EDITORS’ CHOICE

Cell-Mediated Deposition of Porous Silica onBacterial Biofilms

David Jaroch,1,2 Eric McLamore,1,3 Wen Zhang,1,4 Jin Shi,1,2 Jay Garland,5

M. Katherine Banks,1,6 D. Marshall Porterfield,1,6 Jenna L. Rickus1,2,5

1Bindley Bioscience Center and BirckNanotechnology Center, Physiological Sensing Facility,

Purdue, 225 S. University St, West Lafayette, Indiana 47907; telephone: 765-494-1197;

fax: 765-496-1116; e-mail: rickus@purdue.edu2Department of Biomedical Engineering, Purdue University, West Lafayette, Indiana3Department of Agricultural and Biological Engineering, University of Florida,

Gainesville, Florida4Department of Civil Engineering, Purdue University, West Lafayette, Indiana5Dynamac Corporation, Kennedy Space Center, Florida6Department of Agricultural and Biological Engineering, Purdue University,

West Lafayette, Indiana

Received 17 January 2011; revision received 11 April 2011; accepted 22 April 2011

Published online 2 May 2011 in Wiley Online Library (wileyonlinelibrary.com). DOI 10.1002/bit.23195

ABSTRACT: Living hybrid materials that respond dynami-cally to their surrounding environment have importantapplications in bioreactors. Silica based sol–gels representappealing matrix materials as they form a mesoporousbiocompatible glass lattice that allows for nutrient diffusionwhile firmly encapsulating living cells. Despite progress insol–gel cellular encapsulation technologies, current techni-ques typically form bulk materials and are unable to generateregular silica membranes over complex geometries for large-scale applications. We have developed a novel biomimeticencapsulation technique whereby endogenous extracellularmatrix molecules facilitate formation of a cell surface specificbiomineral layer. In this study, monoculture Pseudomonasaeruginosa and Nitrosomonas europaea biofilms are exposedto silica precursors under different acid conditions. Scan-ning electron microscopy (SEM) imaging and electron dis-persive X-ray (EDX) elemental analysis revealed the presenceof a thin silica layer covering the biofilm surface. Cellsurvival was confirmed 30min, 30 days, and 90 days afterencapsulation using confocal imaging with a membraneintegrity assay and physiological flux measurements ofoxygen, glucose, and NHþ

4 . No statistical difference inviability, oxygen flux, or substrate flux was observed afterencapsulation in silica glass. Shear induced biofilm detach-ment was assessed using a particle counter. Encapsulationsignificantly reduced detachment rate of the biofilms forover 30 days. The results of this study indicate that the thinregular silica membrane permits the diffusion of nutrientsand cellular products, supporting continued cellular viabi-

lity after biomineralization. This technique offers a means ofcontrollably encapsulating biofilms over large surfaces andcomplex geometries. The generic deposition mechanismemployed to form the silica matrix can be translated to awide range of biological material and represents a platformencapsulation technology.

Biotechnol. Bioeng. 2011;108: 2249–2260.

� 2011 Wiley Periodicals, Inc.

KEYWORDS: encapsulation; immobilization; biofilm;sol–gel; detachment; physiological sensing

Introduction

We introduce a novel cell-mediated encapsulation techni-que capable of forming a thin, flexible silica membrane onliving cells over large areas and complex geometries. Unlikebulk sol–gel encapsulation methods that entrap cells in athick silica matrix (Fennouh et al., 2000; Nassif et al., 2002),this cell-templating technique employs endogenous extra-cellular matrix and cell surface molecules as sites for thepreferential nucleation of a silica matrix in an aqueousenvironment, resulting in a porous membrane thatencapsulates the biofilm while still allowing for diffusion.

The sol–gel process is a method of producing porousmetal oxide solids from solution phase precursors. Sol–gelproduced silica glasses have excellent material propertiesfor the encapsulation of biomolecules including pharma-ceutical agents (Koehler et al., 2008; Kortesuo et al., 2000),enzymes (Bhatia et al., 2000; Luckarift et al., 2004), and

Correspondence to: J.L. Rickus

Contract grant sponsor: U.S. Army Research Laboratory

Contract grant sponsor: U.S. Army Research Office

Contract grant number: W911NF-09-1-0447

Additional Supporting Information may be found in the online version of this article.

� 2011 Wiley Periodicals, Inc. Biotechnology and Bioengineering, Vol. 108, No. 10, October, 2011 2249

other proteins (Ellerby et al., 1992; Jedlicka et al., 2007). Theinterconnected mesoporous amorphous glass network thatforms due to the polycondensation of silicic acid moleculesstabilizes sensitive biomolecules (Eggers and Valentine,2001), provides a large area for surface/solution interactions(Yang and Zhu, 2005), and allows for the controlled releaseof entrapped agents (Roveri et al., 2005). Sol–gels aretypically synthesized using wet chemistry under ambientatmosphere at standard temperatures and pressures. Theseconditions allow for the encapsulation of biological agentswithout denaturation or destruction (Avnir et al., 2006).During the past two decades, researchers have expanded thescope of sol–gel technology to encapsulate living cells bybulk casting methods for a variety of applications (Avniret al., 2006; Prakash and Bhathena, 2008) including cell-based biosensors (Bottcher et al., 2004; Livage and Coradin,2006), catalysts (Carturan et al., 1989; Inama et al., 1993),and drug delivery devices (Avnir et al., 2006; Bottcher et al.,2004; Coradin et al., 2006).

Bulk casting techniques have several limitations for cellencapsulation. The resulting encapsulation matrix has nospecificity for the cells of interest and the material geometryis governed by the surrounding vessel rather than the cells(Ferrer et al., 2003; Nassif et al., 2003). For example, castingis not feasible for biofilms grown in bioreactors due to thelack of controlled deposition. In addition, the diffusioncharacteristics of bulk matrices are dependent on thethickness and can vary widely from cell to cell (McCainand Harris, 2003; Satoh et al., 1995; Taylor et al., 2004).Some bulk colloidal silica-based sol–gel systems canovercome diffusion limitations due to their large pore size(range 30–600 nm) (Klein, 1988; Perullini et al., 2005, 2007).Alkoxysilane derived sol–gels typically produce glasses withsmall pore sizes (range 10–50 nm; Perullini et al., 2007)reducing the rate of diffusion. While it is potentially possibleto entrap large volumes of cells using these methods,diffusion constraints on nutrient, gas, and product exchangeand the brittle nature of bulk glasses tend to preclude theiruse in large-scale immobilization applications.

In contrast to bulk methods, the BioSil method, develop-ed by Carturan et al. (1997) condenses silica precursors froma high temperature vapor phase onto the surface of a proteinor cell surface to form a silica film (Carturan et al., 1997).This method improves specificity and diffusion properties,but is limited to small sample volumes. More importantlyonly precursors that can be vaporized at reasonabletemperatures and pressures can be used with cells, andtypically even easily vaporized precursors require a relativelyhigh vapor stream temperature (�808C) (Carturan et al.,1997, 2004). BioSil also requires that the cells be exposed toair and the humidity and exposure time must be controlledto prevent cell death.

The goal of the current study was to develop a cellularencapsulation technology that is versatile and easy totranslate from the lab into a working environment. To thatend we have developed a thin, flexible, cell surface-specific,silica encapsulant synthesized using simple solution-based

chemistry. The technique is inspired by the natural processesobserved in the growth of stromatolytic bacterial forma-tions. In silica rich hotsprings, bacteria serve as sites for silicanucleation (Konhauser and Ferris, 1996; Konhauser et al.,2001; Mountain et al., 2003) with extracellular proteins andcarbohydrates mediating the silicification of the cells(Konhauser and Ferris, 1996; Konhauser et al., 2001). Theresultant amorphous silica layer allows for bacterial survivalat depths of over 3mm (Konhauser and Ferris, 1996;Konhauser et al., 2001). It has been proposed that hydroxyl(–OH) groups found on proteins and mucopolysaccharidesengage in ionic and hydrogen bonding interactions with the–SiOH groups present in silicic acid containing solutions,generating preferential sites for mineral deposition (Heckyet al., 1973; Kroger et al., 1999; Lobel et al., 1996). Alternatively,the hydroxyl groups may complex with phosphates to formpreferential nucleation sites (Kroger et al., 2002). Proteinswith hydroxyl groups in conjunction with protonatedamines have been found to have catalytic functionality onthe hydrolysis and polycondensation of alkoxysilane pre-cursors (Roth et al., 2005). Positively charged amine (–NHþ

3 ) groups, such as those found on lysine residues, havealso been found to participate in silica precipitation andaggregation (Coradin et al., 2002). These same hydroxyl andamine bearing groups are found distributed in the proteinsand polysaccharides of all organisms. While such organismshave not evolved to form ordered structures, theseendogenous proteins can serve as a site for silica deposition.

The cell-mediated solution phase encapsulation techni-que described in this manuscript was developed as a simplescalable means of immobilizing cells within a custom fitfilm formed at physiological temperatures for a variety ofindustrial and environmental applications. Biofilms formedby the bacteria Pseudomonas aeruginosa and Nitrosomonaseuropaeawere selected as model systems for this study due totheir widespread distribution in the environment, and usein wastewater treatment/bioremediation. Both organismsare present in soil, sewage, and freshwater systems and arecommonly used as model gram-negative bacteria inconventional bioreactors. The two bacteria were grown inmonoculture biofilms on hollow fiber membrane aeratedbioreactors prior to encapsulation (McLamore et al., 2007,2010b). Electron microscopy and elemental analysis wereused to validate silica layer formation on mature biofilmsafter exposure to mineralizing solutions. Measurements ofsubstrate (glucose or NHþ

4 ) flux were used to characterizethe viability and physiology of the biofilms after silicaformation under different synthesis conditions. Biofilmdetachment rate was monitored for 90 days after encapsula-tion to assess the functionality of the stabilizing silica layer.

Materials and Methods

Bacterial Cell Culture

P. aeruginosa is a chemoheterotrophic gram-negativeopportunistic pathogen of animals, plants, and humans,

2250 Biotechnology and Bioengineering, Vol. 108, No. 10, October, 2011

and is used extensively as a model organism in wastewatertreatment, bioremediation, and biomedical applications(Ganguli and Tripathi, 2002; Sauer et al., 2002;Vijayaraghavan and Yun, 2008; Xu et al., 1998).P. aeruginosa PA01 (ATCC 97) was obtained fromAmerican Type Culture Collection (Manassas, VA), andbiofilms were grown at 378C in modified glucose media(10mM glucose, 50mM HEPES, 3mM NH4Cl, 43mMNaCl, 3.7mM KH2PO4, 1mM MgSO4, and 3.5mM FeSO4).

N. europaea is a chemoautotrophic gram-negativebacteria that is often the rate-limiting step in nitrogencycling within the environment (Brandt et al., 2001; Cociet al., 2005). N. europaea are sensitive to changes in nutrientconditions, light, temperature, and chemical toxins, andthus are widely studied in soil, sewage, and freshwatersystems. N. europaea (ATCC 19718) was obtained fromATCC, and biofilms were grown in ATCC medium 2265(25.0mM (NH4)2SO4, 43.0mM KH2PO4, 1.5mM MgSO4,0.25mM CaCl2, 10mM FeSO4, 0.83mM CuSO4, 3.9mMNaH2PO4, and 3.74mM Na2CO3).

All cells were cultured on semi-permeable siliconmembranes in a hollow fiber membrane aerated bioreactorand grown under aerobic conditions according toMcLamore et al. (2007, 2010b). Intact, mature biofilmswere extracted from the bioreactors via ¼00 ferrules, andtransferred to constructed flow cells prior to biosilificationaccording to McLamore et al. (2009, 2010b). Afterencapsulation, biofilms were returned to the bioreactorsfor use in long-term (90 days) detachment studies andincubated under standard growth conditions.

Enriched Silica Solution Preparation

Tetramethyl orthosilicate (TMOS, Sigma-Aldrich, St. LouisMO) was hydrolyzed in a 1:16mol ratio (TMOS/H2O)deionized water solution using 1mL of 0.04 molar acidinitiator (hydrochloric, nitric, or trifluoroacetic acid—TFA) per 1 g of solution. The mixture was stirred vigorouslyfor 10min until clear. The methanol produced by thehydrolysis reaction was removed from the solution by rotaryevaporation under vacuum at 458C (30% reduction insolution volume). The resulting saturated silica solution wasrefrigerated prior to use or used immediately. During solprocessing the pH of the solution was monitored prior tohydrolysis, after hydrolysis, after rotary evaporation, andafter addition to cellular media. Please see SupplementalFigure S1 for sol pH data.

Physiological Sensing of Oxygen, Glucose, and NHR4

To measure real-time flux of metabolic analytes, a sensingtechnique known as self-referencing (SR) was used(Porterfield, 2007). SR converts concentration sensors intodynamic biophysical flux sensors for quantifying real-timetransport in the cellular to whole tissue domain, and hasbeen used in many fields, including: agricultural (Gilliham

et al., 2006; Porterfield et al., 1999), biomedical (Land et al.,1999; Zuberi et al., 2008), and environmental (McLamoreet al., 2009, 2010b; Sanchez et al., 2008) applications. SRdiscretely corrects for signals produced by ambient drift andnoise by continuously recording differential concentration(DC) while oscillating a microsensor between two locationsseparated by a fixed excursion distance (DX), and calculatinganalyte flux using Fick’s first law of diffusion (Kuhtreiberand Jaffe, 1990).

SR sensors were used to non-invasively quantify biofilmoxygen and substrate flux using established methods(McLamore et al., 2009). Briefly, oxygen flux was measuredusing a SR optical oxygen sensor, which was constructed byimmobilizing an oxygen-quenched fluorescent dye (plati-num tetrakis pentafluoropheynl porphyrin) on the tip of atapered optical fiber. For P. aeruginosa biofilms, substrate(glucose) flux was amperometrically measured using aglucose biosensor that was fabricated by entrapping glucoseoxidase within a Nafion/carbon nanotube layer on the tipof a platinized Pt/Ir wire (McLamore et al., 2010a). ForN. europaea, substrate (NHþ

4 ) flux was measured using amicroelectrode fabricated by immersing a Ag/AgCl wire in atapered glass capillary containing electrolyte and a liquidmembrane selective for NHþ

4 (McLamore et al., 2009).For all experiments, substrate and/or O2 flux were

continuously measured at five positions along the surface ofeach biofilm for 10min unless otherwise indicated (2mmin the lateral direction between each position). For dataconcerning physiological flux, all averages represent thearithmetic mean of at least 10min of continuous recordingat five positions (n¼ 3 replicates), and error bars representthe standard error of the arithmetic mean.

Biofilm Encapsulation

The three acid initiators were screened for biocompatibilitywith biofilms by continuously measuring real-time respira-tion using oxygen optrodes. Oxygen uptake was monitoredfor 10min to determine baseline aerobic respiratory level.The media was then carefully removed and filtered mediacontaining 20mL per mL enriched silica solution was added.The samples were allowed to rest in the saturated silica for20min in order to encapsulate the biofilm. Oxygen fluxmeasurements were monitored throughout the biosilicifica-tion process. After 20min, the solution was again carefullyremoved and replaced with fresh silica-free medium to haltthe biosilicification process. Oxygen flux measurementswere then continuously recorded along the biofilm surfacefor 14 h to monitor biofilm viability. Additional encapsu-lated biofilms were returned to the bioreactor and allowed toincubate for 30 and 90 days before flux analysis. As a controlexperiment, flux was measured in growth media, thesolution was replaced with fresh growth media containingno silica, and physiological flux/viability measured. Adiagram illustrating the hollow fiber membrane, biofilm,and silica encapsulant is depicted in Figure 1.

Jaroch et al.: Cell-Mediated Deposition of Porous Silica 2251

Biotechnology and Bioengineering

SEM Imaging/Electron Dispersive X-Ray (EDX)Elemental Analysis

P. aeruginosa and N. europaea biofilms were encapsulatedwith silica as described previously. Thirty minutes afterbiofilm encapsulation, membranes containing encapsulatedbiofilms were immersed in a 4% glutaraldehyde/sterilephosphate buffer solution for 1 h. The samples were thensoaked in deionized water for 15min, followed by serialdehydration in ethanol solutions (25%, 50%, 75%, 90%, and100%, respectively). Upon removal from the final ethanolwash, the samples were placed in a partially enclosedpolystyrene dish and allowed to dry slowly under ambientconditions for 8 h. Samples were then placed in a desiccatingchamber prior to scanning electron microscopy (SEM)imaging using a FEI NOVA nanoSEM high resolutionFESEM. Encapsulated biofilm samples incubated for 30 and90 days were prepared in an identical manner.

For EDX analysis, biofilms taken 30min, 30 days, and90 days after encapsulation were fixed in 4% glutaraldehyde/sterile phosphate buffer solution for 1 h, washed four timesin deionized water to remove residual media, and driedunder ambient conditions for 8 h. The samples were thenplaced in a desiccating chamber prior to analysis using anOXFORD INCA 250 electron dispersive X-ray detector(EDX). Spectra were collected over a 120mm� 120mmarea (n¼ 3). The spectral contribution of carbon wasremoved prior to analysis due to potential interference fromthe carbon tape fixative and the atomic percentage ofthe remaining elements (reported as atomic % (at%)) wasdetermined.

Confocal Imaging

Confocal microscopy was used to quantify membraneintegrity after exposure to mineralizing solutions usinga BacLight Live/Dead viability kit (Invitrogen MolecularProbes, Carlsbad, CA). The stain consisted of a nucleic acid

(STYO9) and (propidium iodide) stain, with green and redstained cells representing cells with intact and damagedmembranes, respectively. A Zeiss LSM 710 (Thornwood,NY) confocal microscope with multi-wavelength lasers (488and 514 nm) was used for excitation. Zen software (Zeiss)was used for image capture. Nine cross sections of 144mmby 144mm were analyzed over a total biofilm depth of128mm (2mm sections) using a 10� objective.

Cellular Detachment

To characterize biofilm detachment due to shear stress,mature biofilm samples (5 cm in length) were transferred toa constructed flow cell, and exposed to a constant fluidvelocity (flowrate) for 20min. Liquid effluent was collectedin autoclaved bottles containing 5mL of isoton solution,and trypsinized using GIBCO EDTA/trypsin solution(Invitrogen, Carlsbad, CA) for 5min. Preliminary analysisindicated that trypsination did not cause significant cell lysiswithin the first 10min for either species of bacteria. Totalnumber of detached particles (0.5–60mm range) wasquantified using a Beckman Multisizer 4 Coulter Counter(Beckman Coulter, Fullerton, CA). For all plots of detach-ment data, average values represent the total number ofparticles smaller than 2mm in diameter using Multisizersoftware. The particle size was calculated by summing thetotal number of particles smaller than 2mm in diameter.A representative plot of detached particles is included inthe Supplemental Figure S2.

Results

To minimize cellular stress during the encapsulationprocess, biofilms were exposed to sols containing differentacid initiator species. The sol causing the minimalperturbation of biofilm physiology was selected for materials

Figure 1. Diagram of encapsulated biofilm on hollow fiber membrane.

2252 Biotechnology and Bioengineering, Vol. 108, No. 10, October, 2011

characterization and long-term cellular viability, physiolo-gical flux, and mechanical stability analysis. The results ofthis set of experimentation describes how sol compositioneffects biofilm response, the physical and chemical proper-ties of the encapsulant, physiology of encapsulated biofilmsover time, and the stabilizing properties of the encapsulanton cellular detachment.

Acid Initiator Screening

While all three acids produced an initial aqueous solutionwith similar pH, their effect on the hydrolysis andpolycondensation of the silicic acid species appeared tobe a major factor influencing the pH of the media after soladdition (see Supplemental Fig. S1). Measurements of thenitric acid solution pH demonstrate an increase in acidconcentration after hydrolysis due to silicic acid formation.After rotary evaporation, silicic acid is further concentrated,reducing solution pH. Sol addition resulted in a decreasein media pH, indicating free silicic acid was still present.In the case of HCl and TFA catalyzed sols, a different setof responses was observed. As before, the solution pHdecreased after hydrolysis, but not to the extent of nitric acidcatalyzed sol. This can be attributed to partial hydrolysis orpolycondensation occurring at an early stage, consumingfree silicic acid and forming nanocolonies of silica. Rotaryevaporation resulted in an increased pH in HCl catalyzedsol, indication that some of the free silicic acid wasconsumed in polycondensation. TFA sol did decrease in pHafter rotary evaporation, but not to level of nitric acidinitiated sol, indicating that partial hydrolysis or poly-condensation might have occurred. Upon addition tomedia, modest alterations of media pH for HCl and TFAcatalyzed sols were observed, indicating little non-con-densed silicic acid was available.

Respiratory oxygen flux for P. aeruginosa andN. europaeabiofilms was not significantly different amongst non-encapsulated samples for all experiments (P< 0.01,a¼ 0.05) (Fig. 2). Additionally, no significant differencein O2 flux was measured during control experiments (freshgrowth media with no silica). TFA significantly increasedrespiration rate in P. aeruginosa biofilms (285� 14%), buthad no significant effect on N. europaea respiration rate.This difference could be due to activation of stress responsemechanisms expressed by P. aeruginosa (e.g., efflux pumps,neutralizing enzymes/antioxidants), which are not presentin N. europaea (Gilbert et al., 2002; Russell, 2003).Conversely, hydrochloric acid (HCl) caused a significantincrease in average respiratory rate for N. europaea, but hadno significant effect on P. aeruginosa. For all samples usingnitric acid as the initiator, no significant difference wasmeasured between basal respiratory rate, control samples,and encapsulated biofilms. The bacteria respirome isextremely complex, and contains many stress responsemechanisms associated with temperature stress, oxidativedamage, salt stress (Coci et al., 2005; Gilbert et al., 2002;

McLamore et al., 2010b; Sauer et al., 2002), and many otherchanges in conditions. Although it is unclear as to thespecific mechanism(s) behind these increased respiratoryrates, no decrease in aerobic respiration was noted duringencapsulation with any of the acid initiators. Based on this

Figure 2. Acid initiator screening. Nitric acid does not induce a stress response.

Average oxygen influx for N. europaea and P. aeruginosa biofilms before and 5 h after

cell-mediated encapsulation using (A) trifluoroacetic acid, (B) hydrochloric acid, and

(C) nitric acid initiators. �Statistically significant difference between oxygen con-

sumption for pre- and post-encapsulation analyzed using ANOVA (n¼ 3, a¼ 0.05).

Error bars represent �2 SE of the arithmetic mean at five locations along the biofilm.

Jaroch et al.: Cell-Mediated Deposition of Porous Silica 2253

Biotechnology and Bioengineering

screening technique, nitric acid was selected as the initiator,although initiation of silicification with TFA and/or HClshould not be ruled out.

Verification of Encapsulation

SEM images of biofilms taken 30min after encapsulationexhibit morphological differences relative to controls. Cellsencapsulated in silica retained a higher degree of structuralfidelity (Fig. 3B), resisting the collapsing and smoothingeffects of dehydration that occur during sample preparation(Fig. 3A). Figure 3F shows a silica matrix bridging adjacentcells (indicated by a white arrow). A thin silica layer coatingthe exposed surface of the biofilm is apparent. This layer issufficiently ridged to prevent feature collapse on themicroscopic scale while allowing for flexibility of the bulkbiofilm. Given that the individual cells are clearly discern-able in encapsulated samples, the biomineral layer thicknesscan be inferred to be on the scale of these cellular surfacefeatures (�200 nm–1mm). After incubation for 30 days thesilica layer is no longer apparent and both bacterial filmsdisplay morphology consistent with unencapsulated cells(Fig. 3C and G). This observation is attributed to the growthof a secondary biofilm layer covering the encapsulatedsurface. After 90 days of incubation, a thick secondarybiofilm had developed, which was prone to detachmentduring SEM preparation. The 90-day sample shed the outer,secondary biofilm layer re-exposing the silica to producesamples that did not display a high degree of feature collapse(Fig. 3D and H).

Representative EDX spectra of control and encapsulatedbiofilms are presented in Figure 4. Samples displayeddetectable levels of carbon (C), oxygen (O), silica (Si),phosphorus (P), sulfur (S), aluminum (Al), sodium (Na),chlorine (Cl), and potassium (K). The elemental contribu-tion from carbon was removed due to potential backgroundartifact and the atomic percentage of the remaining elementswas then calculated. Analysis of the biofilm surfacesindicated that silica deposition took place after exposureto mineralizing solutions, causing an increase in silicaconcentration from trace levels (�1.8� 0.2% and �0.3�0.6%, P. aeruginosa and N. europaea, respectively) to15.2� 0.7% and 18.5� 0.6% for P. aeruginosa and N.europaea, respectively (Table I). Surface phosphorous, andsulfur levels decreased after the encapsulation of bothbiofilms. These elements are naturally present in both thecellular and extracellular material comprising the biofilm(Beveridge et al., 1983; Lunsdorf et al., 1997) and tend to beretained even after the extensive washing employed toremove residual media and phosphate buffer (Little et al.,1991). After 30 days of incubation, the surface silicaconcentration decreased to trace levels (�2.0� 0.5% and�1.0� 0.1% for P. aeruginosa and N. europaea, respec-tively), indicating the formation of a secondary biofilm layergrowing on top of the silica encapsulant. After 90 days silicawas still detectable in both biofilms. During SEM prepara-tion P. aeruginosa biofilms were especially prone to surfacedetachment (see Fig. 2H). The underlying layers of the filmdisplayed relatively high concentrations of silica, 5.4� 0.9%,indicating that the encapsulant was retained under asecondary biofilm layer. At the 90-day time point both

Figure 3. SEM images of biofilm surface. Silica layer present after encapsulation. Representative electron microscopy images of N. europaea biofilm (A) without silica

deposition, (B) 30 min after encapsulation, (C) 30 days after encapsulation, and (D) 90 days after encapsulation. Images of P. aeruginosa (E) before encapsulation and (F) 30 min,

(G) 30 days, and (H) 90 days after encapsulation are also presented. The arrow highlights a region of silica matrix bridging cells. Scale bars represent 4mm.

2254 Biotechnology and Bioengineering, Vol. 108, No. 10, October, 2011

biofilms also registered detectable levels of aluminum(Fig. 2E and H).

Physiology and Viability

Average oxygen and NHþ4 flux for N. europaea biofilms did

not significantly change following encapsulation using nitricacid as an initiator (P< 0.01) (Fig. 5A). This result issignificant because N. europaea has relatively few physio-logical defense mechanisms and are sensitive to subtlechanges in operating conditions (e.g., salt concentration,substrate shock) (Brandt et al., 2001; Coci et al., 2005).Likewise, metabolic respiratory rates for P. aeruginosa didnot significantly change after encapsulation using nitric acid(P< 0.01) (Fig. 5B). No significant differences in metabolic

Figure 4. EDS spectra of biofilm surface. Silica present in surface layer after encapsulation. Representative EDS spectra of N. europaea (A) before encapsulation, (B) 30 min

after encapsulation, (C) 30 days after encapsulation, and (D) 90 days after encapsulation. Corresponding spectra of P. aeruginosa (E) before encapsulation and (F) 30 min,

(G) 30 days, and (H) 90 days after encapsulation are also presented. Silica was detectable in the spectra of both biofilms for 90 days after encapsulation.

Table I. Average elemental composition (at%) of P. aeruginosa and

N. europaea biofilms before biomineralization and 30min after as

determined by EDS spectral analysis (n¼ 3 biofilms).

Si P S O

P. aeruginosa

Control 1.8� 0.2 4.9� 0.4 2.2� 0.3 88.3� 0.8

Si coated 30min 15.2� 0.7 2.0� 0.4 0.8� 0.2 76.7� 1.0

Si coated 30 days 2.0� 0.5 5.0� 0.7 2.9� 0.4 86.8� 0.8

Si coated 90 days 5.4� 0.9 3.1� 0.8 3.2� 0.8 84.8� 1.5

N. europaea

Control 0.3� 0.6 2.6� 0.4 1.7� 0.2 85.4� 1.9

Si coated 30min 18.5� 0.6 1.2� 0.1 1.0� 0.1 75.5� 0.9

Si coated 30 days 1.0� 0.1 3.6� 0.2 2.4� 0.4 82.4� 1.7

Si coated 90 days 0.7� 0.3 1.5� 0.1 0.9� 0.05 95.0� 0.2

Values are reported as average at%� SD.

Jaroch et al.: Cell-Mediated Deposition of Porous Silica 2255

Biotechnology and Bioengineering

respiration were noted between silicated samples andcontrol samples for either species (P¼ 0.031). Afterincubation for 30 and 90 days both samples maintainedconsistent levels of flux for their respective substrates(NHþ

4 for N. europaea and glucose for P. aeruginosa).Oxygen influx for P. aeruginosa increased after incubationwhile N. europaea O2 flux was not significantly changed.This increase in O2 flux correlates with an increase inbiomass growth rate, which is consistent with previousexperiments (McLamore et al., 2010b).

Viability was verified using a membrane integrity stainfor all biofilm samples, and no significant difference wasnoted between control samples and silicated samples.Representative images at 10� magnification are shownin Figure 6, and orthogonal views indicate no significantdifference in profile and depth images.

Detachment

Detachment of cells from biofilms was quantified beforeand after encapsulation under constant hydrodynamic

conditions (Fig. 7). For control samples, the averagedetachment rate normalized to fluid velocity (viaReynolds number) was 308� 14 particles min�1 for N.europaea biofilms 281� 20 particles min�1 for P. aeruginosabiofilms. This value was calculated by determining the slopeof the curve in Figure 7A, and represents the averagedetachment rate within the fluid velocity operating rangeduring 20min of exposure to steady state fluid flow.

Following encapsulation, biofilm detachment due to fluidshear stress was reduced by approximately 63� 14% and50� 8% for P. aeruginosa and N. europaea biofilms,respectively (Fig. 7B). For both species, the detachmentrate following 30 days of encapsulation was an average of24� 7% lower than control samples, but significantly higherthan newly encapsulated samples. This is due to thedetachment of cells/polymers from the secondary layer,which formed on the surface of the silica layer (confirmed byelectron microscopy, Fig. 3C and G).

After formation of a mature secondary biofilm (afterapproximately 90 days under these operating conditions),detachment rate was not significantly different than non-encapsulated cells (P¼ 0.019, a¼ 0.05), confirming theformation of a secondary biofilm.

Discussion

Monocultures of both P. aeruginosa and N. europaeabiofilms were investigated to determine if the encapsulationtechnology could be applied to organisms exhibitingdifferent species-specific stress response mechanisms, sub-strate and electron acceptor requirements, and extracellularmatrix composition. P. aeruginosa is a chemoorganoheter-otroph with a robust set of stress coping mechanismsincluding efflux pumps, quorum sensing mechanisms, andneutralizing enzymes/antioxidants (Gilbert et al., 2002;Russell, 2003). In contrast, N. europaea is a chemolithoau-totroph that lacks many of these stress response mechanismsand is relatively sensitive to environmental stressors.Biofilms of both species were able to survive silicificationand return to normal nutrient and oxygen consumptionlevels. This finding suggests that the technique may begeneralized to a diverse range of biofilm forming bacteria.

The nanostructure of solution-based silica depositsgoverns the diffusion characteristics of the encapsulant,and can be influenced by several factors including silicaconcentration, pH, curing temperature, and acid initiatorspecies (Brinker and Scherer, 1990). For live cell encapsula-tion applications, research conducted by Perullini et al.(2008, 2010, 2011) illustrates the interdependence of silicaconcentration and solution composition on cellular viabilityand the structure of the resulting glasses. In this work, theauthors demonstrated how shifts in silica concentrationaffect the porous structure of bulk silica gels and the viabilityof organisms before and after encapsulation (Perulliniet al., 2008, 2011). The need to maintain cellular viabilityconstrains parameter type and range that can be varied

Figure 5. Oxygen and substrate flux. Encapsulation does not significantly alter

biofilm metabolism. A: Average oxygen and substrate influx for N. europaea biofilm

before silica encapsulation, after encapsulation, and after 30- and 90-day incubation.

B: Average oxygen and substrate flux for a P. aeruginosa biofilm before silica

encapsulation, after encapsulation, and after 30- and 90-day incubation. Average

values represent 10min of continuous recording at seven different positions along

the biofilm surface. For all biofilms there was no significant spatial heterogeneity of

flux (P< 0.01, a¼ 0.05) along the surface (different positions separated by 1mm).

2256 Biotechnology and Bioengineering, Vol. 108, No. 10, October, 2011

to influence nanoarchitecture. Temperature and finalsolution pH had to remain within the organism’s optimalgrowth range, which is especially important for obligateaerobic bacteria such as N. europaea, which are not capable

of utilizing metabolic/proteomic stress response mechan-isms (such as shifting to anaerobic respiration) to adapt tolarge changes in environmental conditions (Gilbert et al.,2002; Russell, 2003). In preliminary experiments, silica

Figure 6. Membrane integrity imaging. Encapsulation does not impact overall biofilm viability. Representative confocal microscopy images (10� magnification) of

N europaea biofilm (A) without silica deposition and (B) 30 min after encapsulation. Images E and F represent an unmodified P. aeruginosa biofilm and a biofilm encapsulated

for 30min, respectively. Biofilms incubated for 30 (C and G) and 90 days (D and H) for N europaea and P. aeruginosa respectively do not demonstrate a decrease in viability over

time. Biofilms were stained with a BacLight membrane integrity skit, where green represents cells with intact membranes, and red indicates cells which have lysed or damage

membranes. Scale bars represent 200mm. Orthogonal view represents composite of two-dimensional images taken every 2mm over biofilm depth.

Figure 7. Evaluation of biofilm detachment rate. A: Average number of detached particles (0.5–2.0mm) from P. aeruginosa biofilms due to fluid shear (Re¼ 3.0–100). B: Total

number of detached particles per axial Reynolds number (i.e., slope of plot in panel A) for P. aeruginosa and N. europaea biofilms. All values represent an average of three

liquid samples taken from five replicate experiments, and error bars represent� 2 SE. �Samples with a significantly lower number of detached particles relative to basal levels

(P< 0.01, a¼ 0.05).

Jaroch et al.: Cell-Mediated Deposition of Porous Silica 2257

Biotechnology and Bioengineering

concentration was adjusted to provide a window wheredeposition occurred on the cellular substrate prior to bulkgelation. Given these constraints, three types of acid initiator(hydrochloric, nitric, and TFA) were screened to determinetheir effects on biofilm viability.

The respiration rate of cells encapsulated with silicahydrolyzed by nitric acid did not significantly change, whilehydrochloric and trifluoroacetic acid significantly alteredrespiration rate within 30min (Fig. 2). The increasedrespiration rate observed in P. aeruginosa (using trifluor-oacetic acid) andN. europaea (using hydrochloric acid) maybe indicative of oxidative stress. Importantly, respiration didnot decrease for any of the acid initiators.

The reason for the different response to hydrochloric andtrifluoroacetic acid is unclear. Acid initiator species canindirectly influence biofilm metabolism by altering theporosity and diffusional characteristics of the silica matrix(Brinker and Scherer, 1990). Alternatively, the acid initiatormay interact directly with the bacteria, stressing the cells.Different bacteria vary in their ability to adapt to acidicenvironments and the acid type can have specific affectsbeyond alterations in pH (Kobayashi et al., 2000).P. aeruginosa are capable of facultative respiration, whileN. europaea are obligate aerobes; thus selection of anacid initiator which can facilitate the encapsulation ofa broad range of microbial species is required. To fulfillthis requirement, nitric acid initiated silica solutions werechosen for further investigation.

Using nitric acid as an initiator, SEM imaging and EDXelemental analysis confirmed the presence of a silica coatingin both species of biofilm (Fig. 3B and D and Table I).Elemental analysis demonstrated significantly higher con-centrations of silica on the surface of encapsulated cells(18.49� 0.59 and 15.24� 0.67 at% for N. europaea andP. aeruginosa, respectively) than control samples (Table I).The trace levels of silicon detected in control samples(�1.8 at% and �0.3 at% for P. aeruginosa and N. europaea,respectively) may be attributed to the residue left from thesilicone tubing utilized to culture the biofilm. Silica wasfound at detectable levels 90 days after encapsulation.During incubation, a secondary biofilm layer grew on theencapsulated surface, reducing the concentration of silicadetectable by EDS. This secondary layer partially detached inP. aeruginosa biofilms at the 90-day time point (see Fig. 2H)increasing the contribution of the underlying silica layer tothe resulting EDX spectra. The presence of aluminum inconjunction with silica in the 30 and 90 days samples likelyarises from silica’s ability to fix aluminum ions in solution(Birchall, 1995; Dugger et al., 1964). This characteristichas been employed to remove metallic contaminantsfrom wastewater (Van Jaarsveld et al., 1998; Zamzowet al., 1990). When utilized as an encapsulant for wastewatertreatment applications, silica’s interaction with aluminummay present a secondary health benefit. The incidence ofAlzheimer’s disease has been linked to elevated aluminumlevels and lack of silica in the water supply (Rondeauet al., 2009).

Cellular viability was confirmed using both physiologicalflux measurements and live/dead staining. Metabolicoxygen, glucose (P. aeruginosa), and NHþ

4 (N. europaea)flux was confirmed after encapsulation with silica (Fig. 5).These results indicated that the silica matrix was sufficientlyporous to allow for the diffusion of dissolved gasses andnutrients. Biophysical transport of nutrients and electronacceptors regulates synthesis and maintenance of cellswithin the biofilm, and is limited by the concentrationboundary layer formed at the biofilm–fluid interface.No significant change in oxygen flux, substrate flux, orstoichiometric metabolic ratio was observed after encapsu-lation (P< 0.02, a¼ 0.05), suggesting that cells survived theencapsulation process intact. No observable differences werenoted at 10� magnification in stained samples analyzedusing confocal microscopy (Fig. 6). There were no largeregions of lysed cells within the matrix (2mm slices), whichone would expect if diffusion limitations or nutrienttransport was significantly altered by silica encapsulation.Preservation of physiology and cell viability is particularlyimportant for N. europaea, which often limit nitrogencycling in natural and engineered systems due to the lack of awide array of stress response mechanisms. Encapsulation ofsessile bacteria can potentially improve resistance to shockloading, chemical toxin exposure, and detachment. Suchimprovements can then be translated into the developmentof more efficient bioreactors for chemical production(Mohan et al., 2007; Qureshi et al., 2005) and waterrecovery (Jackson et al., 2009).

Encapsulation of biofilms significantly reduced thedetachment rate, although a secondary biofilm attachedafter 30- and 90-day incubation and was prone todetachment. The number of detached particles consistedof a combination of lysed cells, active cells, and polymericmaterial. The specific cause of biofilm detachment is outsidethe scope of this work, although the mechanisms are likely acombination of passive detachment (fluid shear, abrasion)and active detachment (active transport mechanisms thatdenature extracellular polymers) (Beer and Stoodley, 2006;Stoodley et al., 2001, 2005).

Taken together the results of this study indicate that bothpersistent P. aeruginosa and sensitive N. europaea speciessurvive the encapsulation process and retain their viability.Unlike bulk encapsulation processed, the cell-mediatedtechnique generates a thin flexible silica membrane onbiofilm surfaces exposed to the surrounding mineralizingsolution. Unlike the BioSil methods, silica formation isperformed at culture temperatures in solution phase. Thethin evenly distributed silica layer does not serve as a barrierto molecular diffusion of O2 or cellular substrate, with nosignificant variation in physiological flux observed betweenencapsulated and control biofilms. The silica is alsointimately associated with extracellular proteins, reinforcingthe silica matrix and preventing the cracking observed inbulk encapsulated cellular materials. The formation ofthe silica layer is dependent upon cellular contact withthe mineralizing solution. The technique can therefore be

2258 Biotechnology and Bioengineering, Vol. 108, No. 10, October, 2011

applied to systems with complex geometries so long assolution contact can be maintained during encapsulation.The solution volume can also be expanded for large-scaleencapsulation of living cells for industrial and environ-mental applications.

Conclusions

The cell-mediated encapsulation technique is a simplesolution based method for encapsulating living cells in a thinporous silica membrane. Monoculture biofilms of thedistinct bacterial species P. aeruginosa and N. europaeawere able to maintain physiological flux of O2, glucose(P. aeruginosa), and NHþ

4 (N. europaea) and weredetermined to be viable by live/dead staining afterencapsulation. A significant reduction in shear-induced celldetachment was observed in both biofilms after encapsula-tion. The cell-mediated encapsulation technique employsendogenous extracellular material as a site for deposition.As such it can potentially be applied to a wide range ofprokaryotic and eukaryotic cell lines and mixed culturesystems. Future work will explore the long-term viability ofencapsulated biofilms, multilayer biofilm systems, and theapplication of the technique to other cell lines.

This material is based upon work supported by the U.S. Army

Research Laboratory and the U.S. Army Research Office under con-

tract/grant number W911NF-09-1-0447. The authors would like to

thank Debra M. Sherman of the Purdue Life Science Microscopy

Facility for her expert imaging and the use of the SEM/EDS equipment

utilized to conduct this study.

References

Avnir D, Coradin T, Lev O, Livage J. 2006. Recent bio-applications of sol-gel

materials. J Mater Chem 16(11):1013–1030.

Beer D, Stoodley P. 2006. Microbial biofilms. In: Falkow S, Rosenberg E,

Schleifer K-H, Stackebrandt E, Dworkin M, editors. The prokaryotes.

Symbiotic Associations, Biotechnology, Applied Microbiology.

Germany: Springer. p 904–937. DOI: 10.1007/0-387-30741-9_28

Beveridge TJ, Meloche JD, Fyfe WS, Murray RGE. 1983. Diagenesis of

metals chemically complexed to bacteria—Laboratory formation

of metal phosphates, sulfides, and organic condensates in artificial

sediments. Appl Environ Microbiol 45(3):1094–1108.

Bhatia RB, Brinker CJ, Gupta AK, Singh AK. 2000. Aqueous sol-gel process

for protein encapsulation. Chem Mater 12(8):2434–2441.

Birchall JD. 1995. The essentiality of silicon in biology. Chem Soc Rev

24(5):351.

Bottcher H, Soltmann U, Mertig M, Pompe W. 2004. Biocers: Ceramics

with incorporated microorganisms for biocatalytic, biosorptive and

functional materials development. J Mater Chem 14(14):2176–2188.

Brandt KK, Hesselsoe M, Roslev P, Henriksen K, Sorensen J. 2001. Toxic

effects of linear alkylbenzene sulfonate on metabolic activity, growth

rate, and microcolony formation of Nitrosomonas and Nitrosospira

strains. Appl Environ Microbiol 67(6):2489–2498.

Brinker CJ, Scherer GW. 1990. Sol-gel science. San Diego California:

Academic Press. ISBN-10: 0121349705 ISBN-13: 9780121349707.

Carturan G, Campostrini R, Dire S, Scardi V, Dealteriis E. 1989. Inorganic

gels for immobilization of biocatalysts—Inclusion of invertase-active

whole cells of yeast (Saccharomyces-Cerevisiae) into thin-layers of SiO2

gel deposited on glass sheets. J Mol Catal 57(1):L13–L16.

Carturan G, Dellagiacoma G, Rossi M, Monte RD, Muraca M. 1997.

Encapsulation of viable animal cells for hybrid bioartificial organs

by the Biosil method. In: Bruce SD, John DM, Edward JAP,

Helmut KS, Masayuki Y, editors. Sol-Gel Optics IV (SPIE Proceed-

ings), Vol. 3136, SPIE. p 366–373. DOI: 10.1117/12.279168

Carturan G, Dal Toso R, Boninsegna S, Dal Monte R. 2004. Encapsulation

of functional cells by sol-gel silica: Actual progress and perspectives for

cell therapy. J Mater Chem 14(14):2087–2098.

Coci M, Riechmann D, Bodelier PLE, Stefani S, Zwart G, Laanbroek HJ.

2005. Effect of salinity on temporal and spatial dynamics of ammonia-

oxidising bacteria from intertidal freshwater sediment. FEMS Micro-

biol Ecol 53(3):359–368.

Coradin T, Durupthy O, Livage J. 2002. Interactions of amino-containing

peptides with sodium silicate and colloidal silica: A biomimetic

approach of silicification. Langmuir 18(6):2331–2336.

Coradin T, Boissiere M, Livage J. 2006. Sol-gel chemistry in medicinal

science. Curr Med Chem 13(1):99–108.

Dugger DL, Stanton JH, McConnel Bl, Irby BN, Cummings WW,Maatman

RW. 1964. Exchange of 20 metal ions with weakly acidic silanol group

of silica gel. J Phys Chem 68(4):757.

Eggers DK, Valentine JS. 2001. Molecular confinement influences protein

structure and enhances thermal protein stability. Protein Sci 10(2):

250–261.

Ellerby LM, Nishida CR, Nishida F, Yamanaka SA, Dunn B, Valentine JS,

Zink JI. 1992. Encapsulation of proteins in transparent porous silicate-

glasses prepared by the sol-gel method. Science 255(5048):1113–1115.

Fennouh S, Guyon S, Livage J, Roux C. 2000. Sol-gel entrapment of

Escherichia coli. J Sol-Gel Sci Technol 19(1–3):647–649.

Ferrer ML, Yuste L, Rojo F, del Monte F. 2003. Biocompatible sol-gel route

for encapsulation of living bacteria in organically modified silica

matrixes. Chem Mater 15(19):3614–3618.

Ganguli A, Tripathi AK. 2002. Bioremediation of toxic chromium from

electroplating effluent by chromate-reducing Pseudomonas aeruginosa

A2Chr in two bioreactors. Appl Microbiol Biotechnol 58(3):416–420.

Gilbert P, McBain AJ, Bloomfield SF. 2002. Biocide abuse and antimicrobial

resistance: Being clear about the issues. J Antimicrob Chemother 50(1):

137–139.

Gilliham M, Sullivan W, Tester M, Tyerman SD. 2006. Simultaneous flux

and current measurement from single plant protoplasts reveals a strong

link between Kþ fluxes and current, but no link between Ca2þ fluxes

and current. Plant J 46(1):134–144.

Hecky RE, Mopper K, Kilham P, Degens ET. 1973. Amino-acid and sugar

composition of diatom cell-walls. Mar Biol 19(4):323–331.

Inama L, Dire S, Carturan G, Cavazza A. 1993. Entrapment of viable

microorganisms by SiO2 sol-gel layers on glass surfaces—Trapping,

catalytic performance and immobilization durability of Saccharo-

myces-Cerevisiae. J Biotechnol 30(2):197–210.

Jackson VA, Paulse AN, Bester AA, Neethling JH, Khan S, Khan W. 2009.

Bioremediation of metal contamination in the Plankenburg River,

Western Cape, South Africa. Int Biodeter Biodegr 63(5):559–568.

Jedlicka SS, Little KM, Nivens DE, Zemlyanov D, Rickus JL. 2007. Peptide

ormosils as cellular substrates. J Mater Chem 17(48):5058–5067.

Klein LC. 1988. Sol-gel technology for thin films, fibers, preforms, electro-

nics and specialty shapes. Noyes: William Andrew Publishing.

Kobayashi H, Saito H, Kakegawa T. 2000. Bacterial strategies to inhabit

acidic environments. J Gen Appl Microbiol 46(5):235–243.

Koehler JJ, Zhao J, Jedlicka SS, Porterfield DM, Rickus JL. 2008. Compart-

mentalized nanocomposite for dynamic nitric oxide release. J Phys

Chem B 112(47):15086–15093.

Konhauser KO, Ferris FG. 1996. Diversity of iron and silica precipitation by

microbial mats hydrothermal waters, Iceland: Implications for Pre-

cambrian iron formations. Geology 24(4):323–326.

Konhauser KO, Phoenix VR, Bottrell SH, Adams DG, Head IM. 2001.

Microbial-silica interactions in Icelandic hot spring sinter: Possible

analogues for some Precambrian siliceous stromatolites. Sedimentol-

ogy 48(2):415–433.

Kortesuo P, Ahola M, Karlsson S, Kangasniemi I, Yli-Urpo A, Kiesvaara J.

2000. Silica xerogel as an implantable carrier for controlled drug

Jaroch et al.: Cell-Mediated Deposition of Porous Silica 2259

Biotechnology and Bioengineering

delivery—Evaluation of drug distribution and tissue effects after

implantation. Biomaterials 21(2):193–198.

Kroger N, Deutzmann R, Sumper M. 1999. Polycationic peptides

from diatom biosilica that direct silica nanosphere formation. Science

286(5442):1129–1132.

Kroger N, Lorenz S, Brunner E, Sumper M. 2002. Self-assembly of highly

phosphorylated silaffins and their function in biosilica morphogenesis.

Science 298(5593):584–586.

KuhtreiberWM, Jaffe LF. 1990. Detection of extracellular calcium gradients

with a calcium-specific vibrating electrode. J Cell Biol 110(5):1565–

1573.

Land SC, Porterfield DM, Sanger RH, Smith PJS. 1999. The self-referencing

oxygen-selective microelectrode: Detection of transmembrane oxygen

flux from single cells. J Exp Biol 202(2):211–218.

Little B, Wagner P, Ray R, Pope R, Scheetz R. 1991. Biofilms—An ESEM

evaluation of artifacts introduced during SEM preparation. J Ind

Microbiol 8(4):213–221.

Livage J, Coradin T. 2006. Living cells in oxide glasses. Medical mineralogy

and geochemistry. Chantilly: Mineralogical Soc America. p 315–

332.

Lobel KD, West JK, Hench LL. 1996. Computational model for protein

mediated biomineralization of the diatom frustule. Mar Biol 126(3):

353–360.

Luckarift HR, Spain JC, Naik RR, Stone MO. 2004. Enzyme immobilization

in a biomimetic silica support. Nat Biotechnol 22(2):211–213.

Lunsdorf H, Brummer I, Timmis KN, WagnerDobler I. 1997. Metal

selectivity of in situ microcolonies in biofilms of the Elbe river. J

Bacteriol 179(1):31–40.

McCain KS, Harris JM. 2003. Total internal reflection fluorescence-correla-

tion spectroscopy study of molecular transport in thin sol-gel films.

Anal Chem 75(14):3616–3624.

McLamore E, JacksonWA,Morse A. 2007. Abiotic transport in a membrane

aerated bioreactor. J Membr Sci 298(1–2):110–116.

McLamore ES, Porterfield DM, Banks MK. 2009. Non-invasive self-refer-

encing electrochemical sensors for quantifying real-time biofilm analyte

flux. Biotechnol Bioeng 102(3):791–799.

McLamore ES, Shi J, Jaroch DB, Barcus C, Osbourne J, Uchida A,

Jiang Y, Buhman KK, Banks K, Teegarden D. others. 2011. A self

referencing enzyme-based microbiosensor for real time measurement

of physiological glucose transport. Biosens Bioelectron 26(5):2237–

2245.

McLamore ES, Zhang W, Porterfield DM, Banks MK. 2010b. Membrane-

aerated biofilm proton and oxygen flux during chemical toxin expo-

sure. Environ Sci Technol 44(18):7050–7057.

Mohan SV, Bhaskar YV, Sarma PN. 2007. Biohydrogen production from

chemical wastewater treatment in biofilm configured reactor operated

in periodic discontinuous batchmode by selectively enriched anaerobic

mixed consortia. Water Res 41(12):2652–2664.

Mountain BW, Benning LG, Boerema JA. 2003. Experimental studies

on New Zealand hot spring sinters: Rates of growth and textural

development. Can J Earth Sci 40(11):1643–1667.

Nassif N, Bouvet O, Rager MN, Roux C, Coradin T, Livage J. 2002. Living

bacteria in silica gels. Nat Mater 1(1):42–44.

Nassif N, Roux C, Coradin T, Rager MN, Bouvet OMM, Livage J. 2003.

A sol-gel matrix to preserve the viability of encapsulated bacteria.

J Mater Chem 13(2):203–208.

Perullini M, Jobbagy M, Soler-Illia G, Bilmes SA. 2005. Cell growth at

cavities created inside silica monoliths synthesized by sol-gel. Chem

Mater 17(15):3806–3808.

Perullini M, Rivero MM, Jobbagy M, Mentaberry A, Blimes SA. 2007. Plant

cell proliferation inside an inorganic host. J Biotechnol 127(3):542–

548.

Perullini M, Jobbagy M, Moretti MB, Garcia SC, Bilmes SA. 2008. Optimiz-

ing silica encapsulation of living cells: In situ evaluation of cellular

stress. Chem Mater 20(9):3015–3021.

Perullini M, Jobbagy M, Mouso N, Forchiassin F, Bilmes SA. 2010.

Silica-alginate-fungi biocomposites for remediation of polluted water.

J Mater Chem 20(31):6479–6483.

Perullini M, AmouraM, Roux C, Coradin T, Livage J, JapasML, JobbagyM,

Bilmes SA. 2011. Improving silica matrices for encapsulation of Escher-

ichia coli using osmoprotectors. J Mater Chem 21(12):4546–4552.

Porterfield DM. 2007. Measuring metabolism and biophysical flux in

the tissue, cellular and sub-cellular domains: Recent developments

in self-referencing amperometry for physiological sensing. Biosens

Bioelectron 22(7):1186–1196.

Porterfield DM, Kuang AX, Smith PJS, Crispi ML, Musgrave ME. 1999.

Oxygen-depleted zones inside reproductive structures of Brassicaceae:

Implications for oxygen control of seed development. Can J Bot

77(10):1439–1446.

Prakash S, Bhathena J. 2008. Live immobilised cells as new therapeutics.

J Drug Deliv Sci Technol 18(1):3–14.

Qureshi N, Annous BA, Ezeji TC, Karcher P, Maddox IS. 2005. Biofilm

reactors for industrial bioconversion processes: Employing potential of

enhanced reaction rates. Microb Cell Fact 4:24.

Rondeau V, Jacqmin-Gadda H, Commenges D, Helmer C, Dartigues

JF. 2009. Aluminum and silica in drinking water and the risk of

Alzheimer’s disease or cognitive decline: findings from 15-year

follow-up of the PAQUID cohort. Am J Epidemiol 169(4):489–496.

Roth KM, Zhou Y, YangWJ, Morse DE. 2005. Bifunctional small molecules

are biomimetic catalysts for silica synthesis at neutral pH. J Am Chem

Soc 127(1):325–330.

Roveri N,MorpurgoM, Palazzo B, Parma B, Vivi L. 2005. Silica xerogels as a

delivery system for the controlled release of different molecular weight

heparins. Anal Bioanal Chem 381(3):601–606.

Russell AD. 2003. Similarities and differences in the responses of micro-

organisms to biocides. J Antimicrob Chemother 52(5):750–763.

Sanchez BC, Ochoa-Acuna H, PorterfieldDM, SepulvedaMS. 2008. Oxygen

flux as an indicator of physiological stress in fathead minnow (Pime-

phales promelas) embryos: A real-time biomonitoring system of water

quality. Environ Sci Technol 42(18):7010–7017.

Satoh S, Matsuyama I, Susa K. 1995. Diffusion of gases in porous silica-gel.

J Non-Cryst Solids 190(3):206–211.

Sauer K, Camper AK, Ehrlich GD, Costerton JW, Davies DG. 2002.

Pseudomonas aeruginosa displays multiple phenotypes during devel-

opment as a biofilm. J Bacteriol 184(4):1140–1154.

Stoodley P, Wilson S, Hall-Stoodley L, Boyle JD, Lappin-Scott HM,

Costerton JW. 2001. Growth and detachment of cell clusters from

mature mixed-species biofilms. Appl Environ Microbiol 67(12):5608–

5613.

Stoodley P, Purevdorj-Gage B, Costerton JW. 2005. Clinical significance of

seeding dispersal in biofilms: A response. Microbiology 151:3453–3453.

Taylor AP, Finnie KS, Bartlett JR, Holden PJ. 2004. Encapsulation of viable

aerobic microorganisms in silica gels. J Sol-Gel Sci Technol 32(1–3):

223–228.

Van Jaarsveld JGS, Van Deventer JSJ, Lorenzen L. 1998. Factors affecting the

immobilization of metals in geopolymerized flyash. Metall Mater Trans

29(1):283–291.

Vijayaraghavan K, Yun YS. 2008. Bacterial biosorbents and biosorption.

Biotechnol Adv 26(3):266–291.

Xu KD, Stewart PS, Xia F, Huang CT, McFeters GA. 1998. Spatial physio-

logical heterogeneity in Pseudomonas aeruginosa biofilm is determined

by oxygen availability. Appl Environ Microbiol 64(10):4035–4039.

Yang HP, Zhu YF. 2005. A high performance glucose biosensor enhanced

via nanosized SiO2. Anal Chim Acta 554(1–2):92–97.

Zamzow MJ, Eichbaum BR, Sandgren KR, Shanks DE. 1990. Removal of

heavy-metals and other cations from waste-water using zeolites. Sep Sci

Technol 25(13–15):1555–1569.

Zuberi M, Liu-Snyder P, ul Haque A, Porterfield D, Borgens R. 2008. Large

naturally-produced electric currents and voltage traverse damaged

mammalian spinal cord. J Biol Eng: 2:17.

2260 Biotechnology and Bioengineering, Vol. 108, No. 10, October, 2011