2012 Langmuir Pseudomonas Aeruginosa Attachment on QCM-D Sensors the Role Of

7/31/2019 2012 Langmuir Pseudomonas Aeruginosa Attachment on QCM-D Sensors the Role Of

1/7

Pseudomonas aeruginosa Attachment on QCM-D Sensors: The Role ofCell and Surface HydrophobicitiesIan M. Marcus, Moshe Herzberg,*, Sharon L. Walker, and Viatcheslav Freger

Department of Chemical and Environmental Engineering, University of CaliforniaRiverside, Bourns Hall B355, Riverside,California 92521, United StatesZuckerberg Institute for Water Research, Ben-Gurion University of the Negev, Sede Boqer Campus 84990, IsraelThe Wolfson Department of Chemical Engineering, TechnionIsrael Institute of Technology, Haifa 32000, Israel

*S Supporting Information

ABSTRACT: While biofilms are ubiquitous in nature, the mechanism by whichthey form is still poorly understood. This study investigated the process by whichbacteria deposit and, shortly after, attach irreversibly to surfaces by reorienting tocreate a stronger interaction, which leads to biofilm formation. A model forattachment of Pseudomonas aeruginosa was developed using a quartz crystalmicrobalance with dissipation monitoring (QCM-D) technology, along with afluorescent microscope and camera to monitor kinetics of adherence of the cellsover time. In this model, the interaction differs depending on the force thatdominates between the viscous, inertial, and elastic loads. P. aeruginosa, grown tothe midexponential growth phase (hydrophilic) and stationary phase (hydro-phobic) and two different surfaces, silica (SiO2) and polyvinylidene fluoride(PVDF), which are hydrophilic and hydrophobic, respectively, were used to testthe model. The bacteria deposited on both of the sensor surfaces, though on thesilica surface the cells reached a steady state where there was no net increase in deposition of bacteria, while the quantity of cellsdepositing on the PVDF surface continued to increase until the end of the experiments. The change in frequency and dissipationper cell were both positive for each overtone (n), except when the cells and surface are both hydrophilic. In the model threefactors, specifically, viscous, inertial, and elastic loads, contribute to the change in frequency and dissipation at each overtone

when a cell deposits on a sensor. On the basis of the model, hydrophobic cells were shown to form an elastic connection to eithersurface, with an increase of elasticity at higher overtones. At lower overtones, hydrophilic cells depositing on the hydrophobicsurface were shown to also be elastic, but as the overtone increases the connection between the cells and sensor becomes moreviscoelastic. In the case of hydrophilic cells interacting with the hydrophilic surface, the connection is viscous at each overtonemeasured. It could be inferred that the transformation of the viscoelasticity of the cell surface connection is due to changes inthe orientation of the cells to the surface, which allow the bacteria to attach irreversibly and begin bio film formation.

INTRODUCTIONMicrobial biofilms represent an ancient, protected mode ofgrowth that allows microbial survival in hostile environments aswell as the ability to disperse and colonize new niches.1 Thesesophisticated bacterial life forms are widely found on shiphulls,2 power plant piping,3 water supply pipelines,4 irrigationsystems,5 and membranes for water purification,6 within thefood industry,7 on human implants,8 and in the lungs of cysticfibrosis patients.9 In the majority of these cases, removal ofbiofilms is difficult and it implies high energetic input,10,11

application of hazardous chemicals,11 and shutting down of theimpacted system.12 Some of the physiological factors that couldinduce and affect irreversible attachment of bacteria onto asurface include aquatic chemistry, hydrodynamic effects, surfaceproperties, and the physicalchemical properties of the cells.

Biofilms generally form in nutrient-rich environments andremain in biofilm form until the bacteria become nutrient-deprived, which returns the bacteria to a planktonic state.13 Abiofilm could form on a surface in laminar or turbulent flow

systems, though the properties of the resulting biofilm maydiffer in extracellular polymeric substance (EPS) content, shearstrength,14,15 and quantity of biomass.16 Previous work hasshown that surface hydrophobicity and charge have an effect oninitial deposition and attachment of cells onto the surface,17 asdoes the surface roughness, where a rough surface will enhance

the formation of biofilms.18

Hydrophobicity and surface chargeof the cell also have an effect on biofilm formation.1 While thepreviously mentioned cell and surface properties have an effecton how biofilms form, in most cases, there are no strategies thatcould successfully prevent biofilm formation without negativeconsequences.19

Within 1 min of bacterial contact with a surface, the bondbetween the bacteria and the surface strengthens by a factor of1001000.20 Thus, to prevent biofilm formation it is important

Received: January 22, 2012Revised: March 20, 2012

Article

pubs.acs.org/Langmuir

XXXX American Chemical Society A dx.doi.org/10.1021/la300333c | Langmuir XXXX, XXX, XXXXXX
http://localhost/var/www/apps/conversion/tmp/scratch_3/pubs.acs.org/Langmuirhttp://pubs.acs.org/action/showImage?doi=10.1021/la300333c&iName=master.img-000.png&w=164&h=133http://localhost/var/www/apps/conversion/tmp/scratch_3/pubs.acs.org/Langmuir


2/7

to understand the transition from the deposition stage, duringwhich there is potential for removal, to the development ofirreversible interactions with the surface. The processes bywhich the cell approaches, deposits, and attaches to the surfacemay involve hydrodynamic and physicalchemical interac-tions21 in the deposition of cells onto the surface. They involvewater removal from the cellsurface interface22 achieving amaximal gain in free energy through increased strength of

adhesion as well as coating the surface with extracellularpolymeric substances (EPS) from the cells, which form a strongviscoelastic layer.2225

A quartz crystal microbalance with dissipation (QCM-D)allows for real-time monitoring of changes in the frequency anddissipation of an adhering mass onto a piezoelectric quartzsensor to assess cellsurface interactions. Negative frequencyshifts have been correlated with the addition of rigid mass ontoa surface following the classical Sauerbrey relation.26 Althoughthe classical interpretation of a negative frequency as an addedmass applies to a rigid planar layer rather than particles,26,27 ithas been used to model bacterial attachment to thesurface.2831 However, a few recent reports clearly indicated

that in some cases the frequency may also increase with theaddition of adherent bacteria.23,32,33 Even though this could beexplained by a loss of mass from the surface, e.g., via waterremoval from the contact area, the most common explanation isbased on the coupled oscillator model.23,3235 This modelindicates that in some cases an elastic springlike connection ofbacteria36 or micrometer-size particles34 to the sensor surfacemay counterbalance the added inertia and lead to positivefrequency shifts. Thus, frequency shift may be misleading as asole indicator of the amount of bacteria adhering to the surface.In this respect, QCM-D, monitoring changes in dissipation aswell as in resonance frequency, allows an insight intoviscoelastic and inertial characteristics of the depositedlayers.37,38 Availability of several overtone frequencies is

another advantageous feature of QCM-D, since elastic, inertial,and dissipative viscous contributions to frequency anddissipation shifts depend on frequency in a different manner.Thus QCM-D lends insight into adhesion and mechanicalcoupling between the surface and depositing bacteria as well astheir evolution with time until the bacteria are irreversiblyattached.23

The combination of QCM-D with fluorescent microscopyallows for quantifying and visualizing adhered bacteriaconcurrently with frequency and dissipation monitoring. Thisapproach is applied in this study in order to delineate amechanism of cell interaction as they approach and deposit onthe surface. Olsson et al. used this methodology to study

reversible and irreversible attachment of bacteria.

23

In that case,surface appendage lengths were varied for a single strain, andattachment was monitored on one specific surface. The presentstudy examines bacterial attachment from a different angle byvarying the sensor surface characteristics and growth conditionsof the cells. The experiments employed Pseudomonas aeruginosacells at two different growth phases with a QCM-D utilizingtwo different sensors coated either with hydrophilic silica(SiO2) or hydrophobic polyvinylidene fluoride (PVDF).Application of the QCM-D in an experimental matrixcombining cells and substrates with different surface character-istics reveals changes occurring at different bacteriasubstratuminterfaces. These results shed light on the nature of themechanical elements introduced in the coupled oscillator

theory and help connect these elements to the actualcharacteristics of the cells and their adhesion to the substrate.

MATERIALS AND METHODS

Preparation of Model Bacterial Strain. Prior to each attachmentexperiment, a 16 h culture of green fluorescent protein (GFP) labeled

P. aeruginosa PAO139 was grown in LB (LuriaBertani) broth byincubation in a shakerincubator at 150 rpm at 30 C (ZHWY-1102C,

ZHWY). The cells were diluted in LB broth (1:100) and incubated for5 or 16 h, to achieve midexponential and stationary growth phases,respectively. The growth curve of this strain is presented in theSupporting Information. Bacteria were harvested by centrifugation(4K15, Sigma) (4000gand 4 C for 10 min) and the pellet was washedthree times using a 100 mM NaCl solution (Sigma Aldrich). Afterharvesting the bacteria, the optical density (OD600) of the suspension

was adjusted to 0.1 for all subsequent QCM-D and cell character-ization experiments, which corresponds to (5 108) (5 106) cellsmL1. All characterization and QCM-D experiments were conductedusing cell suspensions in 100 mM NaCl, and all chemicals were ACSgrade.

Characterization of Bacterial Hydrophobicity. The MATH(microbial adherence to hydrocarbons) method has been described indetail previously.4042 Briefly, this method is based on the cell

partitioning between an electrolyte phase and a hydrocarbon phase.For each experiment, three replicate suspensions (4 mL) from a single

bacterial culture were transferred into test tubes containing 1 mL ofthe hydrocarbon (n-dodecane, Sigma Aldrich). The test tubes were

vortexed (Vortex Genie 2, Scientific Industries, Bohemia, NY) for 2min, followed by a 15 min rest period to allow for phase separation.The optical density of the cells in the aqueous phase was measured atOD600 (Lambda EZ 201 model, Perkin-Elmer, Waltham, MA) todetermine the extent of bacterial cell partitioning. Hydrophobicity

values are reported as the percentages of total cells that partitionedinto the hydrocarbon. Each MATH experiment was conducted ondifferent days using fresh cultures to ensure reproducibility of results.The average of the six total assays is reported in the Results andDiscussion section.

Investigation of Cell Attachment Using QCM-D. QCM-D (Q-

sense AB, Gothenburg, Sweden) provides real-time, label-freemeasurements of molecular adsorption and/or interactions takingplace on various surfaces. In addition to assessing adsorbed mass (ng/cm2 sensitivity), measured as changes in oscillating frequency (F) ofthe quartz crystal, the energy dissipation (D), which is the reducedenergy per oscillation cycle, provides novel insights regardingstructural properties of adsorbed layers. By employing an ultrasensitivecoin-shaped quartz mass sensor housed inside a flow cell with well-defined geometry and hydrodynamic characteristics, QCM-D was usedto study the initial stages of bacterial deposition and attachment. Foreach experiment, solutions were added sequentially for a period oftime to the QCM-D system using a peristaltic pump (IsmaTecPeristaltic Pump, IDEX, Glattbrugg, Switzerland) operating at a flowrate of 150 L/min in the following order: (i) distilled deionized water(DDW) for 20 min to establish a baseline of frequency, (ii) 100 mMNaCl background solution for 40 min, (iii) suspended bacteria in the

background electrolyte solution for 60 min, (iv) cell-free backgroundelectrolyte for 20 min, and (v) DDW for 20 min. Using the Qsoft 401software (Q-Sense AB, Gothenberg, Sweden), the variations offrequency (f) (Hz) and dissipation factor (D) were measuredand recorded for six overtones (n = 3, 5, 7, 9, 11, 13) throughout eachexperiment. Experiments were conducted with two sensor chemistries,gold-coated crystals coated with a polyvinylidene fluoride (PVDF)layer (catalog number QSX999) and silica coated (SiO2) (catalognumber QSX303) prepared by and purchased from Q-Sense(Gothenberg, Sweden). Each experiment was repeated to ensure thereproducibility of the results.

A QCM-D window module was used to verify and quantify the cellsattaching to the sensor. Pictures were taken with a camera (DigitalSign DS-2MBW, Nikon) mounted on the fluorescent microscope(Eclipse LV100, Nikon) every 5 min over the course of a 1 h

Langmuir Article

dx.doi.org/10.1021/la300333c | Langmuir XXXX, XXX, XXXXXXB


3/7

experiment, and images were subsequently processed in Matlab(Mathworks, Natick, MA) (representative images are presented in theSupporting Information).

Quartz Sensor Hydrophobicity. The surface hydrophobicity ofthe two different quartz crystal sensors (silica or PVDF coated) wasmeasured by adding a sessile droplet of DDW at room temperature tothe surface and measuring the contact angle using a goniometer(OCA-2, Dataphysics). These angles are measured by taking a pictureof the surface immediately after the drop is placed on the surface.

Droplets were added to five different locations on the surface to ensuresurface homogeneity. The angles of the two ends for each of thedroplets are then determined using image processing software(SCAZO Service). These results were averaged to give the values ofthe sensor surface hydrophobicity.

RESULTS AND DISCUSSIONSurface Characterization. When P. aeruginosa culture is

harvested at exponential phase, the bacteria are hydrophilicwith 22 3% of cell partitioning in n-dodecane. At stationaryphase, the bacteria are hydrophobic with 80 3% of cellpartitioning in n-dodecane (Table 1). These results are in

accordance with previous research.43 Since midexponentialphase are notably more hydrophilic than stationary cells, thecells are heretofore referred to as hydrophilic and hydrophobic

when referencing the midexponential and stationary phase cells,respectively. Similarly, the two sensor surfaces used wererepresentative of hydrophilic and hydrophobic ones. Contactangle measurements indicated the silica coated sensor to bemoderately hydrophilic (contact angle of 48 2), and thePVDF coated sensor was hydrophobic (contact angle of 90 4) (Table 1).

Electrokinetic properties of the cells and surface have beenfound previously to affect attachment.44 The results of the zetapotential of the P. aeruginosa cells in this study were measuredto be 16.4 and 11.1 mV for midexponential and stationarycells, respectively, and the sensors surfaces have a similarlynegative charge (15.1 and 11.6 mV for PVDF and SiO2,respectively).18,45 Utilizing these literature values and the

Derjaguin, Landau, Verwey, and Overbeek (DLVO) theory,46

which account for the electrostatic and van der Waal forces, allof the cellsurface interactions were identified as attractive withno energy barrier to interaction or secondary minima. (Detailsof the methods, calculations, and results are presented inSupporting Information.)

QCM-D Experiments Combined with FluorescentMicroscopy. A QCM-D with a window module attached toa fluorescent microscope and camera was used to quantify cellattachment in real-time for four experimental conditions, P.aeruginosa at two different growth phases with the two sensorsurfaces. In all of the experiments cells attached; moreover, theinitial cell flux (cells min1 cm2) was fairly close in all cases;i.e., bacteria were depositing at commensurate rates regardless

of the type of bacteria and surface. This may suggest that theinitial deposition process was governed by diffusive andconvective transport of bacteria toward the surface, whichmainly depends on their size and is not expected to significantlydiffer for all cases.47 However, at longer times two differenttrends emerged, apparently related to the nature of the sensorsurface rather than that of the cells. For the two experimentsusing the silica-coated sensor, cell attachment, as quantified by

the microscope images, increased linearly for 2030 min andthen the number of cells attached to the surface plateaued(Figure 1). On the other hand, both types of cells deposited

onto the PVDF sensor at a steady rate, thereby their numberincreased linearly for the entire experiment and did not result ina plateau.

QCM-D data could provide further insight into themechanism and strength of cell attachment to the two surfaces.Figure 2 displays the frequency and dissipation shiftsnormalized to the number of cells per unit area (cell surfacedensity) at the end of the cell deposition phase as a function ofovertone number. In all cases but one, the attachment ofbacteria described above resulted in positive frequency shifts atall overtones, as shown also in previous studies.23,32,33 Anegative frequency shift expected from the Sauerbrey equationwas only observed for the lowest overtones for hydrophilic cells

depositing on the hydrophilic SiO2 surface. Also, the absolutevalues of the frequency and dissipation shifts were decreasingwith overtone number in all cases in a similar manner.

The QCM-D response is conveniently interpreted using thefollowing relations connecting the shifts in the resonancefrequency F and dissipation D with the change of,respectively, the imaginary and real parts of complex loadimpedance ZL sensed by QCM. When, as a result of celladhesion, the load impedance changes by ZL, the change inthe frequency F = nf0 of the nth overtone and the associateddissipation are given by the following relations48

=

Ff

ZZIm( )0

qL

(1a)

Table 1. Hydrophobicity of Cells and Sensor Surfaces

cell/sensor hydrophobicity a

P. aeruginosa PAO1 (midexponential) 22 3%

P. aeruginosa PAO1 (stationary) 80 3%

silica (SiO2) 48 2

polyvinylidene fluoride (PVDF) 90 4aUnits are specific to the measurement technique, being percent ofcells partitioned in hydrocarbon and contact angle in degrees forsensor surfaces.

Figure 1. Number of cells per cm2 (N) adhered to the sensor surfaceover time for hydrophilic (midexponential growth phase) cells on ahydrophilic surface (SiO2) (), hydrophilic cells on a hydrophobicsurface (PVDF) (), hydrophobic cells (stationary growth phase) ona hydrophilic surface (), and hydrophobic cells on a hydrophobicsurface (). The reported values are the averages of replicateexperiments.

Langmuir Article

dx.doi.org/10.1021/la300333c | Langmuir XXXX, XXX, XXXXXXC
http://pubs.acs.org/action/showImage?doi=10.1021/la300333c&iName=master.img-001.png&w=209&h=170


4/7

=

DnZ

Z2

Re( )q

L(1b)

where f0 = 5 MHz is the fundamental frequency of the crystal,Zq the load impedance of the free crystal equal to 8.8 10

6 kgm2 s1 for AT-cut quartz crystal, n the overtone number, and

Re(

ZL) and Im(

ZL) designate the real and imaginary partsofZL, respectively. The change ZL should be proportionalto the number of deposited cells and the difference in ZL perunit area associated with a deposited cell, which may beexpressed as follows

= Z NA Z Z( )L C F (2)

where Nis the surface density of deposited cells, A is the sensorarea associated with an individual deposited cell (thus NA is thesensor area associated with all deposited cells), and ZC and ZFare the load impedances per unit area associated, respectively,with an individual deposited cell and with the cell-free sensor inthe fluid (water) relative to free unloaded crystal in vacuum.

Since the density of the deposited cells Nwas different in allcases, it is more appropriate to compare normalized shifts F/Nand D/N, representative of the average interaction betweenan individual cell and the surface, rather than absolute valuesF and D. However, even the normalized shifts F/N andD/N might not be straightforward to use, if one has toaddress the variation of the shifts with the overtone, sinceovertone dependence might be affected by the nonuniformityof the oscillation amplitude. In particular, due to strongerenergy trapping at larger overtones,49 cells could contributeunequally across the sensor and the degree of nonuniformitywould vary for each overtone. Thus, it may be beneficial to lookat the ratio F/D, in which both N and nonuniformity arecanceled out, thereby it becomes a truly inherent characteristicsof the specific individual cell deposited on the particular surface.The ratio F/D or its reciprocal are often used to measurethe relative rigidity of deposit material on the sensor in asimpler case of planar deposit layers.50,51

The normalized shifts F/N and D/N as well as the ratioF/D as function of the overtone number are presented in

Figures 2 and 3. The interpretation is facilitated by noting thefollowing:

(1) For a free crystal in liquid, both imaginary and real partsof ZF are equal, positive, and inversely proportional tothe fluid penetration depth that is in turn inverselyproportional to F1/2 and hence also n 1/2. This meansIm(ZF) = Re(ZF) n

1/2; therefore F/N n1/2, D/N n1/2, and F/D n (see eqs 1a and 1b). Asnoted before, the dependence ofF/Nand D/Non nmay be distorted by nonuniform oscillation, but that ofF/D should not.

(2) Since the cell diameter is about 1 m, it is much largerthan 100 nm, and then Re(ZC) > Re(ZF); i.e., thedissipation by the cell, which may be assumed to be

Figure 2. Frequency and dissipation shifts normalized to cell surfacedensity versus overtone number: (A) normalized frequency for eachsurfacecell combination and (B) normalized dissipation shift for each

cell

surface combination. The cell

surface combinations aresymbolized as hydrophilic (midexponential growth phase) cells on ahydrophilic surface (SiO2) (), hydrophilic cells on a hydrophobicsurface (PVDF) (), hydrophobic cells (stationary growth phase) ona hydrophilic surface (), and hydrophobic cells on a hydrophobicsurface () for both A and B. The reported values are the averages ofreplicate experiments.

Figure 3. Slopes of frequency divided by dissipation at the point in theexperiments where there is no increase in cell deposition on thesurface. Slopes are presented for each surface cell combination versusovertone number as follows: hydrophilic (midexponential growthphase) cells on a hydrophilic surface (SiO2) (), hydrophilic cells on ahydrophobic surface (PVDF) (), hydrophobic cells (stationarygrowth phase) on a hydrophilic surface (), and hydrophobic cells ona hydrophobic surface (). The reported values are the averages ofreplicate experiments.

Langmuir Article

dx.doi.org/10.1021/la300333c | Langmuir XXXX, XXX, XXXXXXD
http://pubs.acs.org/action/showImage?doi=10.1021/la300333c&iName=master.img-003.png&w=209&h=179http://pubs.acs.org/action/showImage?doi=10.1021/la300333c&iName=master.img-002.png&w=208&h=346


5/7

similar to that of an oscillating sphere,52 should be largerthan that of bare crystal but still related to and n in asimilar manner. D/N should then be always positiveupon cell deposition and decrease with overtone in aboutthe same manner (D/N n1/2) for all cells andsurfaces.

(3) According to the coupled-oscillator model,34,36 Im(ZC)and F are determined by two counteracting contribu-

tions, inertia of the cell and spring elasticity of theconnection between the surface and cell; this connectionmay be assumed to behave as an elastic spring. Theintertia increases Im(ZC) and leads to a negative F,while the springlike elastic connection produces anopposite effect. The absolute values of inertial and elasticloads have a different dependence on the overtonenumber. Namely, for inertia F n and F/D n3/2, i.e., negative with absolute value increasing with n,and for elastic load F +n1 and F/D +n1/2, i.e.,positive with absolute value decreasing with n.

(4) The overall F is determined by the way the dissipative,inertial, and elastic loads are connected within ZC andtheir values relative to ZF (see Figure 4 and eq2). Since

the elastic connection is the one that transmits the forcefrom the sensor to the cell, the suggested mechanicalcircuit for ZC is shown in Figure 4, which also comparesit with the circuit of clean crystal (ZF). In Figure 4, forelements connected in-series, the displacement is thesame and forces are added up (i.e., complex impedancesare added up), while for elements connected in parallel,the force is the same and displacements are added up.Thus, one should add up reciprocal impedances to obtaintotal reciprocal impedance.

Figure 2 shows that, except for the case when both surface

and cells are hydrophilic, F/N and D/N are both positive,showing a fairly similar sharp decline with n. Such a sharpdecline ofF/Nand D/Nwas not fully expected, as all typesof loads in Figure 4 predict a more moderate decline or increasewith n. The sharp dependence might be an artifact related toinstrumentation, e.g., the oscillation nonuniformity varying withn. However, as seen in Figure 3, the parameter F/D, inwhich such a nonuniformity cancels out, shows more moderatevariations with n.

Hydrophobic Cells Firmly Attach and Induce ElasticLoad on the Sensor. For the hydrophobic cells, the behavioris similar, regardless of the sensors surface, F/D beingpositive and moderately increases with n. The positive F/Nand F/D for hydrophobic cells clearly point to the

dominance of the elastic load. Apparently, the parameter F/D moderately increases with n, which may be explained by eq2 and the equivalent circuit in Figure 4. Indeed, the overallresponse is determined by the difference between ZC and ZF. IfZC is mainly elastic, the overall F/D should approximatelybehave as C1n

1/2 + C2n, where prefactors C1 and C2 are bothpositive and may vary in each case depending on dissipation,strength of elastic connection, cell rigidity, fluid viscosity, etc.

The faster changing second term, C2n, could be responsible forthe elevated F/D with n for hydrophobic cells.

The relatively high magnitude of F/D in the case ofhydrophobic cells compared to hydrophilic cells (Figure 3)implies that a firm elastic connection is formed. Water releasefrom the cellsurface interface was suggested previously to takepart in increasing the strength of cell contact with the surface,22

which is likely to play a role in this case. Note that in the case ofhydrophobic cell attachment, surface hydrophobicity had noeffect on this interaction, with a relatively similar F/D.Apparently, once a firm contact is established, the elasticcontribution is determined only by rigidity of the cell.

Hydrophilic Cells May Not Firmly Attach to theSurface. Interestingly, hydrophilic cells depositing on PVDFshowed a trend opposite to hydrophobic cells depositing oneither surface; F/D was lower and also moderatelydecreased with n. Apparently, elastic loading dominates alsoin the case of hydrophilic cells deposition on hydrophobicsurface. In this case, the equivalent circuit in Figure 4 cannotexplain the decrease in F/D with n. It seems that in thiscase, ZC crosses over from elastic (+n

1/2) behavior to morerapidly increasing inertial load (n3/2) as n increases; however,since inertia and elasticity are parallel, the smaller one, i.e.,elasticity, should actually dominate according to Figure 4, butthat is not observed. The proposed explanation is that a firmconnection between the sensor and hydrophilic cell could not befully established, i.e., the connection was viscoelastic, through afi

nite liquid gap separating the cell and surface, rather thanelastic and becomes predominantly viscous as n increases. As aresult, the elastic branch could be deactivated with increasing nand inertia took over as the dominant term contributing to F,as indeed observed.

The same argument may be extended to the case when bothsurface and cells were hydrophilic. In this case, it may beassumed that the connection was even weaker, i.e., negligiblyelastic and was mainly through viscous shear forces; thus, theelastic branch in the circuit in Figure 4 was largely deactivated.As expected and seen in Figure 3, for hydrophilic cellsdepositing on the SiO2 surface, across all overtones, the F/D is the lowest among all combinations. The absence of astrong elastic connection is consistent with the lowest slope

under these conditions and rapid saturation ofNversus time inFigure 1. Since penetration depth, , decays with n, the viscousconnection was most efficient for the lowest, third overtone,where moderate negative shifts in F/N and F/D wereobserved. The negative shift is consistent with the negligibleelastic contribution. For larger overtones (n > 3), F/N wasclose to zero; however, D/N was larger compared to othersurface-cell combinations. Presumably, without an elastic bondto the sensor, the cells were nearly motionless; thus, comparedto other cases, the viscous shear and dissipation increased dueto viscous friction of a liquid gap between the motionless celland moving surface. It should be noted that in the case ofattachment of hydrophilic cell, higher slopes of F/D wereobserved for PVDF compared to SiO2. This difference can be

Figure 4. The equivalent circuits for the mechanical load impedancesof clean crystal ZF (left) and of the crystal area associated with a singledeposited cell ZC (right) showing connection and overtone depend-ence of viscous, inertial, and elastic loads.

Langmuir Article

dx.doi.org/10.1021/la300333c | Langmuir XXXX, XXX, XXXXXXE
http://pubs.acs.org/action/showImage?doi=10.1021/la300333c&iName=master.img-004.png&w=199&h=101


6/7

explained by a faster release of water from the cell surfaceinterface and formation a more elastic and rigid bacteria PVDFsurface connection.

CONCLUDING REMARKS

In general, a positive frequency shift was shown to occur forinteractions involving hydrophobic surfaces, i.e., when eithercells or sensor surface or both were hydrophobic. Under theseconditions, a relatively firm elastic cellsurface connection issuggested to govern, and intuitively, it could be presumed thathigher hydrophobic interaction may induce more watermolecules to leave the surface, being replaced with the attachedcell wall components. Interestingly, the highest dissipationshifts per cell surface density were observed for thecombination of hydrophilic cells and hydrophilic surface. Itmay be deduced that the cells under these conditions werenearly motionless; thus, the viscous shear and dissipationincreased due to a thin liquid gap remaining between the celland sensor surface.

The highest contribution of attached cells to the increase inthe sensor frequency is shown for the combination of

hydrophobic cells and the hydrophobic PVDF, where it wasshown that elastic load dominates. Lower and similarcontribution to the increase in the sensor frequency is shownfor the combinations of hydrophobic surface and hydrophiliccells as well as hydrophilic surface and hydrophobic cells. Forthe case of hydrophobic cells on hydrophilic surface, the studyshows that still elastic load dominates, as observed by theelevation of ratio F/D as a function ofn. In contrast, for thecase of hydrophilic cells on hydrophobic surface, theconnection between the sensor and the cells becomes moreviscous as n increases. Finally, for the combination ofhydrophilic cells and hydrophilic SiO2, F/D is the lowestand the viscous connection dominates.

This investigation has led to the understanding of bacterial

attachment and the role of cell and surface hydrophobicity inthe type of cellsurface connection. It could be hypothesizedthat further transient changes in the viscoelastic characteristicsof the cel lsurface connection can be attributed toreorientation and molecular rearrangements (including removalof interfacial water) after cell deposition. This speculatedreorientation of bacteria after deposition can ensure the abilityof the cells to firmly attach to the surface and form a biofilm.

In order to reduce microbial attachment and control biofilmgrowth, extensive efforts have been devoted to the developmentof antifouling surfaces during the past decade. Better under-standing of the viscoelastic properties of the contact betweenmicroorganisms and such novel surfaces will probably enable usto predict the mechanical strength of this contact and,

therefore, to improve biofilm control strategies.

ASSOCIATED CONTENT

*S Supporting Information

Growth curve of P. aeruginosa PAO1 in LB medium, DLVOtheory applied to cellsurface interaction, and figures for eachcellsurface interaction according to the DLVO theory. Thismaterial is available free of charge via the Internet at http://pubs.acs.org.

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Phone: +972-8-6563520.

Notes

The authors declare no competing financial interest.

ACKNOWLEDGMENTS

The authors thank the Israeli Science Foundation (grantnumber 1360/10 to M.H.) for funding the research and theBinational Agricultural Research and Development Fund

(BARD GS-13-2010) for funding the fellowship of I.M.M.

REFERENCES

(1) OToole, G.; Kaplan, H. B.; Kolter, R. Biofilm formation asmicrobial development. Annu. Rev. Microbiol. 2000, 54 (1), 4979.

(2) Tribou, M.; Swain, G. The use of proactive in-water grooming toimprove the performance of ship hull antifouling coatings. Biofouling2009, 26 (1), 4756.

(3) Sriyutha Murthy, P.; Venkatesan, R.; Nair, K. V. K.; Ravindran,M. Biofilm control for plate heat exchangers using surface seawaterfrom the open ocean for the OTEC power plant. Int. Biodeterior.

Biodegrad. 2004, 53 (2), 133140.(4) Lazarova, V.; Manem, J. Biofilm characterization and activity

analysis in water and wastewater treatment. Water Res. 1995, 29 (10),

22272245.(5) Huang, T. Y.; Gulabivala, K.; Ng, Y. L. A bio-molecular film ex-vivo model to evaluate the influence of canal dimensions and irrigationvariables on the efficacy of irrigation. Int. Endod. J. 2008, 41 (1), 6071.

(6) Pang, C. M.; Hong, P.; Guo, H.; Liu, W.-T. Biofilm formationcharacteristics of bacterial isolates retrieved from a reverse osmosismembrane. Environ. Sci. Technol. 2005, 39 (19), 75417550.

(7) Chmielewski, R. A. N.; Frank, J. F. Biofilm formation and controlin food processing facilities. Compr. Rev. Food Sci. Food Saf. 2003, 2(1), 2232.

(8) Donlan, R. M. Biofilm formation: A clinically relevantmicrobiological process. Clin. Infect. Dis. 2001, 33 (8), 13871392.

(9) Costerton, J. W. Cystic fibrosis pathogenesis and the role ofbiofilms in persistent infection. Trends Microbiol. 2001, 9 (2), 5052.

(10) Gibson, H.; Taylor, J. H.; Hall, K. E.; Holah, J. T. Effectivenessof cleaning techniques used in the food industry in terms of theremoval of bacterial biofilms. J. Appl. Microbiol. 1999, 87 (1), 4148.

(11) Simoes, M.; Pereira, M. O.; Vieira, M. J. Effect of mechanicalstress on biofilms challenged by different chemicals. Water Res. 2005,39 (20), 51425152.

(12) Lund, V.; Ormerod, K. The influence of disinfection processeson biofilm formation in water distribution systems. Water Res. 1995,29 (4), 10131021.

(13) Costerton, J. W.; Lewandowski, Z.; Caldwell, D. E.; Korber, D.R.; Lappin-Scott, H. M. Microbial Biofilms; University of Cambridge:Cambridge, 1995; Vol. 49.

(14) Vrouwenvelder, J. S.; van Loosdrecht, M. C. M.; Kriuthof, J. C.A novel scenario for biofouling control of spiral wound membranesystems. Water Res. 2011, 45 (13), 28803898.

(15) Vrouwenvelder, J. S.; Buiter, J.; Riviere, M.; van der Meer, W. G.J.; van Loosdrecht, M. C. M.; Kruithof, J. C. Impact offlow regime onpressure drop increase and biomass accumulation and morphology inmembrane systems. Water Res. 2010, 44 (3), 689702.

(16) Simoes, M.; Pereira, M. O.; Sillankorva, S.; Azeredo, J.; Vieira,M. J. The effect of hydrodynamic conditions on the phenotype of

Pseudomonas fluorescens biofilms. Biofouling 2007, 23 (4), 249258.(17) Boks, N. P.; Kaper, H. J.; Norde, W.; van der Mei, H. C.;

Busscher, H. J. Mobile and immobile adhesion of staphylococcalstrains to hydrophilic and hydrophobic surfaces. J. Colloid Interface Sci.2009, 331 (1), 6064.

(18) Pasmore, M.; Todd, P.; Smith, S.; Baker, D.; Silverstein, J.;Coons, D.; Bowman, C. N. Effects of ultrafiltration membrane surfaceproperties on Pseudomonas aeruginosa biofilm initiation for the purposeof reducing biofouling. J. Membr. Sci. 2001, 194 (1), 1532.

Langmuir Article

dx.doi.org/10.1021/la300333c | Langmuir XXXX, XXX, XXXXXXF
http://pubs.acs.org/http://pubs.acs.org/mailto:[email protected]:[email protected]://pubs.acs.org/http://pubs.acs.org/


7/7

(19) Simoes, M.; Simoes, L. C.; Vieira, M. J. A review of current andemergent biofilm control strategies. LWT - Food Sci. Technol. 2010, 43(4), 573583.

(20) Vadillo-Rodrguez, V.; Busscher, H. J.; Norde, W.; de Vries, J.;van der Mei, H. C. Atomic force microscopic corroboration of bondaging for adhesion of Streptococcus thermophilus to solid substrata. J.Colloid Interface Sci. 2004, 278 (1), 251254.

(21) Torkzaban, S.; Bradford, S. A.; Walker, S. L. Resolving thecoupled effects of hydrodynamics and DLVO forces on colloidattachment in porous media. Langmuir 2007, 23 (19), 96529660.

(22) Busscher, H. J.; Weerkamp, A. H. Specific and non-specificinteractions in bacterial adhesion to solid substrata. FEMS Microbiol.

Lett. 1987, 46 (2), 165173.(23) Olsson, A. L. J.; van der Mei, H. C.; Busscher, H. J.; Sharma, P.

K. Novel analysis of bacteriumsubstratum bond maturationmeasured using a quartz crystal microbalance. Langmuir 2010, 26(13), 1111311117.

(24) Dalton, H. M.; Goodman, A. E.; Marshall, K. C. Diversity insurface colonization behavior in marine bacteria. J. Ind. Microbiol.

Biotechnol. 1996, 17 (3), 228234.(25) Fletcher, M. Bacterial Adhesion: Molecular and Ecological

Diversity. Wiley-IEEE: New York, 1996; p 361.(26) Sauerbrey, G. Verwendung von schwingquarzen zur wagung

du

nner schichten und zur mikrowa

gung. Z. Phys. A Hadrons Nuclei1959, 155 (2), 206222.(27) Hook, F.; Rodahl, M.; Brzezinski, P.; Kasemo, B. Energy

dissipation kinetics for protein and antibodyantigen adsorptionunder shear oscillation on a quartz crystal microbalance. Langmuir1998, 14 (4), 729734.

(28) Otto, K.; Elwing, H.; Hermansson, M. Effect of ionic strengthon initial interactions of Escherichia coli with surfaces, studied on-line

by a novel quartz crystal microbalance technique. J. Bacteriol. 1999,181 (17), 52105218.

(29) Serra, B.; Gamella, M.; Reviejo, A.; Pingarron, J. Lectin-modifiedpiezoelectric biosensors for bacteria recognition and quantification.

Anal. Bioanal. Chem. 2008, 391 (5), 18531860.(30) Pavey, K. D.; Barnes, L. M.; Hanlon, G. W.; Olliff, C. J.; Ali, Z.;

Paul, F. A rapid, non-destructive method for the determination of

Staphylococcus epidermidis adhesion to surfaces using quartz crystalresonant sensor technology. Lett. Appl. Microbiol. 2001, 33 (5), 344348.

(31) Nivens, D. E.; Chambers, J. Q.; Anderson, T. R.; White, D. C.Long-term, on-line monitoring of microbial biofilms using a quartzcrystal microbalance. Anal. Chem. 1993, 65 (1), 6569.

(32) Olofsson, A.-C.; Hermansson, M.; Elwing, H. Use of a quartzcrystal microbalance to investigate the antiadhesive potential of N-acetyl-L-cysteine. Appl. Environ. Microbiol. 2005, 71 (5), 27052712.

(33) Olsson, A. L. J.; van der Mei, H. C.; Busscher, H. J.; Sharma, P.K. Influence of cell surface appendages on the bacteriumsubstratuminterface measured real-time using QCM-D. Langmuir 2008, 25 (3),16271632.

(34) Pomorska, A.; Shchukin, D.; Hammond, R.; Cooper, M. A.;Grundmeier, G.; Johannsmann, D. Positive frequency shifts observed

upon adsorbing micron-sized solid objects to a quartz crystalmicrobalance from the liquid phase. Anal. Chem. 2010, 82 (6),22372242.

(35) Dybwad, G. L. A sensitive new method for the determination ofadhesive bonding between a particle and a substrate. J. Appl. Phys.1985, 58 (7), 27892790.

(36) Olsson, A. L. J.; van der Mei, H. C.; Busscher, H. J.; Sharma, P.K. Acoustic sensing of the bacteriumsubstratum interface usingQCM-D and the influence of extracellular polymeric substances. J.Colloid Interface Sci. 2011, 357 (1), 135138.

(37) de Kerchove, A. J.; Elimelech, M. Formation of polysaccharidegel layers in the presence of Ca2+ and K+ ions: Measurements andmechanisms. Biomacromolecules 2006, 8 (1), 113121.

(38) Ying, W.; Yang, F.; Bick, A.; Oron, G.; Herzberg, M.Extracellular polymeric substances (EPS) in a hybrid growth

membrane bioreactor (HG-MBR): Viscoelastic and adherencecharacteristics. Environ. Sci. Technol. 2010, 44 (22), 86368643.

(39) Banin, E.; Lozinski, A.; Brady, K. M.; Berenshtein, E.;Butterfield, P. W.; Moshe, M.; Chevion, M.; Greenberg, E. P.;Banin, E. The potential of desferrioxaminegallium as an anti-Pseudomonas therapeutic agent. Proc. Natl. Acad. Sci. U. S. A. 2008,105 (43), 1676116766.

(40) Bolster, C. H.; Cook, K. L.; Marcus, I. M.; Haznedaroglu, B. Z.;Walker, S. L. Correlating transport behavior with cell properties for

eight porcine Escherichia coli isolates. Environ. Sci. Technol. 2010, 44(13), 50085014.

(41) Pembrey, R. S.; Marshall, K. C.; Schneider, R. P. Cell surfaceanalysis techniques: What do cell preparation protocols do to cellsurface properties? Appl. Environ. Microbiol. 1999, 65 (7), 28772894.

(42) Rosenberg, M.; Gutnick, D.; Rosenberg, E. Adherence ofbacteria to hydrocarbons: A simple method for measuring cellsurfacehydrophobicity. FEMS Microbiol. Lett. 1980, 9 (1), 2933.

(43) Herzberg, M.; Rezene, T. Z.; Ziemba, C.; Gillor, O.; Mathee, K.Impact of higher alginate expression on deposition of Pseudomonasaeruginosa in radial stagnation point flow and reverse osmosis systems.

Environ. Sci. Technol. 2009, 43 (19), 73767383.(44) Walker, S. L.; Hill, J. E.; Redman, J. A.; Elimelech, M. Influence

of growth phase on adhesion kinetics of Escherichia coli D21g. Appl.Environ. Microbiol. 2005, 71 (6), 30933099.

(45) Redman, J. A.; Walker, S. L.; Elimelech, M. Bacterial adhesionand transport in porous media: Role of the secondary energyminimum. Environ. Sci. Technol. 2004, 38 (6), 17771785.

(46) Hermansson, M. The DLVO theory in microbial adhesion.Colloids Surf., B 1999, 14 (14), 105119.

(47) Elimelech, M.; Gregory, J.; Jia, X.; Williams, R. A. ParticleDeposition and Aggregation: Measurement, Modeling and Simulation;Butterworth-Heinemann: Oxford, 1995; p 441.

(48) Johannsmann, D. Viscoelastic, mechanical, and dielectricmeasurements on complex samples with the quartz crystal micro-

balance. Phys. Chem. Chem. Phys. 2008, 10 (31), 45164534.(49) Reviakine, I.; Morozov, A. N.; Rossetti, F. F. Effects of finite

crystal size in the quartz crystal microbalance with dissipationmeasurement system: Implications for data analysis. J. Appl. Phys.2004, 95 (12), 77127716.

(50) Notley, S. M.; Eriksson, M.; Wagberg, L. Visco-elastic andadhesive properties of adsorbed polyelectrolyte multilayers determinedin situ with QCM-D and AFM measurements. J. Colloid Interface Sci.2005, 292 (1), 2937.

(51) Schofield, A. L.; Rudd, T. R.; Martin, D. S.; Fernig, D. G.;Edwards, C. Real-time monitoring of the development and stability of

biofilms of Streptococcus mutans using the quartz crystal microbalancewith dissipation monitoring. Biosens. Bioelectron. 2007, 23 (3), 407413.

(52) Landau, L. D.; Lifshitz, E. M. Fluid Mechanics; Pergamon Press:Oxford, 1959.

Langmuir Article

dx.doi.org/10.1021/la300333c | Langmuir XXXX, XXX, XXXXXXG

Documents

2012 Langmuir Pseudomonas Aeruginosa Attachment on QCM-D Sensors the Role Of