Articulo Para Microbiologia

Encapsulation of Bacterial Cells in Electrospun Microtubes

S. Klein,† J. Kuhn,‡ R. Avrahami,§ S. Tarre,† M. Beliavski,† M. Green,† and E. Zussman*,§

Faculty of Civil and Environmental Engineering, Faculty of Biology, and Faculty of Mechanical Engineering,Technion, Israel Institute of Technology Haifa 32000, Israel

Received February 8, 2009; Revised Manuscript Received April 21, 2009

Encapsulation of whole microbial cells in microtubes for use in bioremediation of pollutants in water systemswas the main focus of this investigation. Coelectrospinning of a core polymeric solution with bacterial cells anda shell polymer solution using a spinneret with two coaxial capillaries resulted in microtubes with porous walls.The ability of the microtube’s structure to support cell attachment and maintain enzymatic activity and proliferationof the encapsulated microbial cells was examined. The results obtained show that the encapsulated cells maintainsome of their phosphatase, �-galactosidase and denirification activity and are able to respond to conditions thatinduce these activities. This study demonstrates electrospun microtubes are a suitable platform for the immobilizationof intact microbial cells.

1. Introduction

Bioremediation is rapidly becoming an increasingly importanttool in reclaiming polluted water resources with which to satisfythe ever growing global water demand. The technologiesemployed in bioremediation are varied; the most promising ofthese is based on the immobilization of bacterial cells capableof removing specific contaminants. Cell immobilization tech-niques have been investigated extensively for numerous ap-plications including biomedical, agricultural systems, and forthe bioremediation of pollutants in soils and water. Immobiliza-tion of microbial cells has been found to enhance the stabilityof cell enzymatic activities, protect the cells from mechanicalor chemical damage, and sustain large bacterial populations forextended periods.1,2

An example of partial cell immobilization is the controlledgrowth of biofilms on small diameter carrier particles such assand or granulated activated carbon (GAC) found in bioreactorsfor the treatment of wastewater. However, due to the nonsterileconditions typical to these treatment systems, specific micro-organisms, such as the atrazine degrading bacteria Pseudomonassp. strain ADP, grown in such bioreactor systems have sufferedfrom contamination, loss of the ability to degrade atrazine, andeventual washout.3,4

A commonly used whole cell immobilization method that canalleviate the problem of contamination is the entrapment of cellsin polysaccharide gels such as alginates, agarose, k-carrageenan,and other polymeric matrices such as gelatin and polyvinylalcohol.5 Calcium-alginate cross-linking represents an exampleof a simple and cheap technique that is very suitable formaintaining cell viability due to its relatively mild and nontoxiceffects on cells.6 However, this material has a relatively lowmechanical strength and the cells entrapped in the Ca-alginatebeads leak out and grow in the surrounding medium. Also, thenonhomogenous cell distribution within these beads leads tolimitations with regard to diffusion and a reduction in the activesurface area.7,8

In the current work we report a novel method for theimmobilization of whole bacterial cell which can be used inbioremediation. Here, the bacteria are encapsulated withinelectrospun core-shell microtubes. Electrospinning is a com-monly used process for generating ultrafine, polymer-basedfibers with diameters ranging from tens to hundreds ofnanometers.9,10 Electrospun fibers can be produced from avariety of synthetic and natural polymers and are currently beingextensively used in biomedical applications such as scaffoldsand carriers for biologically active molecules such as proteinsand enzymes.11-13 Encapsulation of living cells by electrospin-ning is of considerable technical interest not only for bioreme-diation. While encapsulation of cells in nanofibers has beenpreviously achieved, at present, the fabricated fibers are eithersoluble in water or lose their integrity in aqueous medium.Jaysinghe et al.14 and Townsend et al.15 encapsulated mam-malian cells by using a coaxial needle arrangement with theconcentrated biosuspension in the inner needle and medicalgrade polydimethylsiloxane (PDMS) in the outer needle. Thepost electrospinning cells were found to be viable withoutcellular damage. Salalha et al.16 described the encapsulation ofwhole bacterial cells in electrospun polyvinyl alcohol (PVA)nanofibers. There it was shown that bacterial cells and virusescan survive the electrospinning process and that these retaintheir viability even after storing them dry as mats for a numberof months. Gensheimer et al.,17 on the other hand, foundconsiderable variability in bacterial viability after producingelectrospun polyethylene oxide (PEO) nanofibers.

Although demonstrating their potential for live cell encap-sulation, the existing electrospun fibers are of limited use forbioremediation. To increase the applicability of live cellencapsulation by electrospinning, fibers are needed that areinsoluble in water and are not biodegradable or biodegradablefor a predetermined time period. Furthermore, immobilizationof microbial cells in a confined space such as a solid fiber maynegatively affect their behavior as suggested by the findingsfrom the immobilization of microbial cells in sol-gels.18 Thedevelopment of electrospun hollow polymeric microfibers(microtubes) offers a new method for bacterial cell encapsula-tion.19 In this single-step coelectrospinning process, the poten-tially toxic organic phase consists of a water-insoluble polymer(i.e., the outer shell solution) that is separated from the aqueous

* To whom correspondence should be addressed. E-mail: [email protected].

† Faculty of Civil and Environmental Engineering.‡ Faculty of Biology.§ Faculty of Mechanical Engineering.

Biomacromolecules 2009, 10, 1751–1756 1751

10.1021/bm900168v CCC: $40.75 2009 American Chemical SocietyPublished on Web 05/12/2009

phase (i.e., the core solution). It was shown that proper selectionof the polymers and solvents results in hollow polymericmicrofibers. The fabricated microtubes are potentially beneficialto microbial cells by protecting them while providing space forthe cells to divide and accumulate. This technology has beensuccessfully applied for the encapsulation of pure enzymes.20

In that study, the addition of polyethylene glycol (PEG) to theshell solution altered the shell’s morphology and made it moreporous which thereby positively affected the transfer of smallmolecules into and out of the microtubes. In the presentinvestigation we demonstrate the feasibility of as-spun micro-tubes as a method for the immobilization of whole bacterialcells.21 This encapsulation approach has a potentially wide rangeof applications particularly for use in aqueous environmentswhich includes the biodegradation of hazardous materials,bioremediation of pollutants in water systems, and bioconver-sions in the biotechnology industry.

2. Materials and Methods

2.1. Bacterial Strains and Growth Conditions. Three differentspecies of bacteria were used in these studies. An Escherichia coli strainthat can grow on lactose and is also resistant to the antibiotic kanamycinwas examined when �-galactosidase was measured. Pseudomonas ADPis able to grow with the herbicide atrazine as the sole source of nitrogenand synthesizes an alkaline phosphatase whose activity can be measuredin whole cells. The third strain was from Pseudomonas putida intowhich a gene specifying the red fluorescent protein had been introduced(see Supporting Information for details of its construction) and thisstrain was used in conjunction with confocal microscopy when cellsin microfibers were examined.

In Escherichia coli the synthesis of �-galactosidase, which normallysplits lactose to galactose and glucose, can be repressed by the additionof glucose to the medium while, when glucose is lacking, the enzymeis synthesized in large amounts in the presence of isopropyl-thioga-lactoside (IPTG), a synthetic inducer that is not cleaved by this enzyme.

Pseudomonas ADP was propagated at 30 °C in a MOPS basedmedium containing citrate as the carbon source (see SupportingInformation) with excess orthophosphate to suppress the synthesis ofalkaline phosphatase or with limiting phosphate to induce this enzyme.For denitrification assays, the strain was initially grown aerobicallywith atrazine present and then transferred to anaerobic conditions inwhich nitrite was the sole source of nitrogen.

Pseudomonas putida S12::dsred was grown at 30 °C in the minimalsalts medium mentioned above with the addition of kanamycin (30mg/L) to maintain the added genetic element containing the geneencoding the red fluorescent protein.

Doubly distilled water (DDW) was used throughout.Microscopic images of the distribution of P. putida S12::dsred cells

inside the microtubes were obtained with cells grown to stationary phasewith shaking at 30 °C in 2 mL of the media cited above. Cells wereharvested by centrifugation, the pellets washed twice in DDW and thenresuspended with 5% glycerol to a total volume of 50 µL. This wasmixed with 450 µL of core solution. The bacterial concentration usedfor electrospinning corresponded to an optical density at 600 nm [A600]of about 2 as measured with a UVmini-1240 spectrophotometer(Shimadzu, Kyoto, Japan).

E. coli KT4/RP4 cells were grown to stationary phase by shakingovernight at 37 °C in 10 mL of LB-kanamycin-IPTG medium. Cellswere harvested by centrifugation, the pellets washed with 5 mL of Zbuffer (0.1 M sodium phosphate (pH 7.0), 10 mM KCl, 1 mM MgSO4,5 mM mercaptoethanol) and then resuspended with 1 mL of Z buffer.For electrospinning, 40 µL of these concentrated cells were mixed with360 µL of core solution. The bacterial concentration used for electro-spinning was 3.8 × 109 CFU/mL.

Pseudomonas ADP cells were grown to stationary phase with shakingat 30 °C in 8 mL of MOPS medium and were either starved for

phosphate (0.1 mM of K2HPO4) or with excess phosphate (2 mM ofK2HPO4). Cells were harvested by centrifugation and the resultingpellets were washed twice with water. Cells were resuspended with5% glycerol to a total volume of 50 µL and then mixed with 450 µLof core solution. The bacterial concentration used for electrospinningand free cell suspension assays corresponded to an optical density at600 nm [A600] of about 10.

Pseudomonas ADP cells were grown to stationary phase with shakingat 30 °C in 10 mL of a salt-based medium containing nitrite. Cellswere harvested by centrifugation and the resulting pellets were washedtwice with water. Cells were resuspended with 5% glycerol to a totalvolume of 100 µL and then mixed with 900 µL of core solution. Thebacterial concentration used for electrospinning and free cell suspensionassays corresponded to an optical density at 600 nm [A600] of about12.

2.2. Electrospinning. Preparation of Polymer Solutions. The shellsolution was 9 wt % polycaprolactone (PCL) 80 K + 1 wt %polyethylene glycol (PEG) 6 K dissolved in a mixture of chloroformand dimethylformamide (DMF), 90:10 (w/w); the core solution was 5wt % polyethylene oxide (PEO) 600 K in H2O. All polymers werepurchased from Sigma-Aldrich and were used as is.

Electrospinning Set Up. Core-shell fibers were fabricated by acoelectrospinning process using the set up described by Dror et al.19

All experiments were conducted at room temperature (∼23 °C) and arelative humidity of about 40%. The spinning parameters were asfollows: the electrostatic field used was approximately 0.7 kV/cm andthe distance between the spinneret and collector plate was 14 cm. Theflow rates of both the core and shell solutions were controlled by twosyringe pumps and were 3.5 mL/h for the shell and 0.5 mL/h for thecore. The fibers were collected as a strip on the edge of a verticalrotating wheel22 having a velocity of 1.2 m/sec. For fluorescencemicroscopy, a few fibers were collected directly onto a microscopeslide.

2.3. Measurement of Enzyme Activity. �-Galactosidase Acti-Vity. For the assay encapsulated cells, whole mats were weigheddirectly after electrospinning and cut up into pieces that weigh20-30 mg each. Following several washes in Z buffer, the pieceswhere dipped in 1.0 mL of Z buffer for 60 min. The reaction23 wasstarted by the addition of 0.2 mL of orthonitrophenyl-�-D-galactoside(ONPG) at a concentration of 4 mg/mL. Incubation was at 37 °Cuntil the appearance of the reaction’s yellow product, nitrophenol,became visible and then the reaction was stopped by adding 0.5mL of 1 M Na2CO3. A total of 1 mL of the assay mixture wastransferred to a spectrophotometer cuvette. For the assay of freecells, 20 µL of the same cell suspension prepared for electrospinningwas added to 0.98 mL of Z buffer. The cells were allowed toequilibrate for 30 min at 37 °C before the addition of the substrate.The reaction was conducted in the same manner as that forencapsulated cells except for an additional step after stopping thereaction which removes cells from the assay mixture by centrifuga-tion. Assay mixtures, including blanks (reaction without cells), wereread at 410 nm (Genesys 10 UV, Thermo Scientific, MA, U.S.A.).Enzyme activity was computed in Miller Units (MU; the change inabsorbance of light per cm at 410 nm (A410) multiplied by 1000divided by the measured reaction time in minutes (t)) and eithermultiplied by the dilution used (free cell suspension) or divided bythe fraction of the electrospun mat weight sampled.

Phosphatase ActiVity. For the assay of encapsulated cells, whole matswere weighed directly after electrospinning and then cut up into piecesthat weighed 20-30 mg each. Following several washes in 50 mMTris HCl-Trizma (pH 7.3), the pieces where immersed in 1.4 mL ofthe same Tris buffer for 60 min. To start the reaction, 0.6 mL ofp-nitrophenyl phosphate (7 mg/mL in H2O; Calbiochem Corp.) wasadded. The reaction was performed at 25 °C until the appearance ofthe yellow product, nitrophenol. The reaction was stopped by adding0.5 mL of 1 M Na2CO3. A total of 1 mL of the assay mixture wastransferred to a spectrophotometer cuvette. For the assay of free cells

1752 Biomacromolecules, Vol. 10, No. 7, 2009 Klein et al.

in suspension, the cells were prepared in the same manner as that forcells prepared for electrospinning with the exception that the pelletsafter washing were resupended with 5% glycerol to a total volume of50 µL and then mixed with 0.45 mL 0.85% NaCl. A 100 µL aliquot ofthis free cell suspension was added to 0.6 mL of 50 mM Tris HCl-Trizma (pH 7.3). The cells were allowed to equilibrate for 30 min at25 °C before addition of 0.3 mL substrate. The reaction was conductedin the same manner as for encapsulated cells except for the additionalstep of removing the cells from the assay mixture by centrifugation.Assay mixtures and blanks (reaction without cells) were read at 410nm (Genesys 10 UV, Thermo Scientific, MA, U.S.A.). Enzyme activitywas computed as phosphatase units (arbitrary), which were calculatedin the same manner as that for Miller Units with �-galactosidase activity.

Denitrification ActiVity. For both respiration and induction experi-ments, denitrification activity was measured as the disappearance ofnitrite from the growth medium. For assays of encapsulated cells, wholemats were weighed directly after electrospinning and cut up into piecesthat weigh about 200 mg each. Each piece was further divided intoeight small pieces. Following several washes with water, the pieceswere placed in a 15 mL tube filled to the top with only minimal saltbased medium and left at room temperature for 18 h. At the end ofthat period, the medium was discarded and the pieces were washedtwice using a vortex to wash away any cells that might be on the exteriorof the microtubes. The pieces were then transferred to a small vialfilled with 4 mL of minimal salt base medium supplemented withtrisodium citrate at a concentration of 2 g/L and NaNO2

- at concentra-tion of 25 mg/L as the nitrogen source. The medium was then bubbledfor 40 min with nitrogen gas to remove dissolved oxygen. At the endof the nitrogen bubbling period, the first sample (t ) 0) was taken andthe vial was sealed. Samples were subsequently taken at different timeswhile nitrogen gas was bubbled through the medium to maintain anoxicconditions. Samples (100 µL) were diluted immediately with water andanalyzed for nitrite. Nitrite concentration was determined by absorptionat 543 nm (UVmini-1240, Shimadzu, Kyoto, Japan) using the sulfa-nilamide colorimetric method.24 For suspensions of free cells, assaysof both cells grown under denitrifying or aerobic conditions wereprepared in the same manner as that outlined for electrospinning withthe exception that the pellets, after washing, were resuspended with5% glycerol to a total volume of 100 µL and then mixed with 0.9 mL

0.85% NaCl. The assay was conducted in the same manner as that forencapsulated cells with two exceptions: (1) To keep the same ratio ofbacteria to medium used in the encapsulated cells assay, the assayvolume was 12 mL; (2) Samples (100 µL to 1 mL) were diluted withDDW and filtered with 0.45 µm membrane filters (Millipore) to removethe cells before being analyzed for nitrite.

2.4. Imaging. Images of the fibers were obtained using a Leo Geminihigh resolution scanning electron microscope (HRSEM) at an accelera-tion voltage of 3 kV and a sample to detector distance of 3-5 mm.The specimens were coated with a thin gold film to increase theirconductivity. The distribution of encapsulated P. putida S12::dsred cellsinside the microtubes was visualized by a Zeiss Axiovert 200 Mmicroscope (Gottingen, Germany), using filters for detection of DsRed(excitation/emission: 545 nm ( 25/605 nm ( 70). Images were capturedusing Hdcam (Hamamtsu) 1394 ORCA-ERA and processed withAxioVision LE 4.5 software.

3. Results and Discussion

3.1. Microtubes Morphology and Cell Distributionwithin the Microtubes. Core-shell fibers were fabricated withthe polymeric solutions described in the Experimental Section.The resultant fibers are partially collapsed microtubes withporous walls which are due to PEG in the shell solution. Crosssections of the microtubes and their surface morphology areshown in Figure 1a. As previously shown by Arinstein et al.,25,26

the radial collapse of the microtubes is expected for thiscombination of core-shell solutions. This collapse can occurto such an extent that the microtubes are transformed into aribbon like structures. The tubular space available for theencapsulated bacteria seems rather restricted, averaging adiameter of 3.5 µm.

To observe the location of cells within the electrospun fibers,a Pseudomonas putida strain that was modified to express thered fluorescent protein (DsRed) was encapsulated within themicrotubes. Images of the encapsulated fluorescent cells weretaken after immersing the slides carrying nonwoven microtubesin water for 2 h to wash away any cells that might be on the

Figure 1. (a) HRSEM micrographs of microtubes used for encapsulating living cells; (b,c) fluorescent microscopy images of encapsulated P.putida S12::dsred.

Bacterial Cells in Electrospun Microtubes Biomacromolecules, Vol. 10, No. 7, 2009 1753

exterior of the microtubes. Figure 1b,c shows the location ofencapsulated fluorescent cells within the electrospun microtubes.The encapsulated cells appear to be aligned with and unevenlydistributed along the microtube axis. At the concentration ofcells used (109), the microtubes appear to be sparsely populated.

3.2. Enzymatic Activity of Encapsulated Whole Cells. Theimpact of electrospinning on bacterial cells was initially testedon their enzymatic activity after encapsulation. Two enzymeswere selected, alkaline phosphatase and �-galactosidase. Theassays were conducted on two different bacterial species thathad been encapsulated in the same manner. The results areshown in Table 1. The enzymatic activity of the encapsulatedcells within a sample of fibers was compared to that of freecell suspensions (before electrospinning). To induce phosphatasesynthesis, Pseudomonas ADP cells were grown under phosphatelimiting conditions. Phosphatase activity was indeed found inthe encapsulated Pseudomonas ADP cells grown in this way.However, the microtubes contained only about 10% activity ofthat of the free cells prior to electrospinning. To determinewhether this enzyme diffused out of the fibers as a result ofcell disruption during electrospinning, phosphatase activity wasalso measured outside the nonwoven mats. The assay wasconducted on samples excised from the same nonwoven matsused for the phosphatase activity assay. After an initial rinsingof them, the excised pieces were placed in buffer for 24 h.Samples from the buffer were assayed for enzymatic activityafter 8 and 24 h. As shown in the Table 1, the relative activityfound in the buffer was very low and did not change over time.This is in contrast to recent findings with encapsulated purifiedalkaline phosphatase that showed that there is significant leakagefrom these fibers to the buffer in the first 24 h.20 Results fromthe current phosphatase experiment suggest that the encapsulatedcells maintained their membrane integrity and leakage ofintercellular enzymes is thereby minimal. Dror et al.20 suggestedthat the relatively small size (80 kDa) of alkaline phosphataseaccounted for the massive leakage observed in their experimentsbecause a similar experiment with a larger size enzyme,�-galactosidase (465 kDa), showed only a small amount ofenzyme detected in the buffer.

The second enzymatic activity examined was �-galactosidase.Encapsulated E. coli KT4/RP4 cells that had been grown withisopropyl-thiogalactoside (IPTG) prior to electrospinning wereassayed. In this case, 23% of the enzymatic activity remainedafter electrospinning. The slightly better recovery of theenzymatic activity might be related to a greater durability ofthis bacterial species or of this enzyme during the electrospin-ning process.

The enzymatic activity observed for both enzymes does notdepend on the survival of viable cells after the electrospinningprocess. Even though the electrospinning process negativelyaffected enzymatic activity as compared to free cells, the assaysresults were an encouraging first step because they point to thepotential of electrospinning as a platform for encapsulation

enzymes in whole cells. The low recovery of enzymatic activitysuggests the addition of a “microtubes filling” step followingelectrospinning would be very useful. That is, the cells thatremain viable can be subsequently grown inside the tubes byimmersing them in suitable media. This has been done and thoseexperiments will be the subject of a future communication.

3.3. Encapsulated Cell Viability. Determining cell viabilitydirectly is difficult for encapsulated microbial cells in electro-spun core-shell fibers. Enzymatic activity reflects cellularlybound enzyme(s) that remain active regardless of whether thecells are alive or dead. Due to the insoluble character of themicrotube, a direct assessment of cell viability by the usual plate-counting technique is not possible. The physiological propertiesof encapsulated cells were, therefore, evaluated by cell respira-tion and their ability to synthesize proteins. Both these propertiesrely on the membrane integrity of the cells and employ complexsystems.

Respiration of Encapsulated Whole Cells. Respiration underdenitrifying conditions was measured as nitrite disappearancefrom the growth medium by encapsulated cells and theirequivalent free cells suspensions (Figure 2a). As expected, nitriteutilization by a suspension of free cells began immediately andall external nitrite was consumed within 5 h. Encapsulated cellsexhibited about a 9 h delay before measurable nitrite uptakebegan. Complete nitrite utilization was observed after anadditional period of about 11 h. The different kinetics of theencapsulated cells as compared to the free cell suspension maybe due to (1) mass transfer limitation of nitrite into themicrotubes; (2) impact of the electrospinning process on theencapsulated cells; and (3) adaptation of encapsulated cells totheir new environment. Because previous results showed a largereduction of enzymatic activity due to the electrospinningprocess, it is more likely that the number of physiologicallyactive encapsulated cells was insufficient for detecting nitriteutilization at the beginning of the experiment. A subsequentrecovery time or growth period was required before nitriterespiration was significant enough to be measured.

De NoVo Synthesis of Enzymes in Encapsulated Cells. In thefirst set of experiments, encapsulated cells were exposed tomedium that induces the synthesis of phosphatase. Pieces ofthe mat containing Pseudomonas ADP cells, previously incu-bated in MOPS medium containing an excess of phosphate, wereplaced in growth medium containing a limiting amount oforthophosphate (phosphatase inducing conditions) for 4 days.As expected, phosphatase activity in electrospun mats containingcells that had been grown with a relatively high concentrationof orthophosphate was low and averaged 0.146 ( 0.055 PU/20mg of electrospun mat versus cells grown in low phosphatewhich gave 6.71 ( 2.4 PU/20 mg. After incubating pieces ofthis mat in a growth medium containing a limiting amount oforthophosphate, a high level of phosphatase was observed whichaveraged 248 ( 34 APU/20 mg of electrospun mat. The averageincrease in phosphatase activity was about 1700 times. This

Table 1. Enzymatic Activity of Encapsulated Whole Cells

enzyme species source enzyme activity unitsa relative activityb (%)

phosphatase(s) Pseudomonas ADP electrospun mats 124.1 ( 32.3 9.88 h in bufferd 2.9 ( 0.6 0.2/2.3c

24 h bufferd 3.1 ( 0.3 0.2/2.5c

�-galactosidase E. coli electrospun mats 15153.7 ( 566.5 22.8a Phosphatase activity: arbitrary units (PU) that are given as the change in absorbance at a wavelength of 410 nm (A410) multiplied by 1000 and

divided by the time in minutes. This number is then multiplied by either the dilution used (for free cell suspension) or the weight of the entire mat dividedby the weight of the sampled piece. �-galactosidase activity: measured as Miller Units (MU) that are calculated in the same manner as phosphataseunits. b Free cells versus encapsulated cells. Phosphatase activity (PU) of free cells: 1270.88 ( 109.0. �-galactosidase activity of free cells: 66376.67 (7998.5 MU. c Relative activity of electrospun microbial cells. d The buffer samples measure the leakage of phosphatase from the microtubes.

1754 Biomacromolecules, Vol. 10, No. 7, 2009 Klein et al.

clearly shows that the encapsulated cells within the electrospunmat were able to respond to the surrounding environment andsynthesize phosphatase. In addition, the level of phosphataseactivity reached was statistically significantly higher (p < 0.05)than free cells grown under the same limiting phosphateconditions 16.90 ( 1.3 PU and suggests that the encapsulatedcells in the microtubes had indeed divided. That is, if theencapsulated cells had only synthesized phosphatase withoutgrowing and dividing, then the highest level of phosphatasereached should have been that of the free cells previously grownin low levels of orthophosphate.

It is important to note that the large standard deviationobserved reflects variation in the impact of electrospinning oncells. It appears that large numbers of cells are killed orincapacitated by the process, with the number of physiologicallyactive cells being 1-2 orders of magnitude less than the amountprior to electrospinning. Gensheirmer et al.17 suggested that therapid evaporation upon electrospinning may cause a drasticchange in the osmotic environment leading to a decrease in cellviability. However, evaporation in the water-based core ofmicrotubes was found by Arinstein et al.25 to be very slow, asevidenced by the buckling effect of the microtubes. The lowsurvival rates observed in the microtubes are probably due tothe shear and tension stresses that act on these cells.16

A second set of experiments examined the induction ofdenitrifying activity in encapsulated whole cells. Induction ofdenitrification activity was based on the strain’s ability to growunder either aerobic or denitrifying (anaerobic) conditions whereoxygen or nitrite serve as the terminal electron acceptor,respectively. In general, the main exogenous signals that inducethe synthesis of the denitrification system are a low oxygentension and the presence of an N-oxide that can be used inrespiration.27 As shown in Figure 2b, when aerobically grownfree cells were transferred to denitrifying growth conditions,they exhibited about an 8 h adaption period (lag time) untildenitrifying activity became observable. As previously statedabove, cells already grown under denitrifying conditionsexhibited relatively high denitrifying activity without any lagtime, while the same encapsulated cells had a lag time of 9 h.When encapsulated Pseudomonas ADP cells previously grownunder aerobic growth conditions were transferred to denitrifyinggrowth conditions, a much longer lag time of about 15 h wasrequired before denitrifying activity was initiated. At the veryleast, these results clearly demonstrate that the encapsulated cellsrespond to an environment that induces synthesis of nitritereductase.

4. Conclusions

The present study demonstrates the feasibility of novelelectrospun microtubes as a method to immobilize bacterial cells.The ability of the microtube structure to support cell attachmentand maintain enzymatic activity of the encapsulated microbialcells was examined. This preliminary investigation shows thatthe encapsulated cells maintain some of their phosphatase,�-galactosidase and denitrification activity and are able torespond to inducing conditions in their surrounding environment.The low number of viable cells that remain recovered theirenzymatic activity when the relevant growth conditions wereprovided. The positive results obtained with encapsulatedPseudomonas ADP, a species capable of degrading the herbicideatrazine are encouraging. This is a first step toward thedevelopment cell-bearing microtubes for bioremediation.

Bacterial cell encapsulation combined with the ability tocontrol mass transfer through the microtube shell creates abioreactor like structure. In addition, this fabrication methodenables encapsulation of different kinds of bacterial cells in thesame microtube. This technique has the unique benefit of a largeratio of surface to volume and should find use when separationbetween bacterial cells and the external aqueous environmentis desired (such as in water purification processes). The as-spunmicrotubes can be easily integrated into water purificationsystems and should be able to cover standard biofilm carriersurfaces, such as sand, plastic, and rocks. Microtube technologyalso offers significant advantages for studying the behavior ofcells under confinement. Cells, either bacterial or mammalian,were found to alter their behavior in a confined environment.28-30

The flexibility of the microtube system that permits the controlof permeability, fiber diameter (i.e., degree of confinement), andthe ability to confine and track individual cells should providea platform for this type of research.

Acknowledgment. We wish to thank the RBNI (RussellBerrie Nanotechnology Institute) and the GWRI (Grand WaterResearch Institute) at the Technion for supporting this research.

Supporting Information Available. Bacterial strains, growthconditions, and DNA manipulation. This material is availablefree of charge via the Internet at http://pubs.acs.org.

Figure 2. Nitrite utilization by encapsulated whole cells and free cells.(a) Percentage of nitrite remaining in the growth medium withdenitrifying free cells (0) and encapsulated denitrifying whole cells(b), and (b) percentage of nitrite remaining in the growth mediumwith denitrifying free cells (O); aerobically grown free cells (9) andencapsulated aerobically grown whole cells (4).

Bacterial Cells in Electrospun Microtubes Biomacromolecules, Vol. 10, No. 7, 2009 1755

References and Notes(1) Lee, C. M.; Lu, C. J.; Chuang, M. S. Water Sci. Technol. 1994, 30,

87–90.(2) Adam, W.; Lukacs, Z.; Kahle, C.; Saha-Moller, C. R.; Schreier, P. J.

Mol. Catal. B: Enzym. 2001, 11, 377–385.(3) Katz, I.; Dosoretz, C. G.; Mandelbaum, R. T.; Green, M. Water Res.

2001, 35, 3272–3275.(4) Herzberg, M.; Dosoretz, C. G.; Tarre, S.; Beliavski, M.; Green, M.

Water Sci. Technol. 2004, 49, 215–222.(5) Cohen, Y. Bioresour. Technol. 2001, 77, 257–274.(6) Anisha, G. S.; Prema, P. Bioresour. Technol. 2008, 99, 3325–3330.(7) Kourkoutas, Y.; Bekatorou, A.; Banat, I. M.; Marchant, R.; Koutinas,

A. A. Food Microbiol. 2004, 21, 377–397.(8) Zohar-Perez, C.; Chet, I.; Nussinovitch, A. Biotechnol. Bioeng. 2004,

88, 671–674.(9) Reneker, D. H.; Yarin, A. L.; Zussman, E.; Xu, H. AdV. Appl. Mech.

2007, 41, 43–195.(10) Reneker, D. H.; Fong, H. Polymeric Nanofibers; Oxford University

Press: New York, 2005.(11) Ramakrishna, S.; Fujihara, K.; Teo, W. E.; Yong, T.; Ma, Z. W.;

Ramaseshan, R. Mater. Today 2006, 9, 40–50.(12) Agarwal, S.; Wendorff, J. H.; Greiner, A. Polymer 2008, 49, 5603–

5621.(13) Greiner, A.; Wendorff, J. H.; Yarin, A. L.; Zussman, E. Appl.

Microbiol. Biotechnol. 2006, 71, 387–393.(14) Jayasinghe, S. N.; Irvine, S.; McEwan, J. R. Nanomedicine 2007, 2,

555–567.(15) Townsend-Nicholson, A.; Jayasinghe, S. N. Biomacromolecules 2006,

7, 3364–3369.

(16) Salalha, W.; Kuhn, J.; Dror, Y.; Zussman, E. Nanotechnology 2006,17, 4675–4681.

(17) Gensheimer, M.; Becker, M.; Brandis-Heep, A.; Wendorff, J. H.;Thauer, R. K.; Greiner, A. AdV. Mater. 2007, 19, 2480–2482.

(18) Nassif, N.; Bouvet, O.; Rager, M. N.; Roux, C.; Coradin, T.; Livage,J. Nat. Mater. 2002, 1, 42–44.

(19) Dror, Y.; Salalha, W.; Avrahami, R.; Zussman, E.; Yarin, A. L.;Dersch, R.; Greiner, A.; Wendorff, J. H. Small 2007, 3, 1064–1073.

(20) Dror, Y.; Kuhn, J.; Avrahami, R.; Zussman, E. Macromolecules 2008,41, 4187–4192.

(21) Pieper, D. H.; Reineke, W. Curr. Opin. Biotechnol. 2000, 11, 262–270.

(22) Theron, A.; Zussman, E.; Yarin, A. L. Nanotechnology 2001, 12, 384–390.

(23) Miller, J. H. Experiments in Molecular Genetics; Cold Spring HarborLaboratory: Cold Spring Harbor, NY, 1972.

(24) Standard Methods for the Examination of Water and Wastewater, 19thed.; American Public Health Association/American Water WorksAssociation/Water Environment Federation: Washington, DC, 1995.

(25) Arinstein, A.; Avrahami, R.; Zussman, E. J. Phys. D: Appl. Phys. 2009,42 (1), 015507.

(26) Arinstein, A.; Zussman, E. Phys. ReV. E 2007, 76 (5), 056303.(27) Zumft, W. G. Microbiol. Mol. Biol. ReV. 1997, 61, 533–616.(28) Stamenovic, D.; Ingber, D. E. Soft Matter 2009, 5, 1137–1145.(29) Wang, J. H. C.; Thampatty, B. P. Biomech. Model. Mechanobiol. 2006,

5, 1–16.(30) Nassif, N.; Roux, C.; Coradin, T.; Bouvet, O. M. M.; Livage, J. J.

Mater. Chem. 2004, 14, 2264–2268.

BM900168V

1756 Biomacromolecules, Vol. 10, No. 7, 2009 Klein et al.