Electron beam fabrication of a microfluidic device for studying submicron-scale bacteria
- M Charl Moolman†1,
- Zhuangxiong Huang†1,
- Sriram Tiruvadi Krishnan1,
- Jacob WJ Kerssemakers1 and
- Nynke H Dekker1Email author
© Moolman et al.; licensee Springer. 2013
Received: 11 February 2013
Accepted: 3 April 2013
Published: 10 April 2013
Controlled restriction of cellular movement using microfluidics allows one to study individual cells to gain insight into aspects of their physiology and behaviour. For example, the use of micron-sized growth channels that confine individual Escherichia coli has yielded novel insights into cell growth and death. To extend this approach to other species of bacteria, many of whom have dimensions in the sub-micron range, or to a larger range of growth conditions, a readily-fabricated device containing sub-micron features is required.
Here we detail the fabrication of a versatile device with growth channels whose widths range from 0.3 μ m to 0.8 μ m. The device is fabricated using electron beam lithography, which provides excellent control over the shape and size of different growth channels and facilitates the rapid-prototyping of new designs. Features are successfully transferred first into silicon, and subsequently into the polydimethylsiloxane that forms the basis of the working microfluidic device. We demonstrate that the growth of sub-micron scale bacteria such as Lactococcus lactis or Escherichia coli cultured in minimal medium can be followed in such a device over several generations.
We have presented a detailed protocol based on electron beam fabrication together with specific dry etching procedures for the fabrication of a microfluidic device suited to study submicron-sized bacteria. We have demonstrated that both Gram-positive and Gram-negative bacteria can be successfully loaded and imaged over a number of generations in this device. Similar devices could potentially be used to study other submicron-sized organisms under conditions in which the height and shape of the growth channels are crucial to the experimental design.
KeywordsMicrofluidics Poly(dimethylsiloxane) Electron beam lithography Micro-nanofabrication Bacterial growth Lactococcus lactis Escherichia coli
The use of microfluidics in biological research has gained much popularity in recent years. Subfields that have been impacted by this technology range from tissue engineering , cancer stem cell research , gene expression of embryonic stem cells , protein interactions , diagnostic medicine  as well as microbial physiology and behaviour [6–8], to name but a few. A specific contribution to the field of microbiology is the ability to observe and manipulate single cells . Individual cells can significantly differ from one another in terms of their biochemistry and genetics . The ability to observe individual cells under controlled conditions provides one with the ability to investigate the individual functioning of cells as well as their mutual behaviour [11–17]. For example, the use of microfluidics has facilitated the study of molecular behaviour inside individual cells, as demonstrated by e.g. Taniguchi et al. in their study of protein and mRNA expression at the single-molecule level inside individual living cells. An additional advantage of microfluidics is that it provides one with the ability to observe many more generations than with conventional agarose pads .
Recently Wang et al. utilized a microfluidic device to quantitatively study steady-state growth and division of individual Escherichia coli (E. coli) cells at a defined reproductive age grown in Luria-Bertani (LB) medium. Such a device makes it possible to study a large number of cells that inherit the same cell pole over multiple generations. In their design, cells are confined in growth channels oriented perpendicularly to a trench through which growth medium (LB) is flown. The width and height of the channels are similar to the dimensions of E. coli, which has a diameter of ca. 1 μ m and a length of ca. 2.5 μ m under these conditions [20, 21]. Cells are immobilized, in the absence of chemical fixation, at the far end of such a growth channel (ca. 25 μ m in length). The length of the growth channels is chosen so as to ensure sufficient supply of nutrients to the bacteria by diffusion. Such an immobilization scheme allows one to simultaneously study numerous different cells for extensive periods of time.
A microfluidic device that would allow one to probe smaller microorganisms would greatly enhance the applicability of this approach. Notably, many bacterial species have submicron-scale dimensions for which growth channels would require significantly reduced widths. Examples of such species include e.g. Mycoplasma (diameter 0.2−0.4 μ m ), Prochlorococcus (diameter 0.5−0.7 μ m ), and Lactococcus lactis (diameter of ca. 0.75−0.95 μ m ), for which growth channels would require significantly reduced widths. A single device with growth channels of variable widths would furthermore provide maximal flexibility for studying different types of bacteria under a variety of growth conditions. A recent advance along these lines described the fabrication of sub-micron channels in agarose . However, both this approach as well as the device utilized by Wang et al. are fabricated using conventional photolithography. While this is a widely available and convenient technique, for the fabrication of devices with smaller dimensions it becomes more cumbersome and alternative approaches such as electron beam lithography (EBL)  become more suitable. EBL can readily fabricate smaller features (ca. 20 nm in lateral dimensions) compared to conventional photolithography (ca. 1 μ m) , while simultaneously affording greater control of the structure size and shape. An additional advantage of EBL is the reduced time from design to final device, which is convenient in a research environment where it is frequently required to change and improve a device on a relatively short time scale. The structural control and rapid-prototyping needs are thus more easily met by EBL than by conventional photolithography.
Results and discussion
The first pattern, i.e. the small growth channels, can now be written into the wafer (Figure 3, Step 4). We make use of a Leica EBPG 5000+ (acceleration voltage 100 kV, aperture 400 μ m) to write the pattern on the wafer. Here we use a spot size of ca. 25 nm and a current of ca. 46 nA. We choose the beam step size (BSS) to be 20 nm and the dose 1400 μ C/cm2.
Following electron beam exposure, we develop the exposed PMMA (Figure 3, Step 5) by using methyl isobutyl ketone (MIBK) and isopropyl alcohol (IPA). We place the wafer in a beaker containing a 3:1 ratio of IPA and MIBK for 60 s. Directly afterwards, we place the wafer in a beaker containing IPA only for 30 s, and subsequently spin it dry. We then clean the wafer by exposing the wafer to an O2 plasma in a microwave plasma system (Tepla 100) with the power set to 100 W and the pressure maintained at approximately 0.15 mbar.
Following PMMA development, we perform the dry etching of the growth channels (Figure 3, Step 6). This is achieved by using an inductive coupled plasma (ICP) reactive-ion etcher (RIE) (Adixen AMS 100 I-speeder) with a mixture of 15 sccm sulfur hexafluoride(SF6), 20 sccm octafluorocyclobutane(C4F8), 10 sccm methane(CH4) that is diluted in 100 sccm helium (He). We set the ICP power to 2000 W, and the capacitive coupled plasma (CCP) power (biased power) to 250 W. We maintain the sample holder at 0°C during the entire process. The wall of the main chamber is maintained at 200°C to inhibit polymer deposition there. We fix the sample holder height at 200 mm and maintain a low pressure of ca. 1 Pa. We perform this etching process for 3 min. At an etching rate of ca. 390 nm/min, this results in approximately 1.2 μ m deep growth channels.
The final step we perform before the wafer can be used as a mold is a silanization step. This is necessary to reduce the adhesion between PDMS and Si in the curing step and is achieved as follows. We expose the wafer to an O2 plasma for 10 s. We then immediately place the wafer in a desiccator together with 15 μ L of silanizing agent (tridecafluoro-1,1,2,2-tetrahydrooctyl trichlorosilane) (TFOCS) . We place the desiccator under a vacuum, which results in evaporation of the silanizing agent and formation of a monolayer on the surface of the Si wafer. This layer renders the Si wafer extremely hydrophobic, preventing the PDMS from adhering to it. After 2 hours under vacuum, the Si wafer is ready to be used as a mold.
In order to fabricate the structures in PDMS we perform the following steps. If the Si wafer is stored for longer than 24 h, we first ultrasonically clean it in 100% HNO3 for 15 min, rinse with DI water and spin it dry (Figure 6, Step 1). Secondly, we prepare PDMS (Mavom Chemical Solutions DC Sylgard 184 elastomer kit) by mixing an elastomer base and curing agent in a ratio of 1:5 to obtain a relatively stiff mold. Afterwards we mix the PDMS thoroughly, we pour it over the clean Si wafer and degas it in a desiccator (Figure 6, Step 2). We subsequently bake the PDMS and Si wafer for 2 h at 85°C, and afterwards leave it to cool down for ca. 30 min. In the final step, we carefully peel off the PDMS from the Si wafer (Figure 6, Step 3).
Finally we fabricate the positive structures in PDMS. We mix PDMS in a 1:10 ratio, degas, and pour it onto the previously cured PDMS mold (Figure 6, Step 5). We then again degas and allow to cure for 2 h at 85°C. Once the curing is complete, we leave the PDMS to cool down for at least 30 min. Subsequently we carefully separate the two PDMS layers from each other (Figure 6, Step 6). At this point the PDMS mold can be stored for later use.
The cured PDMS layer contains 24 positive structures, each of which can be used in an experiment. To study an organism under the microscope utilizing the device, the PDMS device should have an inlet and outlet port for media exchange, as well as a cover glass that seals the device. To fabricate the inlet and outlet ports we first carefully cut out a single PDMS device and punch holes at the two sides of the main trench using a 0.75 mm Harris Uni-Core puncher (Figure 6, Step 7). To bind the cover glass to the PDMS, we simultaneously expose the clean cover glass (ultra-sonicated in acetone and IPA) and the PDMS devices to an O2 plasma using a microwave plasma system (Plasma-Preen I, Plasmatic Systems Inc.) (Figure 6, Step 8). We then bring the two exposed surfaces into contact and press slightly. It is believed that when the surfaces are exposed to plasma, silanol groups (−OH) are developed, which form covalent siloxane bonds (Si −O−Si) when the two surfaces are brought into contact [35, 36]. We then bake the PDMS and attached cover slips for ca. 30 min at 85°C, after which they are ready to be used in an experiment. In the following section we demonstrate the utilization of this device in two types of experiments.
Utilizing the PDMS device
We illustrate the functionality of the microfluidic devices by injecting a fluorescent liquid (Invitrogen Alexa Fluor 514 Goat Anti-Rabbit IgG 2 mg /mL) into the growth channels. First, we attached tubing to the inlet and outlet of the device. We inject phosphate buffered saline (PBS) (Sigma, 0.01 M PBS - NaCl 0.138 M, KCl 0.0027 M, pH 7.4), into the device. After this, we simultaneously autoclave (120°C for 15 min) the device and tubing. This is done both to ensure sterile conditions when working with micro-organisms and to remove any air bubbles present inside the device. After the autoclaving process is complete, we flush through 50 μ L bovine serum albumin (BSA) (10 mg/mL New England Biolabs) through the device and allow it to incubate for at least 15 min. This surface passivation step is done to reduce unwanted sticking of the specimen being studied to the glass and PDMS surfaces. After this incubation period we injected the dye (diluted 1:50 in PBS) into the device and image on a fluorescence microscope.
We successfully wet the growth channels as shown in Figure 8b,c. For illustration purposes we show only the largest growth channels, ca. 0.8 μ m (Figure 8b) and smallest ones ca. 0.3 μ m (Figure 8c). One can clearly observe that dye was successfully injected into both types of channels and can readily visualize their differences in size.
Next we demonstrate that sub-micron size bacteria can successfully be observed in the microfluidic device. For this purpose we use both L. lactis and E. coli. L. lactis has a diameter of ca. 0.8 μ m , as do E. coli when cultured in a minimal medium . It is thus possible to immobilize both these species of bacteria in the largest growth channel of this type of microfluidic device.
We have presented a detailed protocol based on EBL together with specific dry etching procedures for the fabrication of a microfluidic device suited to study submicron-sized bacteria. In comparison to approaches based on conventional optical lithography, our method provides enhanced versatility and control of the dimensions of the growth channels while satisfying the rapid-prototyping needs in a research environment. The widths of the submicron growth channels allow for the potential immobilization and study of different size bacteria with widths ranging from 0.3 μ m to 0.8 μ m. We verified by means of SEM that these structures are successfully transferred from Si into PDMS as well as from PDMS into PDMS. As a proof-of-principle, we demonstrated that both Gram-positive and Gram-negative bacteria can successfully be loaded and imaged over a number of generations in this device. Similar microfluidic devices could potentially be used to study other submicron-sized organisms under conditions in which the height and shape of the growth channels are crucial to the experimental design.
The microscope setup used during the experiments consists of a commercial Nikon Ti, a customized laser illumination path, and a personal computer (PC) running Nikon NIS elements. Different illumination schemes were used for the different measurements. A Cobolt Fandango 515 nm continuous wave (CW) diode-pumped solid-state (DPSS) laser is used to excite the fluorescent dye to verify device wettability. The experiments with L.lactis and E. coli are performed using standard brightfield illumination. In all the experiments a Nikon CFI Apo TIRF 100x oil (NA 1.49) objective is used for imaging. Bright field images are acquired every 5 minutes.
Cell culture preparation for microscopy
The L.lactis cultures are grown directly from plate in Luria-Bertani medium (LB) at 30°C until an OD 600≈0.2 is reached. The cell culture is then concentrated by centrifugation and injected into the PDMS device. A syringe pump is used to inject fresh LB medium.
E. coli are grown in M9 medium supplemented with 0.3% glycerol at 37°C overnight with shaking, and sub-cultured in the morning until an OD 600≈0.2 is reached. The cell culture is then concentrated by centrifugation and injected into the PDMS device. A syringe pump is used to inject fresh M9-glycerol medium.
The authors thank Felix Hol, Fabai Wu, Peter Galajda, Jaan Männik, Simon van Vliet, Mats Walldén and Johan Elf for initial help with the fabrication process, Luping Xu and Peipei Chen for initial help with loading of the cells, Arnold van Run, Hozanna Miro, and Marc Zuiddam for technical assistance with the fabrication instruments, Francesco Pedaci for his help with the acquisition of the SEM images, Greg Schneider for initial help with the silanization procedure, Serge Donkers for initial help with L. lactis, and Jelle van der Does, Dimitri de Roos and Jaap Beekman for help with instrumentation, Melanie Roemer for her help with the illustrations. This work was supported by the Netherlands Organization for Scientific Research (NWO), Delft University of Technology and the European Community’s Seventh Framework Programme FP7/2007–2013 under grant agreements n° 241548 (MitoSys).
- Kim SM, Lee SH, Suh KY: Cell research with physically modified microfluidic channels: a review. Lab Chip. 2008, 8 (7): 1015-1023. 10.1039/b800835c.View ArticleGoogle Scholar
- Huang Y, Agrawal B, Sun D, Kuo JS, Williams JC: Microfluidics-based devices: New tools for studying cancer and cancer stem cell migration. Biomicrofluidics. 2011, 5: 013412-10.1063/1.3555195.View ArticleGoogle Scholar
- Zhong JF, Chen Y, Marcus JS, Scherer A, Quake SR, Taylor CR, Weiner LP: A microfluidic processor for gene expression profiling of single human embryonic stem cells. Lab Chip. 2008, 8: 68-74. 10.1039/b712116d.View ArticleGoogle Scholar
- Maerkl SJ, Quake SR: A systems approach to measuring the binding energy landscapes of transcription factors. Science. 2007, 315 (5809): 233-237. 10.1126/science.1131007.View ArticleGoogle Scholar
- Yager P, Edwards T, Fu E, Helton K, Nelson K, Tam MR, Weigl BH: Microfluidic diagnostic technologies for global public health. Nature. 2006, 442 (7101): 412-418. 10.1038/nature05064.View ArticleGoogle Scholar
- Danino T, Mondragón-Palomino O, Tsimring L, Hasty J: A synchronized quorum of genetic clocks. Nature. 2010, 463 (7279): 326-330. 10.1038/nature08753.View ArticleGoogle Scholar
- Männik J, Driessen R, Galajda P, Keymer JE, Dekker C: Bacterial growth and motility in sub-micron constrictions. PNAS. 2009, 106 (35): 14861-14866. 10.1073/pnas.0907542106.View ArticleGoogle Scholar
- Gan M, Su J, Wang J, Wu H, Chen L: A scalable microfluidic chip for bacterial suspension culture. Lab Chip. 2011, 11 (23): 4087-4092. 10.1039/c1lc20670b.View ArticleGoogle Scholar
- Locke JCW, Elowitz MB: Using movies to analyse gene circuit dynamics in single cells. Nat Rev Microbiol. 2009, 7 (5): 383-392. 10.1038/nrmicro2056.View ArticleGoogle Scholar
- Weibel DB, Diluzio WR, Whitesides GM: Microfabrication meets microbiology. Nat Rev Microbiol. 2007, 5 (3): 209-218. 10.1038/nrmicro1616.View ArticleGoogle Scholar
- Yin H, Marshall D: Microfluidics for single cell analysis. Curr Opin Biotechnol. 2011, 23: 1-10.View ArticleGoogle Scholar
- Lidstrom ME, Meldrum DR: Life-on-a-chip. Nat Rev Microbiol. 2003, 1 (2): 158-164. 10.1038/nrmicro755.View ArticleGoogle Scholar
- Bennett MR, Hasty J: Microfluidic devices for measuring gene network dynamics in single cells. Nature Rev Gen. 2009, 10 (9): 628-638. 10.1038/nrg2625.View ArticleGoogle Scholar
- Wang Y, Chen Z: Microfluidic techniques for dynamic single-cell analysis. Microchimica Acta. 2010, 168: 177-195. 10.1007/s00604-010-0296-2.View ArticleGoogle Scholar
- Zare RN, Kim S: Microfluidic platforms for single-cell analysis. Annu Rev Biomedx Eng. 2010, 12: 187-10.1146/annurev-bioeng-070909-105238.View ArticleGoogle Scholar
- Cookson S, Ostroff N, Pang WL, Volfson D, Hasty J: Monitoring dynamics of single-cell gene expression over multiple cell cycles. Mol Syst Biol. 2005, 1: 2005.0024-View ArticleGoogle Scholar
- Guo P, Hall EW, Schirhagl R, Mukaibo H, Martin CR, Zare RN: Microfluidic capture and release of bacteria in a conical nanopore array. Lab Chip. 2012, 12 (3): 558-561. 10.1039/c2lc21092d.View ArticleGoogle Scholar
- Taniguchi Y, Choi PJ, Li GW, Chen H, Babu M, Hearn J, Emili A, Xie XS: Quantifying E. coli proteome and transcriptome with single-molecule sensitivity in single cells. Science. 2010, 329 (5991): 533-538. 10.1126/science.1188308.View ArticleGoogle Scholar
- Wang P, Robert L, Pelletier J, Dang WL, Taddei F, Wright A, Jun S: Robust Growth of Escherichia coli. 5. Curr Biol. 2010, 20 (12): 1099-1103. 10.1016/j.cub.2010.04.045.View ArticleGoogle Scholar
- Grossman N, Ron EZ, Woldringh CL: Changes in cell dimensions during amino acid starvation of Escherichia coli. J Bacteriol. 1982, 152: 35-41.Google Scholar
- Nelson D, Young K: Penicillin binding protein 5 affects cell diameter, contour, and morphology of Escherichia coli. J Bacteriol. 2000, 182: 1714-1721. 10.1128/JB.182.6.1714-1721.2000.View ArticleGoogle Scholar
- Zhao H, Dreses-Werringloer U, Davies P, Marambaud P: Amyloid-beta peptide degradation in cell cultures by mycoplasma contaminants. BMC Res Notes. 2008, 1: 38-10.1186/1756-0500-1-38.View ArticleGoogle Scholar
- Partensky F, Hess WR, Vaulot D: Prochlorococcus, a marine photosynthetic prokaryote of global significance. Microbiol Mol Biol Rev. 1999, 63: 106-127.Google Scholar
- Kokkinosa A, Fasseas C, Eliopoulos E, Kalantzopoulos G: Cell size of various lactic acid bacteria as determined by scanning electron microscope and image analysis. Lait. 1998, 78 (5): 491-500. 10.1051/lait:1998546.View ArticleGoogle Scholar
- Moffitt JR, Lee JB, Cluzel P: The single-cell chemostat: an agarose-based, microfluidic device for high-throughput, single-cell studies of bacteria and bacterial communities. Lab Chip. 2012, 12: 1487-10.1039/c2lc00009a.View ArticleGoogle Scholar
- Altissimo M: E-beam lithography for micro-/nanofabrication. Biomicrofluidics. 2010, 4: 026503-10.1063/1.3437589.View ArticleGoogle Scholar
- Qin D, Xia Y, Whitesides GM: Soft lithography for micro- and nanoscale patterning. Nat Protoc. 2010, 5 (3): 491-502. 10.1038/nprot.2009.234.View ArticleGoogle Scholar
- McDonald J, Duffy D, Anderson J, Chiu D, Wu H, Schueller O, Whitesides G: Fabrication of microfluidic systems in poly (dimethylsiloxane). Electrophor. 2000, 21: 27-10.1002/(SICI)1522-2683(20000101)21:1<27::AID-ELPS27>3.0.CO;2-C.View ArticleGoogle Scholar
- Whitesides G, Ostuni E: Soft lithography in biology and biochemistry. Annu Rev Mater Sci. 2001, 28: 153-Google Scholar
- Bender M, Plachetka U, Ran J, Fuchs A, Vratzov B, Kurz H, Glinsner T, Lindner F: High resolution lithography with PDMS molds. J Vac Sci Technol B Microelectron Nanometer Struct. 2004, 22 (6): 3229-10.1116/1.1824057.View ArticleGoogle Scholar
- Jansen HV, de Boer MJ, Unnikrishnan S, Louwerse MC, Elwenspoek MC: Black silicon method X: a review on high speed and selective plasma etching of silicon with profile control: an in-depth comparison between Bosch and cryostat DRIE processes as a roadmap to next generation equipment. J Micromech Microeng. 2009, 19 (3): 033001-10.1088/0960-1317/19/3/033001.View ArticleGoogle Scholar
- Wei J: Silicon MEMS for detection of liquid and solid fronts. PhD thesis. Delft University of Technology: Micro-elektronica & Computer Engineering, 2010.Google Scholar
- Duffy DC, McDonald JC, Schueller OJ, Whitesides GM: Rapid Prototyping of Microfluidic Systems in Poly(dimethylsiloxane). Anal Chem. 1998, 70 (23): 4974-4984. 10.1021/ac980656z.View ArticleGoogle Scholar
- Odom T, Love J, Wolfe D, Paul K: Improved pattern transfer in soft lithography using composite stamps. Langmuir. 2002, 18: 5314-10.1021/la020169l.View ArticleGoogle Scholar
- Eddings M, Johnson M: Determining the optimal PDMS–PDMS bonding technique for microfluidic devices. J Micromech. 2008, 18: 067001-10.1088/0960-1317/18/6/067001.View ArticleGoogle Scholar
- Bhattacharya S, Datta A, Berg J: Studies on surface wettability of poly (dimethyl) siloxane (PDMS) and glass under oxygen-plasma treatment and correlation with bond strength. J MEMS. 2005, 14: 590-View ArticleGoogle Scholar
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.