Aqueous Two-Phase System-Derived Biofilms for ... - ACS Publications

Jul 13, 2012 - (S.T.) E-mail: [email protected]; Fax: +1-734-936-1905; Tel: +1-734-615-5539. ... Due to the fully aqueous and gentle noncontact print...
0 downloads 4 Views 1MB Size
Download PDF
Article pubs.acs.org/Biomac

Aqueous Two-Phase System-Derived Biofilms for Bacterial Interaction Studies Toshiyuki Yaguchi,†,‡,∥ Mohammed Dwidar,†,∥ Chang Kyu Byun,† Brendan Leung,† Siseon Lee,† Yoon-Kyoung Cho,† Robert J. Mitchell,*,† and Shuichi Takayama*,†,§ †

School of Nano-Bioscience and Chemical Engineering, Ulsan National Institute of Science and Technology, Banyeon-ri 100, Ulsan, 689-798, Republic of Korea ‡ Biomechanics Laboratory, Nagoya Institute of Technology, Gokiso-cho, Showa-ku, Nagoya, 466-8555, Japan § Dept Biomedical Engineering and Macromolecular Science & Engineering Program, University of Michigan, 2200 Bonisteel Boulevard, Ann Arbor, Michigan 48109-2099, United States ABSTRACT: We describe patterning of bacterial biofilms using polymer-based aqueous two-phase system (ATPS) microprinting protocols. The fully aqueous but selectively bacteria-partitioning nature of the ATPS allows spatially distinct localization of suspensions of bacteria such as Pseudomonas aeruginosa and Escherichia coli with high precision. The ATPS patterned bacterial suspensions form spatially distinct biofilms over time. Due to the fully aqueous and gentle noncontact printing procedures employed, coculture biofilms composed of multiple types of bacteria could be printed not only adjacent to each other but also directly over another layer of existing biofilm. In addition, the ATPS environment also allows free diffusion of small molecules between spatially distinct and localized bacterial suspensions and biofilms. This enables biofilms to chemically affect or be affected by neighboring biofilms or planktonic cells, even if they consist of different strains or species. We show that a β-lactamase producing biofilm confers ampicillin resistance to neighboring nonresistant planktonic cells, as seen by a 3,600-fold increase in survival of the ampicillin-sensitive strain. These examples demonstrate the ability of ATPS-based biofilm patterning methods to enable unique studies on commensalistic effects between bacterial species.

■

INTRODUCTION Biofilms play a critical role in function and behavior of bacteria.1 These biofilms often consist of mixed consortia of bacteria that can be affected by other species present within or near the biofilm,2−4 metabolically complement each other,5,6 and communicate with each other through quorum signaling molecules.7 Biofilm bacteria can have different modes of living from the planktonic forms as characterized by gene expression and other characteristics.8 For example, biofilms are usually more resistant to desiccation,9 deleterious environmental conditions including nutrient deprivation,10 and up to 1,000 times more resistant to antimicrobial agents than planktonic cells.11,12 What would accelerate understanding of such beneficial or detrimental bacteria and biofilm interactions is a versatile method to pattern biofilms. While there are a number of methods to pattern biofilms of homogeneous bacterial compositions, typically using substrate surface-directed bacterial patterning methods,14−18 mechanistic studies and analysis of bacterial community interactions in biofilms would benefit from the ability to generate coculture biofilms with distinct interaction geometries between different types of bacteria on versatile substrates including other biofilms. Here, we describe a technique for coculture biofilm patterning using polymer-based aqueous two phase systems (ATPSs) composed of aqueous dextran (DEX) and poly© 2012 American Chemical Society

ethylene glycol (PEG) solutions. The gentle, fully aqueous method allows direct patterning and localization of suspensions of different species or strains of bacteria adjacent to each other without intermixing.19 Furthermore, because of the gentle and noncontact nature of the patterning procedure, the technique allows patterning of bacterial suspensions on top of biofilms formed by another type of bacteria to give multilayered biofilms. In addition to basic characterization of the biofilm patterning technique, we demonstrate the technique’s usefulness through a demonstration of neighbor biofilm-endowed antibiotic resistance. The technique requires no special tools or equipment beyond what is found in a typical microbiology laboratory, making it very practical. We envision these capabilities of ATPS-based biofilm patterning to open new avenues for systematic studies of multispecies bacterial interactions.

■

MATERIALS AND METHODS

Bacterial Strains, Plasmids and Culturing Conditions. The strains used in this study were Escherichia coli strains MG1655 and DH5α, and Pseudomonas aeruginosa KCOM 1465, which was obtained Received: April 4, 2012 Revised: June 27, 2012 Published: July 13, 2012 2655

dx.doi.org/10.1021/bm300500y | Biomacromolecules 2012, 13, 2655−2661

Biomacromolecules

Article

Figure 1. Biofilm patterning using ATPSs. (a) Schematic depiction of patterning bacterial suspension using ATPSs. A DEX-rich phase containing a bacterial suspension was dispensed into a PEG-rich phase solution. Patterned bacterial biofilm was formed within 24 h. (b) Confocal image of a DEX-rich phase droplet containing P. aeruginosa st. KCOM 1465 with pHKT3 bacteria suspension. (c) Bright field image of an unstained patterned biofilm. (d) A CV-stained biofilm. Bar: 0.5 mm. from the Korean Collection for Oral Microbiology (www.chdc-kcom. co.kr). These strains were grown fresh from frozen stocks (20% glycerol/-80 °C) when needed. The E. coli and P. aeruginosa strains were routinely streaked from their stocks on Luria−Bertani (LB) agar plates and incubated at 30 °C for 24 h. A single colony was then inoculated into LB broth and allowed to grow overnight (16 h) before using it as inoculum for ATPS micropatterning. When required, the following antibiotics were added to the LB agar plates and broth; ampicillin (100 μg/mL), kanamycin (50 μg/mL), tetracycline (100 μg/mL). For the P. aeruginosa tests, this strain was transformed with plasmid pHKT3, which encodes for the constitutive expression of a red fluorescent protein.20 Positive clones were selected using tetracycline. Likewise, the E. coli strains were transformed with either pAmCyan (Clontech, USA) or pLacCherry.21 The pAmCyan plasmid encodes for the cyan fluorescent protein (CFP) and confers ampicillin resistance while pLacCherry encodes for the mCherry fluorescent protein and provides resistance to kanamycin. Both strains were always cultured at 30 °C for these tests. Preparation of ATPS Solutions and Bacterial Suspensions. To prepare the ATPS solutions, 14% (w/w) PEG (Mw: 200 000, Sigma-Aldrich, Co.) and 14% (w/w) DEX (Mw: 20 000, Pharmacosmos, Denmark) solutions were first prepared in 50% LB medium and autoclaved separately to sterilize them. After a brief centrifugation to remove any insoluble components, both solutions were mixed equally and allowed to separate into two phases, by centrifuging them (3000 X g, 60 min) at 25 °C, with the PEG-rich phase on top and the DEX-rich phase on the bottom. After separating into the two phases, the DEX-rich phase was then used to prepare the bacterial suspensions. The overnight culture optical density (OD) was measured using a Biophotometer (Eppendorf). A sample of the overnight bacterial culture was then centrifuged (3000 X g, 5 min), the culturing medium was discarded and the remaining bacterial pellet was resuspended in the DEX-rich

phase solution to give the required densities of bacterial cells, i.e., the reconstituted OD (rOD). Complete removal of the supernatant was necessary to ensure that the downstream ATPS solution compositions were well-defined. Unless mentioned otherwise, the rOD was adjusted to 0.05, a value commonly used in biofilm studies.22,23 Bacterial patterning and biofilm formation using ATPS. DEX-rich phase solutions containing bacterial cells were patterned as described previously.19 Briefly, the volume of droplets was 0.2−0.8 μL, and they were patterned into PEG-rich phases using a conventional manual pipet (Eppendorf) dispensing the droplets at desired positions on culture plates coated with polydimethylsiloxane (PDMS). The plates were coated by spin coating 0.5 g of freshly prepared PDMS within a 35 mm Petri-dish. The dish was then cured at 65 °C for at least 12 h. For forming biofilms, particularly the E. coli biofilms, we tested conventional polystyrene dish surfaces against the same type of dish coated with PDMS. The evaluation clearly revealed that the PDMS surfaces resulted in more prominent biofilms (data not shown). After patterning the bacteria droplets, the plates were kept stationary at room temperature with ∼100% humidification for 6 to 24 h. The resulting biofilms were observed after staining with 0.1% crystal violet (CV) for 20 min. Bright field images with or without CV staining were obtained using an SZX16 stereoscope connected to a DP72 CCD camera operated by the DP2-BSW imaging software (Olympus). The diameters of all the patterned biofilms were measured using the DP2BSW analysis software, with at least five independent samples used for each measurement. To visualize the extracellular DNA present within the biofilm, a 0.3 ul drop of DH5α /pAmCyan in DEX phase was spotted inside a 3.5 mm cover-glass bottom Petri dish (SPL) filled with 2.5 mL PEG-rich phase solution as usual. After 24 h, the PEG-rich solution was removed, and the ATPS-derived biofilm was rinsed gently with sterile distilled water. Two microliters of a 2 mM ethidium homodimer-1 (EthD-1) stock solution (Life Technologies, USA) was added to 1 mL of sterile phosphate buffered saline (PBS, pH 7.2), vortexed, and then 2656

dx.doi.org/10.1021/bm300500y | Biomacromolecules 2012, 13, 2655−2661

Biomacromolecules

Article

added to the dish. This was stored in the dark for 30 min. The dish was washed again gently, after which the biofilm and its extracellular DNA were observed using an LSM700 confocal microscope (Carl Zeiss) operated by ZEN 2009 software. Bacterial Biofilm-on-Biofilm Formation. E. coli str. MG1655 carrying pAmCyan was inoculated first in 50% reduced LB medium at an OD600 of 0.05 and cultured in PDMS coated 35 mm Petri dish for 24 h. Afterward, the medium was discarded, and the plate was rinsed with sterile distilled water to remove any planktonic cells. The plate was then filled with 2.5 mL PEG-rich phase solution, and droplets of E. coli str. DH5α with pHKT3, in DEX-rich phase solution, were patterned on top of the CFP-expressing biofilm using the techniques previously described, and the plate was left for 24 h at room temperature once more. After that, the ATPS solutions and the planktonic cells were discarded gently while keeping the fragile layered biofilms as follows: the dish was placed upside-down in water preheated to 37 °C for 5 to 10 min. This allowed the ATPS solutions to drain away by gravity without damaging the layered biofilms. The biofilms were imaged using an LSM700 confocal microscope (Carl Zeiss) operated by ZEN 2009 software. Rescue of an Ampicillin-Sensitive Strain in the Presence of Ampicillin When Patterned over a β-Lactamase Producing Biofilm. To test the ability of an ampicillin-sensitive strain to survive in the presence of ampicillin when located on a biofilm made by a βlactamase producing strain, E. coli str. MG1655 was transformed with pLacCherry,24 which confers a cherry red fluorescence and only kanamycin resistance. This pLacCherry strain was used as the ampicillin sensitive strain. E. coli str. MG1655/pAmCyan was used to make a β-lactamase-producing patterned biofilm using a modified version of the protocol described above. Three 0.3 μL DEX-rich drops of E. coli pAmCyan were first patterned in a 35 mm Petri dish containing 2.5 mL of PEG and allowed to grow for 24 h and form biofilms. Without washing or removal of the ATPS solutions, 0.3 μL drops of E. coli str. MG1655/ pLacCherry were spotted on the preformed biofilms as well as separately in the same plate, approximately 2 cm away from the E. coli pAmCyan biofilms, giving six E. coli pLacCherry spots in total. After half an hour, a 0.5 μL drop of ampicillin (2,000 μg/mL) was spotted on each one of these six drops: three with the E. coli pAmCyan biofilms, and three of E. coli pLacCherry planktonic cells alone. The ampicillin stock solution was prepared in water and diluted into the DEX-rich solution to give the working concentration. The plates were then incubated at room temperature for 24 h longer, and images for these drops were taken using confocal microscopy as described in the previous section. Next, each spot was aspirated independently into centrifuge tubes containing 1 mL of sterile PBS, serially diluted and then spread out on agar plates containing either kanamycin (50 μg/ mL) to enumerate the surviving E. coli pLacCherry cells, ampicillin (100 μg/mL) for the E. coli pAmCyan cells and both kanamycin and ampicillin to test for the transfer of plasmids as based upon a recent finding that bacterial biofilms form nanotubes between the bacteria, allowing plasmids and cytoplasmic contents to be transferred between them.25

Figure 2. Effect of cell density and droplet size on biofilm formation in ATPS droplets. (a) Patterned P. aeruginosa st. KCOM 1465 biofilms of varying reconstituted OD evaluated by CV staining. The ranges of reconstituted ODs were 0.5 to 13 and culture times were 6/12/24 h at room temperature. Representative images are shown from four replicates. Staining of biofilms formed at OD 5 and above at 6 h was variable, but no staining was observed for any replicates with the two lowest ODs. (b) The area of the patterned P. aeruginosa st. KCOM 1465 biofilms increases with increasing volumes of bacterial droplets dispensed. Biofilm areas were determined from microscope images of CV stained biofilms using the DP2-BSW image analysis software.

solutions and planktonic cells were removed to reveal circular biofilms that were easily visualized by both direct phase contrast microscopy as well as by 0.1% CV staining (Figure 1c,d). We observed that use of PDMS substrates enhanced biofilm formation over use of non PDMS-coated culture dishes (data not shown). Many researchers use low initial ODs, i.e., an OD of approximately 0.05 at 600 nm, when growing biofilms within various formats, such as 96-well plates and flow cells.13,22,23 Since the ATPS environment is quite different from other conventional methods, the effects of the initial ODs and residence times on ATPS-based biofilm formation were consequently investigated. For this, P. aeruginosa KCOM 1465 cells were spun down, concentrated and suspended in the DEX-rich solution at several different rODs. As shown in Figure 2a, P. aeruginosa KCOM 1465 biofilms needed 6 h for them to be seen with CV staining when the initial rOD was over 5. No biofilm formation was obvious for the two lowest rODs tested at 6 h. All of the cell concentrations, however, led to visible biofilms after 12 h and fully formed and defined

■

RESULTS ATPS-Generated Biofilms. In a previous research article, we demonstrated that it was possible to generate and maintain micropatterns of suspensions of multiple bacterial strains within polymer-based ATPSs.19 The duration used for culturing, however, was short (up to 8 h), and conditions were not conducive to the formation of biofilms. Using a similar protocol but with extended culture times of up to 1 day, biofilms were formed with bacterial suspensions in DEX-rich phase droplets as little as 0.3 μL (Figures 1c, 1d, and 2a). A confocal image of the patterned DEX-rich phase droplet containing RFP expressing P. aeruginosa KCOM 1465 is shown in Figure 1b and illustrates that the bacterial suspension is confined within the DEX-rich drop. After incubation for 24 h, the ATPS 2657

dx.doi.org/10.1021/bm300500y | Biomacromolecules 2012, 13, 2655−2661

Biomacromolecules

Article

Figure 3. Staining of the extracellular DNA present in an ATPS-derived E. coli biofilm. Images showing the 3D structure of a CFP-expressing E. coli strains DH5α/pAmCyan biofilm. The diameter of the biofilm was approximately 1 mm. The panels on the right show the split and overlay fluorescence signals obtained from CFP (green) and EthD-1 (red), which binds to the extracellular DNA.

biofilms at 24 h. In fact, all of the biofilms were visually identical regardless of the initial rOD when a 24 h incubation protocol was implemented. It was also found that the size of the biofilm formed is proportional to the volume of the DEX-rich cell suspension droplet used (Figure 2b), with larger volumes giving proportionately larger areas of biofilms. Similar biofilms could also be constructed using an E. coli culture (Figure 3). Furthermore, staining of these ATPSderived biofilms with EthD-1 showed that they include extracellular DNA (Figure 3), which is used as a structural support during the maturation of biofilms. The presence of this extracellular DNA implies that these ATPS-derived populations are biofilms and not just bacterial cell mats or aggregates. Generation of a Biofilm on a Biofilm. The gentle, fully aqueous, and noncontact nature of ATPS-based bacteria patterning also allows patterning of biofilms on top of another preformed biofilm. Here, the E. coli strains MG1655 and DH5α were selected to show versatility of the patterning technology. To distinguish between the two E. coli strains, the DH5α strain was engineered to express red fluorescence (pHKT3)20 and MG1655 to express cyan fluorescence (pAmCyan, Clontech, USA). The protocol developed for forming a biofilm-on-abiofilm is illustrated in Figure 4. Initially a biofilm consisting entirely of CFP-expressing E. coli strain MG1655 was formed on a PDMS-covered dish. After washing the planktonic and loosely attached bacterial cells away, the biofilm was covered with a PEG-rich solution and a DEX-rich phase containing an RFP-expressing E. coli strain DH5α culture was spotted on top. After 24 h, the biofilms were washed and visualized using a confocal microscopy (Figure 4). The secondary patterned biofilms, i.e., the RFP-expressing E. coli strain DH5α, are clearly visible and have defined edges, demonstrating that the bacteria were confined to the DEX phase even when spotted on top of another biofilm. The z-stack image of the RFP biofilm edge confirms that the RFP-encoding E. coli cells are in fact on top of the CFP-expressing cells and did not merely dislodge them, evidence that a biofilm, at least in

its early stages, can be generated on another biofilm using this technique. This is, to the best of our knowledge, the first reported study where biofilms have been formed in distinct user-defined patterns on another pre-existing biofilm. Ampicillin Resistance Is Conferred onto Planktonic Cells by Biofilms. The ability of a biofilm to confer benefits to other bacteria present within close proximity was tested. It is known that for β-lactamase producing bacterial strains, such as E. coli strain MG1655/pAmCyan, this enzyme is transported into the periplasmic membrane where it deactivates the antibiotic.26,27 It was hypothesized, therefore, that this strain could confer ampicillin resistance to neighboring E. coli strain MG1655/pLacCherry cells, which is kanamycin resistant but susceptible to ampicillin. Initially a spot biofilm of E. coli strain MG1655/pAmCyan was prepared using the ATPS protocols listed above. After 24 h, another DEX-rich droplet containing E. coli strain MG1655 carrying pLacCherry was spotted on top of the original droplet, which was followed 30 min later by the addition of a third DEX-rich droplet containing 2,000 μg/mL ampicillin. In parallel, within the same plate, a separate E. coli pLacCherry droplet was spotted alone, i.e., without the E. coli pAmCyan biofilm, and exposed to ampicillin in an identical manner. As a control, tests were also performed where the E. coli pLacCherry droplet was spotted on a wild-type, i.e., no plasmid or resistance, E. coli MG1655 biofilm. The number of surviving E. coli pAmCyan and E. coli pLacCherry cells for each condition is shown in Figure 5. It is clear that the addition of ampicillin had little effect on the survival of the E. coli pAmCyan cells, while the number of E. coli pLacCherry CFU decreased dramatically when present alone and exposed to ampicillin, with less than 1 in 175 000 MG1655/pLacCherry surviving after a 24 h incubation. When spotted on top of a wild-type MG1655 culture, the number surviving the addition of ampicillin was somewhat higher compared to when present alone, but still led to a significant loss in viability. 2658

dx.doi.org/10.1021/bm300500y | Biomacromolecules 2012, 13, 2655−2661

Biomacromolecules

Article

Figure 4. Layered patterning of secondary bacterial biofilms over a primary biofilm. (a) An E. coli MG1655/pAMCyan biofilm was allowed to form across a dish in conventional LB medium. (b) This represents the resulting primary biofilm. (c) The conventional LB medium was replaced with a PEG-containing LB medium and a suspension of E. coli DH5α/pHKT3 in LB medium containing DEX was patterned on top of the primary biofilm. (d) After incubation, the secondary biofilm forms in distinct patterns over the primary biofilm. (e) Confocal microscope images of a 5 × 4 array of circular E. coli DH5α/pHKT3 biofilms patterned over E. coli MG1655/pAMCyan biofilm blanket. (Left) Bar: 1 mm. (f) A magnification of the yellow rectangle shown in panel e. (g) A magnification of the yellow rectangle show in panel f. The 3D structure including the layered and patterned nature of the bacterial biofilms can be observed.

By contrast, when the E. coli pLacCherry culture was spotted on top of the E. coli pAmCyan biofilm, the number of E. coli pLacCherry surviving in presence of ampicillin was 3600-fold higher than when they were present alone and only represented a 7-fold loss in colony forming unit (CFU) numbers when compared to a control with no antibiotic. These results plainly demonstrate the commensal effects when the planktonic MG1655/pLacCherry cells were located within close proximity to a β-lactamase-producing biofilm. Although the number of E. coli pLacCherry within the antibiotic-free tests when present alone was roughly 7-fold higher when compared to that spotted on the biofilm, the higher CFU numbers are likely the result of less competition for nutrient resources, allowing the E. coli pLacCherry culture to grow further during the 24 h incubation. The confocal images additionally showed that the E. coli pLacCherry cells in the control spot were normal size and shape while those exposed to ampicillin were elongated and filamentous (Figure 5 inset). This phenotypic change was reported before and was seen only when the ampicillin concentration exceeded the minimal inhibitory concentration.28 Although numerous such cells were observed under the microscope, the number of viable cells was very low (Figure

5), suggesting that these many of the cells did not survive the antibiotic treatment but were still fluorescing.

■

DISCUSSION

This study presents several novel protocols and findings regarding bacterial biofilm patterning. Initially, we characterized the use of ATPS techniques to develop bacterial biofilm spots. This technique is useful in that it does not need special tools, e.g. cell flow chamber system22 or microfluidic devices,29 while also allowing researchers to generate multiple biofilms within a single system, something that is difficult or not possible within the conventional methods. Furthermore, ATPS droplet patterned bacteria can be spatially separated but chemically interconnected, as small molecules can easily diffuse between the DEX and PEG-rich phases, allowing the bacteria to potentially communicate.19 The polymers at the molecular weights and weight/volume concentrations used (3.5 mOsm of the 7% of 20 000 Mw DEX compared to 170 mOsm of the sodium chloride content alone) do not significantly alter the osmolarity of the cell culture solutions and a wide range of DEX and PEG solutions with different polymer concentrations 2659

dx.doi.org/10.1021/bm300500y | Biomacromolecules 2012, 13, 2655−2661

Biomacromolecules

Article

Finally, we show an application of the ATPS-derived biofilms to evaluate the propensity of a biofilm to confer antibiotic resistance to nonresistant planktonic cells. To determine this, the number of MG1655/pLacCherry surviving after an exposure to high doses of ampicillin when present alone or spotted on a biofilm consisting of wild type MG1655 or MG1655/pAmCyan, which expresses β-lactamase, were measured. Our results illustrate that a close association between the sensitive and resistant bacterial species significantly increased the viability of the nonresistant strain under antibiotic treatment conditions. These results clearly demonstrate a commensalistic benefit conferred to these bacteria as nonampicillin-resistant MG1655/pLacCherry cells spotted on the ampicillin resistant biofilm had a 3600-fold increased survival than when plated alone.

■

CONCLUSIONS This study presents research at the crossroads of the biofilm and micropatterning fields, using the unique properties of ATPS systems20,30 to evaluate biofilm formation processes. Biofilms were rapidly and consistently made using an ATPS system without the need for complex or expensive equipment. A wide range of initial bacterial suspension concentrations could be used to form biofilms. The resulting areas or diameters of the biofilms were also reasonably controlled although less defined compared to methods using precision microfabrication methods. For example, our method generally generates circular biofilms whereas other microfabrication methods can generate other shapes. Although not extensively analyzed, the extent or thickness of biofilm formation appears somewhat controllable depending on duration of biofilm formation although initial bacteria density seemed to have minimal effects except at the very early stages of biofilm formation. We also generated biofilms consisting of two different strains, with one strain patterned on top of another. This study represents the first report of a user-defined patterned biofilm-on-another-biofilm being generated. As an application of multiple bacteria type interactions, the ability of a biofilm to confer ampicillin resistance to planktonic cells was demonstrated, with a 3,600fold increase in survival of the sensitive strain when spotted onto a biofilm formed by a resistant strain. This technique can be applied to generate biofilms consisting of diverse bacterial strains within a single Petri dish, allowing researchers to study intercellular communication and behavior, such as mutualism and commensalism.

Figure 5. An antibiotic resistant biofilm helps a nonresistant planktonic neighbor. CFUs obtained for each strain or coculture with and without ampicillin treatment. E. coli str. MG1655/ pLacCherry survival is very low when cultured alone or on a biofilm formed with an ampicillin sensitive strain (MG1655) but has significantly increased survival when cultured on top of a biofilm consisting of resistant bacteria. No dual resistant (AmpR, KanR) colonies were isolated. The inset shows the abnormally elongated and filamentous morphology observed for E. coli str. MG1655/pLacCherry cells after exposure to ampicillin when spotted alone, a phenomenon that was reported previously.28

have been shown previously to be compatible, as measured by proliferation rates, with bacterial culture and growth.19 The average area of a biofilm formed using a 0.2 μL droplet was less than 0.5 mm2, corresponding to an average diameter of 0.75 mm (Figure 2b). This is significantly larger than what the Weibel’s group’s 0.1 mm diameter biofilms prepared using a jig and stencil platform.14 The size, resolution, and patterning precision of the droplets were limited by specifications of the pipetter and the manual dexterity of the experimenter. This size difference, however, is not an indication of intrinsic limits of resolution of the ATPS-based biofilm patterning method but rather an experimental design consideration for creating robust assay readouts and for demonstrating what can be done simply by manual use of pipettes that are common in any biology laboratory. The results in Figures 2 and 3 also present several other interesting findings and hint at some of the differences between biofilms in nature, conventional biofilm patterning methods, and ATPS. The clear circular shapes of the biofilms in Figure 2a demonstrate that the bacteria were confined to the DEX-phase. While such distinct patterns and shapes of biofilms may not necessarily resemble the extensively interspersed and comingling structures of biofilms commonly observed in nature, the ability to generate defined shapes are envisioned to be useful for systematic and quantitative mechanistic studies where reproducibility with adjustability are key. Another difference with some naturally formed biofilms is the lack of fluid flow in our experimental conditions. Static conditions, however, also do exist in some natural environments and are experimentally convenient. The ATPS-based biofilms can be patterned individually and even on a previously developed biofilm. The biofilm patterning capability opens many research possibilities, including studies on both intercellular communication and commensalistic/mutualistic/detrimental relationships between different bacterial species.

■

AUTHOR INFORMATION

Corresponding Author

*(R.J.M.) E-mail: [email protected]; Fax: +82-52-217-2509; Tel: +82-52-217-2313. (S.T.) E-mail: [email protected]; Fax: +1-734-936-1905; Tel: +1-734-615-5539. Author Contributions ∥

These authors contributed equally.

Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This work was supported by the WCU (World Class University) program (No. R322008000200540) through the National Research Foundation of Korea (NRF) as funded by the Ministry of Education, Science and Technology (MEST). M.D., S.L., and R.J.M. are grateful for the support from the 2660

dx.doi.org/10.1021/bm300500y | Biomacromolecules 2012, 13, 2655−2661

Biomacromolecules

Article

Korea-Israeli Joint Fund Program and by the Basic Science Research Program (Nos. K21001001804-10B1200-21610, 2010-0015377 and 2010-0028073) all of which are also supported by NRF as funded by MEST.

■

REFERENCES

(1) Domka, J.; Lee, J.; Wood, T. K. Appl. Environ. Microb. 2006, 72 (4), 2449−2459. (2) McEldowney, S.; Fletcher, M. . Arch. Microbiol. 1987, 148 (1), 57−62. (3) Ciardi, J. E.; McCray, G. F.; Kolenbrander, P. E.; Lau, A. Infect. Immun. 1987, 55 (6), 1441−6. (4) Lamont, R. J.; Rosan, B. Infect. Immun. 1990, 58 (6), 1738−43. (5) Nielsen, A. T.; Tolker-Nielsen, T.; Barken, K. B.; Molin, S. . Environ. Microbiol. 2000, 2 (1), 59−68. (6) Kim, H. J.; Boedicker, J. Q.; Choi, J. W.; Ismagilov, R. F. Proc. Natl. Acad. Sci. U.S.A. 2008, 105 (47), 18188−93. (7) Lee, J. H.; Lee, J. FEMS Microbiol. Rev. 2010, 34 (4), 426−444. (8) Wood, T. K. Environ. Microbiol. 2009, 11 (1), 1−15. (9) Flemming, H. C.; Wingender, J. Nat. Rev. Microbiol. 2010, 8 (9), 623−633. (10) Oh, Y. J.; Jo, W.; Yang, Y.; Park, S. Ultramicroscopy 2007, 107 (10−11), 869−874. (11) Stewart, P. S.; Costerton, J. W. Lancet 2001, 358 (9276), 135− 138. (12) Ceri, H.; Olson, M. E.; Stremick, C.; Read, R. R.; Morck, D.; Buret, A. J. Clin. Microbiol. 1999, 37 (6), 1771−6. (13) Kadouri, D.; O’Toole, G. A. Appl. Environ. Microbiol. 2005, 71 (7), 4044−4051. (14) Eun, Y. J.; Weibel, D. B. Langmuir 2009, 25 (8), 4643−54. (15) Weibel, D. B.; Lee, A.; Mayer, M.; Brady, S. F.; Bruzewicz, D.; Yang, J.; DiLuzio, W. R.; Clardy, J.; Whitesides, G. M. Langmuir 2005, 21 (14), 6436−6442. (16) Kappell, G. M.; Grover, J. P.; Chrzanowski, T. H. Electron. J. Biotechnol. 2009, 12 (3), No. DOI: 10.2225/vol12-issue3-fulltext-8. (17) Bos, R.; van der Mei, H. C.; Gold, J.; Busscher, H. J. . FEMS Microbiol. Lett 2000, 189 (2), 311−315. (18) Hou, S.; Burton, E. A.; Wu, R. L.; Luk, Y. Y.; Ren, D. Chem. Commun. 2009, 10, 1207−9. (19) Yaguchi, T.; Lee, S.; Choi, W. S.; Kim, D.; Kim, T.; Mitchell, R. J.; Takayama, S. Analyst 2010, 135 (11), 2848−52. (20) Tomlin, K. L.; Clark, S. R. D.; Ceri, H. J. Microbiol. Methods 2004, 57 (1), 95−106. (21) Elliott, N.; Lee, T.; You, L.; Yuan, F. Integr. Biol. 2011, 3 (6), 696−705. (22) Wood, T. K.; Gonzalez Barrios, A. F.; Herzberg, M.; Lee, J. Appl. Microbiol. Biotechnol. 2006, 72 (2), 361−7. (23) Lee, J. T.; Jayaraman, A.; Wood, T. K. BMC Microbiol. 2007, 7, 42. (24) Elliott, N.; Lee, T.; You, L. C.; Yuan, F. Integr. Biol. 2011, 3 (6), 696−705. (25) Dubey, G. P.; Ben-Yehuda, S. Cell 2011, 144 (4), 590−600. (26) Suvorov, M.; Vakulenko, S. B.; Mobashery, S. Antimicrob. Agents Chemother. 2007, 51 (8), 2937−42. (27) Koshland, D.; Botstein, D. Cell 1980, 20 (3), 749−60. (28) Magnussen, C. R.; Hruska, J. F. J. Clin. Microbiol. 1980, 12 (5), 690−4. (29) Groisman, A.; Lobo, C.; Cho, H.; Campbell, J. K.; Dufour, Y. S.; Stevens, A. M.; Levchenko, A. Nat. Methods 2005, 2 (9), 685−9. (30) Tavana, H.; Jovic, A.; Mosadegh, B.; Lee, Q. Y.; Liu, X.; Luker, K. E.; Luker , G. D.; Weiss, S. J.; Takayama, S. Nat. Mater. 2009, 8, 736−741.

2661

dx.doi.org/10.1021/bm300500y | Biomacromolecules 2012, 13, 2655−2661