Electrochemical Polarization-Induced Changes in the Growth of Individual Cells and Biofilms of Pseudomonas fluorescens (ATCC 17552)

The effect of surface electrochemical polarization on the growthof cells of Pseudomonas fluorescens (ATCC 17552) on gold electrodeshas been examined. Potentials positive or negative to the potentialof zero charge (PZC) of gold were applied, and these resultedin changes in cell morphology, size at cell division, time todivision, and biofilm structure. At –0.2 V (Ag/AgCl-3M NaCl), cells elongated at a rate of up to 0.19 µm min^–1,rendering daughter cells that reached up to 3.8 µm immediatelyafter division. The doubling time for the entire population,estimated from the increment in the fraction of surface coveredby bacteria, was 82 ± 7 min. Eight-hour-old biofilmsat –0.2 V were composed of large cells distributed inexpanded mushroom-like microcolonies that protruded severalmicrometers in the solution. A different behavior was observedunder positive polarization. At an applied potential of 0.5V, the doubling time of the population was 103 ± 8 min,cells elongated at a lower rate (up to 0.08 µm min^–1),rendering shorter daughters (2.5 ± 0.5 µm) afterdivision, although the duplication times were virtually thesame at all potentials. Biofilms grown under this positive potentialwere composed of short cells distributed in a large number ofcompact microcolonies. These were flatter than those grown at–0.2 V or at the PZC and were pyramidal in shape. Polarizationeffects on cell growth and biofilm structure resembled thosepreviously reported as produced by changes in the nutritionallevel of the culture medium.

INTRODUCTION

Top

Introduction

In nature, surfaces often develop a potential difference withthe surrounding aqueous medium, as a consequence of which anelectrical double layer is formed at the interface. The thicknessof this double layer depends on both the ionic properties ofthe electrolyte and the interfacial excess of charge. The potentialdifference decays mostly within the thickness of the doublelayer, inducing strong electric fields on the order of 10⁵ V/cmdue to the extremely short distances through which the potentialchange occurs. An analogous situation can be found in biologicalmembranes, where a potential difference of 0.14 to 0.20 V dropswithin a typical distance of 10 nm across the thickness of themembrane, generating field strengths of the same order of magnitude(10).

Potential differences across biological membranes (i.e., membraneelectrochemical potential, ${Delta}$ ${Psi}$ ) have been related to several fundamentalphysiological processes, including ATP synthesis, flagellarrotation, and growth yield (7). Interestingly, it has been shownin eukaryotic cells that externally applied electric fieldscan modulate some cellular processes, including development,regeneration, and motility. These effects have often been relatedto local perturbations in the cell membrane potential (7).

It is generally accepted that most bacteria in nature live formingbiofilms on virtually any kind of immersed surfaces, includingglass, polymers, minerals, and metals. When growing on metals,biofilms are often related to corrosion problems. Charged metalsurfaces can produce potential differences of several hundredmillivolts within the bulk of the surrounding electrolyte (9).Bacteria growing at the metal-electrolyte interface are thereforeexposed to potential perturbations that can modify their behavior,physiology, and growth.

The aims of the present work were to determine both the effectsof low-amplitude direct-current polarization on the growth anddivision of individual bacterial cells of Pseudomonas fluorescensand the influence of polarization on biofilm formation by thesebacteria.

MATERIALS AND METHODS

Top

Materials and Methods

Biological material.
Pure cultures of P. fluorescens (ATCC 17552) were grown overnightat 32°C with continuous shaking in a culture broth containingpeptone (Sigma) at 0.5 g liter^–1 dissolved in 0.1 M NaCl,pH 6.8. After 18 h of growth, cultures were diluted with halfa volume of fresh medium and cells were left to reach the exponentialphase of growth, as determined by measuring the absorbance at600 nm. Exponentially growing cells were harvested by centrifugationfor 10 min at 10,000 x g in a Jouan BR4i refrigerated centrifuge,washed with 0.1 M NaCl, pH 6.8, and suspended in an NaCl solutionof the same ionic strength after being centrifuged again.

Electrochemical thin-film flow cell.
All the experiments were performed using a cell described elsewhere(3). Briefly, a working electrode constructed by sputteringa thin film (10 to 20 nm) of gold onto a glass coverslip modifiedwith (3-mercaptopropyl)trimethoxysilane (6) was placed facingdown as a lid of a shallow (2-mm) acrylic chamber of 14-mm diameterinto which a 0.75-mm platinum wire counterelectrode and a referenceelectrode were fitted. This arrangement allows the observationfrom the back of bacteria attached on the working electrode.Observations were made by phase-contrast microscopy using anOlympus BH-2 transmission microscope equipped with a seriesof D Achromat positive low-contrast lenses (Olympus, Tokyo,Japan), an IF550 green filter, and a phase-contrast condenser.The integrity of the gold films was verified by microscopicobservation, and their electric resistance was measured in orderto ensure electric conductivity. The reference electrode wasAg/AgCl-3 M NaCl, and all potentials are reported against thisreference. The electrochemical cell was connected to the flowcontrol system by silicone rubber tubing through stainless steelinlets.

Voltammetric measurements.
The voltammetric response of gold electrodes exposed to a flowing0.5% NaCl solution was determined using an Autolab PGP201 potentiostat.The potential was cycled between 0.6 and –0.6 V at a scanrate of 10 mV · s^–1, starting from the open circuitpotential and in the anodic direction.

Potentiostatic growth experiments.
The flow system was inoculated with a bacterial suspension containing2 x 10⁸ to 3 x 10⁸ cells ml^–1. The bacterial cells wereallowed to adhere to the gold working electrode at the restpotential for 2 minutes or until 30 to 60 attached bacteriacould be observed in the microscope field (3). After attachment,the working electrode was polarized at the desired potentialby using an EG&G 362 scanning potentiostat (Princeton AppliedResearch). Measurements were performed at potentials of –0.5,–0.2, 0.1, 0.5, and 0.8 V. A continuous flow of sterilemedium was supplied, and the growth of bacteria was registeredas a function of time. Experiments were carried out under alaminar flow of 0.7 ml min^–1 (Reinolds coefficient = 8.10),which was created by hydrostatic pressure. Each experiment wasrecorded with a charge-coupled-device camera (500 by 582 activepixels) for 3 to 5 h in order to observe at least two consecutivedivision events for each cell. The optical magnification wasx2,000.

Sequential gray-scale images were captured from the videotapeswith 3-min time intervals in order to determine the kineticsof bacterial growth under the different potential conditions.Micrographs were analyzed using the standard measurement toolsof the public domain NIH-Image software (National Institutesof Health), which can be obtained free at http://rsb.info.gov/nih.image/.

The growth of the entire population in the field of view wasestimated from the measurement of the area covered by the cells,on the assumption that the cell diameter did not change duringeach experiment. The percentage of cell coverage in every imagewas calculated after use of a specific background correctiontool for the improvement of microscopic images, a thresholdfiltering, and a binary conversion to obtain black-and-whitepatterns on which the percentage of black pixels (cell bodies)on a white background was determined. After a linear fittingof experimental results, the doubling time (g) was calculatedas g = 0.301/S, where S is the slope of the bacterial growthwith time in a semilog plot (16). Values are reported as themeans ± standard deviations of at least three independentdeterminations. The cell length and time to division were measuredfor each growing cell. The cell length was determined by drawinga straight line between cell poles and measuring the line length.Measurements were previously calibrated with the appropriatepixels/distance ratio. Time to duplication was estimated asthe time from the initial division needed to observe the slipout of alignment between daughter cells, as described by Donachieand Begg (4). Cell length data were fitted to an exponentialgrowth expression of the form l = l₀ + A₁ exp(x/t₁), where lis the calculated cell length, l₀ is the length offset, andA₁ and t₁ are the amplitude and the time constant of the exponentialgrowth, respectively. The elongation rate was obtained as thefirst derivative of the calculated data.

After 3 and 8 h of growth, biofilms at three different areasrandomly selected on the electrode surface were analyzed. Thebiofilm thickness was measured from the determination of thevertical displacement between focal planes at the metal surfaceand at the biofilm-liquid interface, using the method of Bakkeand Olsson (1). In addition, sequential micrographs at a magnificationof x800 were taken at focal planes located at stepped distancesfrom the electrode surface in order to compare the biofilm structuresdeveloped at selected applied potentials. The distance stepwas 4.5 µm. The area covered by cells at every focal planeof the biofilms was determined by digital analysis as describedabove. Reported values are the means ± standard deviationsof at least three independent measurements.

RESULTS AND DISCUSSION

Top

Results and Discussion

Voltammetric behavior of gold in 0.5% NaCl.
The typical voltammetric response of gold electrodes in 0.5%NaCl is shown in Fig. 1. A wide interval of low current densitycorresponding to nonfaradaic processes was observed at potentialsabove –0.15 and below 0.5 V. At potentials more negativethan –0.15 V, a cathodic current wave in the reverse scancorresponding to the reduction of oxygen dissolved in the solutionwas observed. To investigate the effect of the applied potentialon cell growth in the absence of faradaic processes, growthexperiments were carried out at –0.2, 0.1, and 0.5 V.Since values ranging from 0.00 to 0.07 V (Ag/AgCl-KCl saturated)were found in the literature for the potential of zero charge(PZC) of gold in NaCl solutions (14), the experiments at 0.1V were taken as a control for unperturbed growth in the absenceof an electric field. Experiments at lower (–0.5 V) andhigher (0.8 V) potentials were also performed to evaluate theinfluence of faradaic processes on cell growth.

View larger version (13K):
[in this window]
[in a new window]
FIG. 1. Cyclic voltammetry of gold thin-film electrodes in 0.5% NaCl, pH 6.8. Arrows indicate the scan direction.

Bacterial population growth.
Results obtained for the growth under various potentials areshown in Fig. 2. During control experiments at a potential of0.1 V ( ${cong}$ PZC), the bacterial population grew exponentially witha doubling time of 82.6 ± 7 min, with no evidence ofa lag phase. When the applied potential was –0.2 V (i.e.,negative to the PZC), the population growth kinetics did notshow any difference from the control values. In contrast, ata more negative potential of –0.5 V, most of the bacteriadid not adhere directly on the surface. Instead, they formedclusters loosely connected to the electrode, making it difficultto quantify the population growth from the measurement of thecovered area. After 3 h of incubation, no evidence of growthof directly attached cells was observed, and neither was theformation of a biofilm after 24 h.

View larger version (19K):
[in this window]
[in a new window]
FIG. 2. Growth of P. fluorescens on a polarized gold surface at –0.2 V (•), the PZC ( ${blacktriangledown}$ ), 0.5 V ( ${circ}$ ), and 0.8 V ( ${square}$ ). Growth was measured every 3 min as the fraction of the total area covered by cells at every image (A_i) and was normalized to the area covered by cells at time zero (A₀). Lines indicate the linear fitting results used to calculate the doubling time (g) (see Materials and Methods).

It has been previously shown that polarization to cathodic potentialsof –0.2 V and –0.5 V prevents bacterial adhesionto an extent directly correlated to the magnitude of the appliedpotential (3). The effect was related to the electrostatic repulsiveforces at the interface. In the present work, in addition tothe negative surface charge induced on gold at both potentials,at –0.5 V oxygen reduction takes place (Fig. 1) by thereaction 2O₂ + 4H⁺ + 4e^– $->$ 4OH^–, raising the interfacialpH. The exposure of cells to an alkali stress can produce atransient failure in the intracellular pH homeostasis that hasbeen directly related to the cessation of growth and cell division(17). For Escherichia coli the upper pH limit for growth hasbeen reported to be pH 8.5 (12). In the present case, the electrochemicaloxygen reduction can increase the interfacial pH at values muchhigher than pH 8.5, stopping cell division and growth. It istherefore proposed that this effect, together with the electrostaticprevention of cell adhesion to the surface, is responsible forimpeding biofilm formation at an applied potential of –0.5V.

When potentials positive to the PZC were applied, two differenteffects were observed, depending on the value of the appliedpotential. At 0.5 V a lag phase of 30 to 40 min in the populationgrowth was followed by an exponential growth phase with a doublingtime of 103 ± 8 min, higher than the value previouslyobserved (Fig. 2). At 0.8 V, on the other hand, the bacteriadid not grow, resembling the situation observed when a potentialof –0.5 V was applied. In this case, the bacteria werein direct contact with the surface, allowing the area measurement.A slow decrease of the covered area with time was observed,indicating that initially adhered bacteria were disengaged fromthe surface by the flowing solution (Fig. 2). A biofilm hadformed at this potential after 24 h of incubation.

Growth of individual cells.
After applying a potential, the time and cell size before divisionand the elongation of individual cells were analyzed as a functionof time; these results are shown in Fig. 3 to 5.

View larger version (15K):
[in this window]
[in a new window]
FIG. 3. Duplication time for cells growing on a gold surface polarized at –0.2 V (•), the PZC ( ${blacktriangledown}$ ), and 0.5 V ( ${circ}$ ), as a function of the time of cell division during the experimentation time.

View larger version (27K):
[in this window]
[in a new window]
FIG. 5. Evolution of cell length during the growth of cells on a gold surface at a potential of –0.2 V (closed symbols) or 0.5 V (open symbols). Different symbols correspond to cells originated at increasing experimentation times (•, 42 min; ${blacksquare}$ , 72 min; ${blacktriangleup}$ , 96 min; ${circ}$ , 57 min; ${square}$ , 81min; and ${triangleup}$ , 93min). Lines indicate the exponential growth fitting results used to calculate the elongation rates of individual bacteria (see Materials and Methods). Inset, elongation rates calculated by differentiation of the exponential growth fitting results.

As can be seen in Fig. 3, the time to division (i.e., duplicationtime) of a given cell was dependent on the time at which thatparticular cell was generated, showing an adaptation of thepopulations to the applied potential. When the applied potentialwas at the PZC or more negative values (–0.2 V), the timeto division for the cells dividing first (after $~$ 30 min) variedgreatly, from 50 to 72 min. Beyond 70 min from the start ofthe experiment, the time to division showed a lower dispersionand approached 40 min. On the other hand, for experiments ata positive potential, initial duplication times were between70 to 92 min and decreased markedly with the progress of theexperiment, also reaching values of approximately 40 min. Itis important to note that duplication times were always longerfor cells at a potential positive to the PZC for a given experimentationtime, compared to those measured for cells at a negative potentialat the same time from potential application (Fig. 3).

As shown in Fig. 4, the size at division also indicated an adaptationto the imposed potential conditions. Sizes at division wereobserved to increase with the progress of the experiment atall potentials. At the PZC or at –0.2 V, sizes increasedfrom initial values ranging between 2.4 and 3.0 µm tovalues of between 3.1 and 3.8 µm. In contrast, at positivepotentials, sizes changed from 2.3 to 2.7 µm to 2.4 to3.2 µm during the progress of the experiment. The lengthwas always shorter for cells grown at positive potentials atany given time than for cells grown at negative potentials (Fig.5).

View larger version (17K):
[in this window]
[in a new window]
FIG. 4. Length at cell division of cells growing on a gold surface polarized at –0.2 V (•), the PZC ( ${blacktriangledown}$ ), and 0.5 V ( ${circ}$ ), as a function of the experimentation time.

In order to gain additional information regarding the effectof applied potentials on cell growth, the elongation of cellswas measured for experiments at –0.2 and 0.5 V. As canbe seen in Fig. 5, cells grown at negative potentials producedlonger daughter cells as a consequence of a higher rate of elongationduring growth compared with those developing at a positive potential.Calculated elongation rates reached values of up to 0.19 and0.08 µm/min at potentials of –0.2 and 0.5 V, respectively(inset in Fig. 5).

Evidence presented in Fig. 2 to 4 suggests that growth of cellsis not affected by an applied potential of –0.2 V. Incontrast, at a potential of 0.5 V, a set of deleterious effectson growth were observed. The response at this potential willbe compared to that observed at –0.2 V.

As discussed elsewhere (11), bacterial cells become much biggeras they grow faster. During a typical shift-up experiment inwhich cells are shifted from a poorer to a richer medium, therate of mass accumulation rapidly increases to a new value.The preshift rate of cell division is maintained for approximately60 min after the shift (5), but during this time cell size increases.When the bacteria are placed in a poorer medium (shift down),the number of cells continues to increase at the preshift rate,while the increase in cellular mass almost ceases (11). Aftera lag period during which the size of the cells is adjusted(diminishes) appropriately for maximum growth rate in the newmedium, growth of cellular mass proceeds at the postshift rate(11).

For cells growing at –0.2 V, the covered area grew exponentiallywith no observable lag period (Fig. 2). This was accompaniedby an adaptation period, during which the duplication time wasprogressively reduced (Fig. 3) while the length at divisionincreased (Fig. 4) due to the faster increase in the mass accumulationrate. All these changes observed at negative polarization aresimilar to those observed after a shift up in nutrient level(11).

Changes in the covered area at 0.5 V, however, resemble theresponse to a shift down. A lag period of 30 to 40 min was observedafter applying the potential. After this, growth was resumedat a rate lower than those at the other potentials (Fig. 2).Variations in cell length and time to division in this casediffer from the typical response to a shift down, since an increasein the former (Fig. 4) and a decrease in the later (Fig. 3)were observed. Beside this, it is important to note that thecells were always shorter and older at division when exposedto a positive potential than they were at a negative potential(Fig. 3 and 4), indicating that applying a positive potentialinterferes with the bacterial growth.

Although the origin of the dependence of bacterial growth onpotential remains unclear, similarities found with the responseto changes in nutrient availability deserve to be analyzed.As demonstrated in the work by Gross et al. (7), externallyapplied electric fields modulate the membrane potential of animalcells in a way that produces a hyperpolarization on the anode-facingside and a depolarization on the cathodal side of cells. Neglectingmembrane conductance, those authors interpreted the membranepotential change as the difference between the interior unperturbedpotential and the change of the local exterior potential dueto action of the electric field. If these arguments are applicableto bacteria, the question arises as to how hyperpolarizationof cell membrane due to the surface polarization to 0.5 V wouldinfluence cell growth. It is proposed that due to the high electriccomponent of membrane potential ( ${Delta}$ ${Psi}$ , negative inside) inducedby the field, proton translocation to the outer side of themembrane would be hindered, thus lowering both ${Delta}$ pH and ATP synthesis.Additionally, the transport of electrons down the transportchain would also be prevented by the increased ${Delta}$ ${Psi}$ , and reducedcompounds such as NADH or FADH would be accumulated, loweringboth the rates of main oxidative metabolic pathways and thenutrient consumption (16). This effect together with the lowATP yield could be responsible for the shift-down-like responseto positive polarization.

Biofilm structure.
To evaluate the influence of changes of individual cell growthon biofilm structure, biofilms grown at potentials of –0.2and 0.5 V were optically sectioned (magnification, x800) after3 and 8 h of incubation. As can be observed in Fig. 6a to d, apatchy biofilm grew at both potentials but showed differentdistribution patterns. After 3 h of growth at –0.2 V,microcolonies consisted of long cells distributed in an expandedand diffuse way. These were surrounded by large void spacesand channels (Fig. 6a). At a potential of 0.5 V, on the otherhand, a large number of small microcolonies regularly distributedon the surface and composed of tightly packed short cells wereobserved (Fig. 6b). For biofilms grown at both potentials, 22%of the surface was covered by cells in the basal plane (Fig.7a). The quantity of cell material was reduced markedly at thebiofilm-liquid interface. In both cases, the biofilms formedreached an estimated thickness of 13 µm after 3 h of growth(Fig. 7a). After 8 h of growth, the colony and cell featuresdescribed above at both potentials were still observed, togetherwith the changes in the three-dimensional biofilm structurediscussed below.

View larger version (171K):
[in this window]
[in a new window]
FIG. 6. Representative images showing the distribution of biofilms grown on a polarized gold surface at –0.2 V (a and c) and 0.5 V (b and d) for 3 h (a and b) and 8 h (c and d). Optical magnification, x800. The bar indicates 10 µm, and the scale applies to all the panels.

View larger version (11K):
[in this window]
[in a new window]
FIG. 7. Percentage of area covered by cells in biofilm focal planes at increasing distances from the electrode surface. Biofilms were developed at –0.2 V (gray bars) and 0.5 V (white bars) and were analyzed after 3 (a) and 8 (b) h of growth.

At –0.2 V, the formation of expanded irregular colonieswas observed. These protruded into the liquid phase, reachingan estimated thickness of 24 ± 2 µm (Fig. 7b).The higher cell density was registered at a distance of approximately8 µm away from the surface (Fig. 7b), producing microcoloniesresembling mushrooms. In contrast, at 0.5 V, colonies grew untilthey reached a maximum cell density of 27% ± 3% at thesurface level. The quantity of cell material progressively diminishedfrom the surface to the upper levels of the biofilms, renderingmicrocolonies pyramidal in shape (Fig. 7b). The maximum thicknessobserved at this potential was 16 ± 2 µm.

The changes in microcolony features and distribution can beexplained if the synergic influence of the modified growth ofindividual cells, the electrophoretic effects on charged bacteria,and the electro-osmotic induction of cell clustering (2, 13,15) is considered. All these processes are thought to contributeto the formation of tightly packed microcolonies under positivepolarization (Fig. 6b and d). First, negatively charged bacteria(3) are expected to migrate towards the surface in the electricfield associated with the anodically polarized electrode (0.5V). Second, the electrostatic interaction between surfaces oppositelycharged will favor irreversible adhesion (3). Finally, the participationof electro-osmotic ionic fluxes induced by disruption of theelectrode double layer by primary colonizing bacteria will helpwith cell clustering (13). It is important to note that allthese forces (electrophoretic, electrostatic, and electro-osmotic)will also work on newly originated bacteria, thus helping toretain daughter cells in the biofilm in a close relationshipto the surface, leading to the formation of pyramidal clusterswith a high density of cells in the basal layer (Fig. 7b).

On a negatively charged surface, on the other hand, field-inducedforces will tend to spread newly formed cells in open microcolonies(Fig. 6a and 6c), with a weak association with the surface (Fig.7b).

Even when this combination of physical processes could explainthe observed response of a biofilm structure to polarization,the participation of growth-related effects in determining biofilmstructure cannot be excluded. Interestingly, James et al. (8)reported that the irrigation of attached Acinetobacter sp. strainGJ12 cells with a high-nutrient medium led to the spread ofsurface-colonizing bacteria and to the formation of a dispersedbiofilm, while irrigation with a starvation medium led to theformation of an aggregated biofilm composed of discrete microcolonies(8). These results can clearly be connected to the observationsreported here to show a relationship between the effects ofchanges in nutrient levels and those derived from surface polarization.

Conclusions.
The growth of cells of P. fluorescens attached to a gold surfaceis in some way influenced by external polarization of the surface.The effects of polarization depend on the magnitude and polarityof the applied potential. Major effects are observed under positivepolarization and are similar to those produced by a nutritionalshift down.

ACKNOWLEDGMENTS

The present research was supported by a grant from FundaciónAntorchas, Argentina (project no. 14116-134).

Encouragement by and useful discussion with David J. Schiffrin,Liverpool, United Kingdom, are gratefully acknowledged.

FOOTNOTES

* Corresponding author. Mailing address: División Corrosión- INTEMA, UNMdP, Juan B. Justo 4302, B7608FDQ Mar del Plata, Argentina. Phone: 54 223 4816600. Fax: 54 223 4810046. E-mail: jbusalme{at}fi.mdp.edu.ar.

REFERENCES

Top