activated sludge deflocculation. First, N. capsulatum and activated sludge exhibited dynamic K⁺ efflux-uptake-efflux responses that were similar to those of E. coli K-12 when exposed to the electrophilic stressors NEM and CDNB and the thiol reductant DTT. In the case of NEM shock loads to N. capsulatum and activated sludge cultures, the data suggest that the cytoplasmic K⁺ efflux system was inhibited by DTT addition with subsequent reaccumulation of cytoplasmic K⁺ due to the activity of cellular K⁺ uptake systems, and this was consistent with previous studies demonstrating GGKE in E. coli K-12 (3, 19). However, DTT did not reverse K⁺ efflux caused by CDNB, and this is also consistent with results reported by Booth and coworkers for GGKE in E. coli K-12 (10). These results suggest that a similar K⁺ efflux mechanism is activated in response to electrophilic stressors in N. capsulatum, activated sludge, and E. coli K-12, the organism for which the GGKE stress response was elucidated. Second, fluorescent membrane-permeable and -impermeable nucleic acid stains supported previously reported results for activated sludge (6) and indicated that cell lysis did not contribute significantly to the K⁺ efflux observed in response to electrophile addition. Third, the membrane-permeable chemical nigericin induced both cytoplasmic K⁺ efflux in activated sludge and deflocculation in activated sludge, supporting the notion that release of cytoplasmic K⁺ causes a reduction in floc strength and deflocculation because of turbulent action in the system. Nigericin addition resulted in near-complete release of cytoplasmic K⁺ into the bulk liquid, and the degree of deflocculation was comparable to that observed with NEM. Finally, a quantitative analysis of data from several independent K⁺ efflux experiments supports the notion that the electrophile-induced GGKE system has the potential to increase the intrafloc potassium concentration to the range previously shown to cause floc structure deterioration due to external addition of K⁺ as an inorganic salt. It was also observed that NEM stress resulted in much faster rates of K⁺ release into the bulk liquid for N. capsulatum than for activated sludge, suggesting that diffusion limitations slowed the release of K⁺ from flocs. Collectively, these data are consistent with the notion that cytoplasmic potassium is released in response to electrophilic shock and results in activated sludge deflocculation and strongly implicate the GGKE stress response as a significant causal mechanism.