Nanoscale Hydrodynamic Film for Diffusive Mass Transport Control in

Subscriber access provided by UNIV OF NEWCASTLE

Article

Nanoscale Hydrodynamic Film for Diffusive Mass Transport Control in Compartmentalized Microfluidic Chambers Minseok Kim, Ji Won Lim, Sung Kuk Lee, and Taesung Kim Anal. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.analchem.7b01966 • Publication Date (Web): 20 Jul 2017 Downloaded from http://pubs.acs.org on July 22, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Analytical Chemistry is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 28

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

Nanoscale Hydrodynamic Film for Diffusive Mass Transport Control in Compartmentalized Microfluidic Chambers Minseok Kim1†, Ji-Won Lim2, Sung Kuk Lee2,3, and Taesung Kim1,2* 1

Department of Mechanical Engineering, Ulsan National Institute of Science and Technology (UNIST), 50 UNIST-gil, Ulsan 44919, Republic of Korea 2 Department of Biomedical Engineering, Ulsan National Institute of Science and Technology (UNIST), 50 UNIST-gil, Ulsan 44919, Republic of Korea 3 Department of Chemical Engineering, Ulsan National Institute of Science and Technology (UNIST), 50 UNIST-gil, Ulsan 44919, Republic of Korea *

Correspondence:

Taesung Kim Department of Mechanical Engineering Ulsan National Institute of Science and Technology (UNIST) 50 UNIST-gil, Ulsan 44919, Republic of Korea E-mail: [email protected] Tel.: +82-52-217-2313 Fax: +82-52-217-2409 †

Current affiliation:

Daegu Research Center for Medical Devices and Rehab Engineering, Korea Institute of Machinery and Materials, 330 Techno Sunhwan-ro, Yuga-myeon, Dalsung-gun, Daegu 42994 Republic of Korea.

1 ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Abstract A compartmentalized microfluidic chamber array that not only offers separate cell culture environments but also independent control of the diffusion of small molecules provides an extremely useful platform for cell cultivations and versatile cellular assays. However, it is challenging to incorporate both cell compartmentalization and active diffusion control in real-time and precise manners. Here, we present a novel nanoscale hydrodynamic film (NHF) that is formed between a solid substrate and a polydimethylsiloxane (PDMS) surface. The thickness of the NHF can be adjusted by varying the pressure applied between them so that the mass transfer through the NHF can also be controlled. These novel phenomena are characterized and applied to develop a compartmentalized microchamber array with a diffusion-tunable and solution-switchable chemostat-like versatile bacterial assays. The NHF-based compartmentalization technique is ideal for not only continuous bacterial cultivation by consistently refreshing various nutrient sources, but also various diffusion-based microbial assays such as chemical induction of synthetically engineered bacterial cells and selective growth of a specific bacterial strain with respect to chemical environments. In addition, we show that tight compartmentalization protects cells in the chambers, while biofilm formation and nutrient contamination are eliminated with a loading lysis buffer, which typically hinders long-term continuous cultures and accurate microbial assays on a chip. Therefore, we ensure that the NHF-based compartmentalization platform proposed in this work will not only facilitate fundamental studies in microbiology but also various practical applications of microbes for production of valuable metabolites and by-products in a high throughput and highly efficient format.


Page 2 of 28

Page 3 of 28

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Introduction Microfluidic compartmentalization has been an extremely useful technique to characterize cellular behavior (e.g., cell growth, cell-fate decisions, and antibiotic resistance) under well-defined extracellular chemical environments1-3. Notably, recent progress in microfluidic devices that physically compartmentalize bacterial cells in a small chamber (pL–nL) and exquisitely control extracellular chemical environments by virtue of a laminar flow at low Reynolds numbers and small characteristic volumes, allowing rapid adaptation to a defined physiochemical condition1,4-10. Therefore, microfluidic compartmentalization that can be considerably parallelized within a small chip was considered as a highly efficient method to perform multiple bacteriological cultures/assays in a simultaneous, versatile, and high-throughput manner, resulting in a highly sophisticated and near-real-time microbial analysis without significant reagent consumption or operational challenges7,11-14. In general, microfluidic compartmentalization techniques minimize convective flows and allow for purely diffusive mass transport13,15-17, particularly preferred for suspended cells in liquid media (e.g., bacteria). This is because the shear stress caused by convective flow can flush away the subject cells and often induce physical damages to the cells18. To enable the diffusion-controlled compartmentalization of a small sample of micron-sized bacterial cells, many nanofabricated semipermeable structures have been integrated into microfluidic channels as diffusion-permitting and convection-preventing barriers1. For example, commercialized nanoporous membranes, such as polycarbonate, polysulfone, or polyethylene, were integrated between top and bottom PDMS microchannels, physically isolating bacterial cells in a microfluidic chamber, but allowing controlled chemical communications via diffusion19-21. A variety of commercial membranes allowed for the optimization of membrane characteristics such as the pore size, porosity, biocompatibility, and membrane thickness according to a specific application. However, the manual integration process requires a skilled bonding process, leading to high fabrication-to-fabrication variation and 3 ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

even resulting in fluid leakage or bubble generation. Another alternative was the integration of nanoporous hydrogels or microbead clusters by selectively patterning and solidifying the precursor solutions15,22-25. These techniques showed relatively low variations in diffusive mass transport controls compared to direct membrane integration, but spontaneous degradation of hydrogel structures and low robustness became a problem for long-term bacterial cultures and versatile bacteriological assays26. To overcome these issues, a microbioreactor was integrated with a nanochannel (or nanoslit) array for both bacterial compartmentalization and continuous chemical supplementation6,27, which was produced using a combination of standard photolithography and conventional nanofabrication techniques (e.g., electron or focused ion-beam lithography)28-30. The nanochannel-based compartmentalization exhibited a high degree of freedom in defining geometric features of both microchambers and nanochannels, showing high robustness and accuracy in the mass transport control of chemical compounds31. However, the nanochannel-based devices were difficult to employ for versatile and specific bacteriological assays because the fabrication required skilled operators, intensive labor, and expensive facilities. Furthermore, the rate of diffusive transport was predefined in the design of the nanochannels, which often seemed to be insufficient to feed a highdensity bacterial population due to rapid chemical consumption and generation in a microfluidic test chamber. In other words, the rates of nutrient consumption and metabolite generation exceed the rate of chemical refreshment through nanochannels, resulting in disruption of the chemically controlled environment. Therefore, an innovative microfluidic compartmentalization technique is urgently needed to tackle the current weaknesses of both the membrane-integrated and nanochannel-based approaches, resulting in long-term and versatile bacteriological assays even for high bacterial populations with constantly maintained chemical environments. Here, we propose novel, easy-to-use, versatile, and robust nanoscale hydrodynamic films (NHFs) that can physically isolate bacterial cells but allow tuneable chemical delivery via accurate, 4 ACS Paragon Plus Environment

Page 4 of 28

Page 5 of 28

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


rapid, and controllable diffusive transport control. First, we develop a diffusion-controlled bacterial compartmentalization mechanism using transport-tunable NHFs and then visualize/characterize the transport of small biomolecules according to various pressurization and geometric conditions. Second, we employ NHFs as a semi-permeable barrier for the continuous refreshment of nutrients and metabolites in a compartmentalized microfluidic chamber array. This chemostat-like chemical refreshment enables continuous bacterial proliferation capable of sustaining ultra-high bacterial populations with various nutrient feeding conditions. Third, we demonstrate versatile diffusionbased bacteriological assays in a compartmentalized microchamber, including chemical gene induction of synthetically engineered bacterial cells and selective growth of cocultured bacterial strains with regard to antibiotic resistance. Finally, we demonstrate selective lysis of bacterial biofilms by contaminating loaded chemicals through microchannels, while preserving the subject cells in compartmentalized chambers by temporally breaking NHFs to cut off the chemical transport.

Experimental Chemicals and reagents. We purchased all the chemicals and reagents used in this work from Sigma-Aldrich Corporation (Natick, MA, USA) unless otherwise noted. Green and red food-dyes were used to visualize diffusive transports through NHFs. Bacterial compartmentalization was characterized using 1 µm microparticles (No. MFCD00131492) owing to their similar critical dimension to the microbes. Fluorescein isothiocyanate, 50 µM, (FITC, No. 3326-32-7) in 1× phosphate buffered saline (PBS, 150 mM, pH 7.4, No. MFCD00131855) was used to quantify the rate of diffusive transport in various pressurization conditions. For the bacterial culture, Tryptone broth (TB, 1% tryptone, and 0.5% NaCl, No. 211705), Luria broth (LB, No. L3022), and M9 (No. 248510) media were purchased from BD biosciences (Woburn, MA, USA). Pluronic F-127 (0.02 w%) was flushed through the microchannels to minimize non-specific clogging of the bacterial cells 5 ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

to the PDMS and glass walls. A mineral oil (No. M5904) was used as the immersive solution to prevent diffusive transport across the NHFs. A bleach solution (10 w% sodium hypochlorite, No. 7681-52-9) was used for testing the selective bacterial lysis. Fabrication of the master mold. The microfluidic device was fabricated using standard photolithography with SU-8 (SU-8 2010, MicroChem, Newton, MA, USA) and soft-lithography with PDMS (Dow Corning, Midland, Michigan, USA), followed by a monolithic layer-by-layer aligning and bonding process32. Briefly, the SU-8 2010 was spin-coated on a silicon wafer to form a 10-µm-thick photoresist film (3000 rpm for 30 s), followed by soft-baking on a hotplate (96° C) for 3 min. Then, the SU-8 was exposed to UV light (365 nm, 150 mJ cm-2) with the first photomask (Microimage, Ansan, Gyeonggi, Korea) and a mask aligner (MA6, Suss MicroTec AG, Schleissheimer, Garching, Germany). Next, another 10-µm-thick SU-8 layer was spin-coated, softbaked and UV-exposed on the top of the first layer with the second photomask using the same protocols, followed by a post-exposure bake on a hotplate (96° C) for 3 min. After developing the unexposed region in a SU-8 developer solution (MicroChem, Newton, MA, USA) for 5 min, duallevel microchannel structures appeared; the shallow and deep channels were approximately 10 µm and 20 µm, respectively. Another SU-8 master mold with 50 µm thickness was prepared to cast control channels using the SU-8 2050 (MicroChem, Newton, MA, USA) with the standard procedure as described in a manufacture’s protocol. Then, all the SU-8 master molds were coated with the trichloro(3,3,3-trifluoropropyl)silane (Sigma-Aldrich, Natick, MA, USA) for easy release of the PDMS chip from the master mold. Fabrication of microfluidic compartmentalization device. PDMS pre-polymer was spin-coated on a dual-level master mold under 500 rpm for 30 s (Spin-1200D, Midas System, Daejeon, Korea) followed by curing in a 65 °C convection oven for 1 h to fabricate membrane-like thin main channels (~1 mm thickness). At the same time, another PDMS slab (~5 mm thickness) with control channels was cured at the same temperature over 4 h. After curing, the 5-mm-thick PDMS slab and 6 ACS Paragon Plus Environment

Page 6 of 28

Page 7 of 28

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


1-mm-thick PDMS membrane were treated on a silicon substrate with an oxygen plasma (50 sccm O2, 50 W for 3 s, Cute-MP, Femto Science, Hwaseong, Gyeonggi, Korea). Subsequently, the control channel was mounted on top of the main channels through manual alignment and then irreversibly bonded together. Because the control channel was simply designed as a large rectangular chamber to completely cover all microchambers simultaneously, the alignment process required no skilled handling and/or additional equipment. Lastly, the monolithic control and main channels were detached from the SU-8 mold and bonded to a glass substrate to form dual-level main channels using the same oxygen plasma conditions. Preparation of bacterial cells. We used MG1655, an E. coli stain, for most bacterial assays in this work. The E. coli was transformed with plasmids that contain a constitutive promoter (pLtetO-1) and green fluorescent protein (GFP) genes to quantitatively analyze the bacterial population. For a dual induction study, the same stain was transformed with plasmid pZBRG, producing both red fluorescent proteins (RFPs) and GFPs in the presence of 40 mM arabinose and 10 µM tetracycline, respectively33. We also used a DH10B strain having kanamycin resistance to test the selective growth with respect to antibiotic chemical conditions. The same culture and preparation protocols as those used in our previous studies10,15 were employed for all the cells. Shortly after this, a single colony of E. coli grown on a LB agar plate was inoculated into 5 mL of LB, TB, or M9 buffer with 1% glucose, which was subsequently shaken in a rotary incubator shaker (200 rpm, 36 °C) over 12 h. A final optical density at a 600 nm wavelength (OD600) was gauged from 0.4 (TB and M9) to 1.0 (LB). Then, the bacterial cells were introduced to the microfluidic devices after appropriate dilution with working buffer solutions. On-chip bacterial cultivation and assays. The dual-level main channels were filled with bacterial solutions through punched inlet reservoirs, then a pneumatic pressure (50–60 kPa) was applied through an upper control channel to isolate the bacteria within the culture chambers. The critical pressure for the bacterial compartmentalization slightly varied in accordance with device 7 ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

geometries, but can be monitored and tweaked by microscopic observation and in-situ pressure regulation. Excessive solutions in the reservoirs were completely removed by pipetting, followed by increasing the pressure gage to over 250 kPa to protect the compartmentalized bacterial during the lysis-buffer treatment along the loading channels. This is an essential step to eliminate all residual bacteria in the reservoirs and loading channels that can contaminate culture buffers and eventually affect the bacterial assays in the compartmentalized microchambers. After the selective lysis step, test chemicals mixed in a nutrient media such as TB, LB, or M9 buffers were constantly flowed with a constant flow rate (5–25 µl/min) using a syringe pump (PHD ULTRA™ 4400, Harvard Apparatus, Holliston, MA, USA). Experimental setup and data analysis. An inverted fluorescence microscope (Ti-U, Nikon, Minato-ku, Tokyo, Japan) equipped with a charge-coupled device (CCD) camera (ORCA R2, Hamamatsu Photonics, Hamamatsu, Shizuoka, Japan) was used to obtain the optical and fluorescence images using software (NIS-Elements Advanced Research, Nikon, Minato-ku, Tokyo, Japan). The same software was also used for controlling a motorized stage (96S209-N2, LUDL Electronic Products LTD, NY, USA). Solenoid valves (S10MM-20-24-2, Pneumadyne Inc, Plymouth, MN, USA) and pneumatic manifolds (MSV10-4, Pneumadyne Inc, Plymouth, MN, USA) were controlled by a PC using a customized LabVIEW program (National Instruments, Austin, TX, USA). For data analysis and post processing, MATLAB (Mathworks, Natick, MA, USA), Image J (NIH, USA) and OriginPro 8.5 (OriginLab, Wheeling, IL, USA) were used.

Results NHFs for diffusive mass transport control. Figure 1a shows a mass-transport control mechanism using size-tuneable NHFs generated by pressurizing aqueous fluids between two hydrophilic surfaces. First, the dual-level PDMS microfluidic device, consisting of deep (20 µm) and shallow (10 µm) channels, was fully wetted and easily filled with an aqueous solution because


Page 8 of 28

Page 9 of 28

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


the polydimethylsiloxane (PDMS) and glass surfaces were hydrophilic after O2 plasma bonding (i). Then, the dual-level elastomer was deformed by pressurizing the monolithically integrated upper control channel (50 µm in height), which was filled with deionized water under a specific pressure using a pressure regulator. The pressurization in the control channel led to the roof collapse of the main channels where microparticles were loaded, and formed an isolated microchamber (compartmentalized chamber) and a NHF for diffusive mass transport control. Herein, the shallow channels were compressed to a nanometer scale, while the deep microchannels, such as the compartmentalized chambers and loading channels, remained at a micrometer scale (approximately 10 µm). The NHF formed at the rim of the compartmentalized chamber permitted chemical transport via diffusion34, resulting in rapid and continuous chemical exchanges without hydrodynamic drag on the microparticle motion (iii). For the repeatable compartmentalization and rapid diffusion control, there were several design parameters that should be carefully considered during the fabrication of the dual-level compartmentalization structures. First, the ratio of deep and shallow channels needs to be about 2:1 because the ratio may affect adequate pressurization ranges for the formation of NHFs. Second, the circular chamber shape and small chamber volume (e.g., 100 µm in diameter and 75 pL in volume) was determined to completely refresh the solution in each chamber with surrounding flow within several minutes by maximizing diffusive transport flux. Third, a post structure array in the compartmentalized chambers are recommended for a large chamber diameter to avoid the roof buckling problems. Lastly, as the width of chamber walls (i.e., shallow channel width) decreases, rapid diffusion is allowed due to the reduced diffusion resistance. However, in case that the width is less than 10 µm, the robustness of the chamber wall structures is not guaranteed when strongly pressurized. We visualized the diffusive transport through the NHFs using a solution mixture containing microparticles and colored food dyes (Figure 1b). After loading the solution, we pressurized the microchamber to compartmentalize the microparticles with a characteristic length (dmp = 1 µm) similar to that of microbes, followed by the introduction of a 9 ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

particle-free, red-colored solution at the loading channel. The number of microparticles (Nmp = 17) in the chamber was maintained, during which the chemical environment (green color) was rapidly adapted to a given red-colored buffer within 3 min. We again introduced a green-colored and particle-free buffer, which reconfirmed the switch in the chemical environment of the compartmentalized chamber without convective particle motion. Characterization of microparticle compartmentalization. In addition, we further investigated the motion of particles in the compartmentalized chamber under various pressure and flow conditions to determine the critical pressure required to isolate the microparticles without any convective shear flows (Supplementary Videos S1 and S2). Figure 2a shows the representative particle motion with respect to the pressurization of the upper control channel. The microparticles moved freely across the shallow channel without pressurization (0 kPa, control) by a pressure-driven flow along the loading channel. As the applied pressure increased (10–30 kPa), the average directional moving velocity of the microparticles within the compartmentalized chamber was gradually reduced due to the increased flow resistance across the shallow channel. At higher pressures (>50 kPa), the particles in the compartmentalized chamber showed no directional movement, while the particle velocity along the loading channel was slightly enhanced because the flow stream bypassed the chamber to satisfy an incompressible and continuous flow. As a result, the compartmentalization represented two predominant transport types: convection dominant and diffusion dominant (Figure 2b). Interestingly, the critical pressure was the same, regardless of the flow velocity at the loading channels. This can be explained by the fact that the applied pressure at the control channel (~104 Pa) was much higher than that at the loading channel (101–102 Pa in typical laminar flows)35, implying that the flow velocity and pressure along the loading channel may not alter the NHF thickness. We also confirmed that the particle compartmentalization could be significantly parallelized by simply adding more chambers and only by employing a common pressurized control channel. As shown in


Page 10 of 28

Page 11 of 28

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 2c, the number of particles in the serially connected 12 chambers was almost equal regardless of the position of the individual chamber, only showing a small stochastic difference. Characterization of the rate of diffusive transport. We quantified the rate of diffusive transport through NHFs using fluorescent molecules at various diffusion-dominant pressures and shallowchannel widths (Figure 3). In the control experiment, in which the chamber was surrounded by an immersive mineral oil, we observed no noticeable diffusive transport of FITC (Figure 3a). The fluorescent intensity of the compartmentalized chamber was maintained over the observation time (1 h). By contrast, the fluorescent intensity was gradually diminished when the chamber was surrounded by a water solution due to the diffusive transports of FITC. Consequently, the rate of diffusion was inversely proportional to the applied pressure (Figure 3b,c) while the microparticles were completely compartmentalized from the convective flow at all diffusion-dominant pressurization. We estimated the NHF thickness (hNHF) by comparing the experimental results of the diffusive transport rates with those obtained by numerical simulations, resulting in 700, 330, and 90 nm at the pressurization conditions of 70, 90, and 110 kPa, respectively (see Figure S1 for more detailed information). Thus, we confirmed that the pressurization can regulate the thickness of NHFs in submicron resolution, eventually determining the net rate of diffusive transport. In addition, the shallow-channel widths also determined the rate of diffusive transport; a wider shallow channel exhibited slower diffusion (Figure 3d). This was attributed to the decreased transport rate due to the reduced NHF height at high pressurization (50–110 kPa) and long diffusion distance with wide shallow-channel widths (5–40 µm), which can be accorded by the Fick’s law (J = –DA∇ ∇c, where J is the rate of diffusion, D is the diffusion coefficient, and c is the concentration). Consequently, the rate of diffusion was spatiotemporally tunable using both real-time regulations of applied pressures (active control of the diffusion height) and shallow-channel designs (passive control of the diffusion length). Although the critical pressures may vary depending on microfluidic channel designs, PDMS thickness and stiffness, and other tubing conditions, we minimized the run11 ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

to-run variation of the diffusive transport by constantly maintaining the aforementioned conditions and in situ pressure adjustment. Continuous bacterial cultivation in the diffusion-controlled microchamber. We cultured recombinant E. coli with constitutively expressing GFPs to demonstrate chemostat-like continuous bacterial growth using the constant nutrient delivery through the NHFs. We note that the florescent intensity from E. coli was assumed to be proportional to the cell population/density in this work. As shown in Figure 4a,b, we directly compared multiple continuous and fed-batch cultures in a single device having two separated inlets to individually flush the loading channels with fresh culture medium or an oil solution (Supplementary Video S3). Because the NHF allowed continuous refreshment of chemical environments (e.g., nutrients and metabolites) to the compartmentalized chamber (culture chamber) via diffusion, a few bacterial cells that were initially compartmentalized continuously proliferated over 24 hours and densely packed within the culture chamber, resulting in ultra-high bacterial density over 100 OD600 (Figure 4a, continuous culture). In contrast, the bacterial cells that were already fully grown in the test tube did not proliferate further when the culture chamber was surrounded by oil, owing to the lack of additional nutrient supply (Figure 4b, fedbatch culture). We further characterized bacterial growth behaviors under different nutrient sources and initial seeding numbers by measuring the fluorescence intensities of cells in the chamber. Figure 4c shows bacterial growth curves in the various nutrient supply, which clearly appeared to be the lag, exponential, and stationary growth phases. The nutrient sources significantly affected the period taken to reach the exponential phases and the slopes of the exponential phase curves indicating bacterial growth rates. Figure 4d shows the effects of the initial seeding number on the lag phases when the nutrient source was fixed as a Luria-Broth medium; the higher the initial seeding number, the earlier the exponential phase started, despite similar slopes (growth rates). Interestingly, the bacterial densities at the stationary phase substantially exceeded an OD600 of 100 (approximately 1 12 ACS Paragon Plus Environment

Page 12 of 28

Page 13 of 28

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


× 1011 cells/mL), representing a unique advantage of constant chemical refreshment using NHFs that could not be achieved using conventional fed-batch or semi-continuous culture methods. Another main advantage of the NHF-based continuous bacterial cultivation is that any microfluidic components, such as pumps, valves, mixers, and flow distributors, can easily be integrated with the high degree of freedom in designing the numbers and positions of compartmentalized chambers. To demonstrate this ability, we substantially increased the number of culture chambers (500 units) with a Christmas-tree-shaped microchannel36 to uniformly distribute the nutrients, fluids, and cells. As we already determined that the initial seeding numbers of microparticles (dmp = 1 µm) were uniform regardless of the position of individual chambers (Figure 2c), it is reasonable to assume that there is a uniform bacterial population during the initial seeding step. Consequently, Figure 4c demonstrates that all 500 culture chambers achieved ultra-high bacterial populations with high uniformity, regardless of the location and small variations in initial seeding numbers, enabling high-throughput continuous bacterial cultures by only employing a common control channel connected with a solenoid valve. This feature appears to be important because the compartmentalized chamber array enabled high throughput and massively parallelized bacterial cultivation, which may facilitate the production of valuable metabolites using synthetically engineered microbes37. Tightly sealed compartmentalized chambers for selective lysis. As shown in Figure 5a, longterm cultivation of bacterial cells on a chip usually led to bacterial biofilm formation in microchannels, which may alter the chemical and cellular conditions compared to the initial conditions. For enabling accurate and long-term bacteriological assays, this biofouling and bacterial contamination when loading microchannels and other fluidic reservoirs should be carefully managed and disinfected. The compartmentalized chamber with NHFs demonstrated the ability to actively control the diffusive transport and even completely cut off diffusion by using simple pressure regulation. For example, provided higher pressurization (i.e., 250 kPa) than the one 13 ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

required for enabling diffusive transport (50–110 kPa), the rate of diffusive mass transport became almost zero because the PDMS shallow channel surface made complete contact with the glass substrate, breaking the NHFs. As a result, bacterial chucks and biofilms at the loading channels were selectively and promptly eliminated after introducing a lysis buffer without causing any damage to the compartmentalized cells by blocking the diffusive transport of the lysis buffer into the chamber (Figure 5b and Supplementary Video S4). This was confirmed by comparing fluorescent intensities between the inside and outside of the culture chambers; the intensity from the cells at the loading channel completely and selectively disappeared within 8 seconds (Figure 5c). This selective lysis appears to be a unique function of the NHF-based compartmentalized chamber for allowing not only high-throughput and long-term bacterial cultivation but also accurate and versatile bacteriological assays without any chemical and cellular cross contaminations. Chemical induction of synthetically engineered bacterial cells. The diffusion-controlled and convection-free chemical environment of the compartmentalized chamber array could be also utilized as a versatile tool for bacteriological assays. As an exemplary application, we implemented a genetic induction of a synthetically engineered bacterial strain that harbored an artificial genetic plasmid (tzBRG). The tzBRG plasmid was synthetically designed for bacterial cells to express different fluorescent proteins with respect to a chemical inducer. As shown in Figure 6ab, we introduced 40 mM arabinose or 10 µM tetracycline with a culture media (TB) into the compartmentalized chamber through NHFs, which resulted in the expression of RFPs and GFPs, respectively. In contrast, no production of GFPs or RFPs was observed without the treatment of chemical inducers, while bacterial cells were continuously proliferated at all conditions (Figure 6c,d). Thus, we successfully demonstrated that the arabinose and tetracycline chemically switched on each promoter in a selective manner without significant crosstalk. Interestingly, the GFPs/RFPs expression level was not saturated but continuously increased over the long observation time (15 h). However, the same synthetically engineered strain was characterized using the same chemical 14 ACS Paragon Plus Environment

Page 14 of 28

Page 15 of 28

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


conditions in our previous work where the fluorescent signals approached a plateau in 4 hours13. This may be attributed to the fact that the NHF-based compartmentalized chamber allowed continuous bacterial proliferation during the chemical induction, leading to a higher bacterial population than our previous work. This population increase played important role in enhancing GFPs/RFPs expression levels, which was confirmed by obtaining both florescent and blight-field images at the final states (15 h). Consequently, the high production yield using the NHF-based continuous chemical refreshment would be applied for not only achieving high sensitivity in wholecells-based microbial sensors10 but also rapid and efficient quantification of various diffusion-based bacteriological assays13. Selective bacterial growth by chemical environment control. Another application of the NHFbased compartmentalized chamber was directed and selected bacterial growth according to the chemical preference of microbes. We used two bacterial strains for the assay: GFP-expressing MG1655 and RFP-expressing DH10B. The MG1655 and DH10B were engineered to be resistive to ampicillin and kanamycin, respectively. As shown in Figure 6a,b, we initially compartmentalized the bacterial strains in the same chamber with a different population ratio; the majority of the population was MG1655, while a single DH10B was seeded. Then, we continuously flushed a nutrient solution (TB) containing 1× kanamycin to continuously coculture the cells. As a result, the RFP signal from the kanamycin resistive DH10B continuously increased while the GFP intensity gradually reduced. We found that the population dominancy was changed from MG1655 to DH10B in approximately 24 hours, and the chamber was predominantly occupied by the DH10B after 42 hours. Although the antibiotic selection is a general method to separate a specific strain from complex bacterial mixtures, the microfluidic compartmentalized chamber may enable more rapid and efficient selection because the chemical environments could be almost constantly maintained without accumulation of metabolites generated by the non-selected dominant cells (e.g., MG1655). This means that the metabolites (e.g., quorum sensing molecules) generated by the dominant strain 15 ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

may affect the growth of minority cells via quorum sensing38. In addition, the diffusion-controlled compartmentalized chamber was convenient to supply fresh nutrients and antibiotics in a controllable manner, while the conventional fed-batch antibiotic dose should consider degradation and/or consumption of antibiotics and nutrients over time; this usually required multiple subcultures and more complicated and repeated handling. Therefore, we demonstrated that our compartmentalized chamber may have applications in the selection/screening of a specific target bacterial strain from a complex bacterial mixture using the resistance against antibiotics or preference to a chemical environment.

Discussion It is worthwhile to discuss the advantages of the compartmentalized chamber array using sizetunable NHFs compared to the conventional, membrane-based and/or nanochannel-based microfluidic devices in diffusion controls. First, the NHF-based technique controlled the rate of diffusion of small biomolecules in a real-time and accurate manner by simply adjusting a pressure gage so that the influence of shear flows was mostly eliminated. However, the conventional membrane-based and nanochannel-based devices appeared to be difficult to accurately control the rate of diffusion in a real-time manner because the commercial membranes and nanochannels have pre-determined diffusive transport areas; this feature enabled unique functions such as a selective lysis and spatiotemporal control of diffusive transport rates. Second, the NHF-based compartmentalized chamber could easily be integrated with other microfluidic components for more complex and multiplexed bacteriological assays. However, it would not be easy to incorporate other microfluidic components with most membrane-based devices due to geometric and observing constraints with integrated opaque membranes. Finally, the HHF-based compartmentalization device could be produced by a standard protocol such as photolithography and soft lithography


Page 16 of 28

Page 17 of 28

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


without any expensive nanofabrication facility and skilled manual microprocessing, thereby guaranteeing high reproducibility and low device-to-device variation.

Conclusions In this study, we conceived a novel diffusion-control mechanism using NHFs that can be accurately and promptly controlled using simple pressure regulation, and then applied the transportcontrol element toward a microfluidic compartmentalization platform for microparticles and microbes. This NHF-based compartmentalization technique appears to be ideal for not only continuous bacterial cultivation by consistently refreshing various nutrient sources but also versatile diffusion-based bacteriological assays such as chemical induction of synthetically engineered bacterial cells and selective growth of a specific bacterial strain according to chemical environments. Therefore, the unique diffusion-controlled compartmentalization enabled obtaining ultra-high bacterial populations (>100 OD600nm) suspended in a liquid media, which can be considerably parallelized (>500 chambers) and easily combined with other microfluidic components (e.g., micromixer) for high-throughput and versatile downstream analysis. Furthermore, the NHF-based microchamber allowed in-situ regulation of diffusion and convection, resulting in tunable and selective mass transport controls. This successfully resolved the issues related to biofilm formation and nutrient contamination that typically hindered long-term continuous culture and accurate bacteriological assays on a chip. Hence, we believe that the proposed NHF-based compartmentalization platform will facilitate not only fundamental studies in microbiology (e.g., characterization of synthetic microbes, antibiotic susceptibility test, and directed evolution) but also industrial applications for high-throughput and highly efficient production of valuable metabolites and by-products. Furthermore, the diffusion-control principle described in this work will be useful for developing in vitro type bioreactor platforms to investigate other microbial species and higher orders of cells in the near future.



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

SUPPORTING INFORMATION The Supporting Information is available free of charge on the ACS Publications website at DOI: Supplementary Figure S1 for numerical simulations of NHF thicknesses; Supplementary Movies S1 through S4.

Acknowledgements This work was supported by the National Research Foundation of Korea (NRF) grant funded by the Korea government (MSIP) (Mid-career Researcher Program: NRF-2014R1A2A1A10050431 and Basic Research Lab Program: 2017R1A4A1015564). This work was also supported by a grant from the Next-Generation BioGreen 21 program (SSAC, PJ01118601), Rural Development Administration, Republic of Korea.


Page 18 of 28

Page 19 of 28

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


References (1) Wu, F. B.; Dekker, C. Chem Soc Rev 2016, 45, 268-280. (2) Keymer, J. E.; Galajda, P.; Muldoon, C.; Park, S.; Austin, R. H. P Natl Acad Sci USA 2006, 103, 17290-17295. (3) Balagadde, F. K.; You, L. C.; Hansen, C. L.; Arnold, F. H.; Quake, S. R. Science 2005, 309, 137140. (4) El-Ali, J.; Sorger, P. K.; Jensen, K. F. Nature 2006, 442, 403-411. (5) Weibel, D. B.; DiLuzio, W. R.; Whitesides, G. M. Nat Rev Microbiol 2007, 5, 209-218. (6) Groisman, A.; Lobo, C.; Cho, H. J.; Campbell, J. K.; Dufour, Y. S.; Stevens, A. M.; Levchenko, A. Nat Methods 2005, 2, 685-689. (7) Sun, P.; Liu, Y.; Sha, J.; Zhang, Z. Y.; Tu, Q.; Chen, P.; Wang, J. Y. Biosens Bioelectron 2011, 26, 1993-1999. (8) Lee, K. S.; Boccazzi, P.; Sinskey, A. J.; Ram, R. J. Lab Chip 2011, 11, 1730-1739. (9) Zhang, Z.; Boccazzi, P.; Choi, H. G.; Perozziello, G.; Sinskey, A. J.; Jensen, K. F. Lab Chip 2006, 6, 906-913. (10) Kim, M.; Lim, J. W.; Kim, H. J.; Lee, S. K.; Lee, S. J.; Kim, T. Biosens Bioelectron 2015, 65, 257-264. (11) Lee, P. J.; Hung, P. J.; Rao, V. M.; Lee, L. P. Biotechnol Bioeng 2006, 94, 5-14. (12) Eyer, K.; Kuhn, P.; Hanke, C.; Dittrich, P. S. Lab Chip 2012, 12, 765-772. (13) Kim, M.; Kim, T. Anal Chem 2015. (14) Kim, M.; Ha, D.; Kim, T. Nat Commun 2015, 6, 6247. (15) Kim, M.; Kim, T. Anal Chem 2010, 82, 9401-9409. (16) Si, G. W.; Yang, W.; Bi, S. Y.; Luo, C. X.; Ouyang, Q. Lab on a Chip 2012, 12, 1389-1394. (17) Atencia, J.; Morrow, J.; Locascio, L. E. Lab on a Chip 2009, 9, 2707-2714. (18) Kim, T.; Pinelis, M.; Maharbiz, M. M. Biomedical Microdevices 2009, 11, 65-73. (19) Abaci, H. E.; Gledhill, K.; Guo, Z. Y.; Christiano, A. M.; Shuler, M. L. Lab on a Chip 2015, 15, 882-888. (20) Diao, J. P.; Young, L.; Kim, S.; Fogarty, E. A.; Heilman, S. M.; Zhou, P.; Shuler, M. L.; Wu, M. M.; DeLisa, M. P. Lab on a Chip 2006, 6, 381-388.



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

(21) Chueh, B. H.; Huh, D.; Kyrtsos, C. R.; Houssin, T.; Futai, N.; Takayama, S. Analytical Chemistry 2007, 79, 3504-3508. (22) Jeong, H. H.; Lee, S. H.; Lee, C. S. Biosensors & Bioelectronics 2013, 47, 278-284. (23) Gumuscu, B.; Bomer, J. G.; van den Berg, A.; Eijkel, J. C. T. Lab on a Chip 2015, 15, 664-667. (24) Lee, S. H.; Heinz, A. J.; Shin, S.; Jung, Y. G.; Choi, S. E.; Park, W.; Roe, J. H.; Kwon, S. Analytical Chemistry 2010, 82, 2900-2906. (25) Hong, J. W.; Song, S.; Shin, J. H. Lab on a Chip 2013, 13, 3033-3040. (26) Hahn, S. K.; Park, J. K.; Tomimatsu, T.; Shimoboji, T. Int J Biol Macromol 2007, 40, 374-380. (27) Cho, H. J.; Jonsson, H.; Campbell, K.; Melke, P.; Williams, J. W.; Jedynak, B.; Stevens, A. M.; Groisman, A.; Levchenko, A. Plos Biol 2007, 5, 2614-2623. (28) Fouad, M.; Yavuz, M.; Cui, B. Journal of Vacuum Science & Technology B 2010, 28, C6i11C16i13. (29) Harnett, C. K.; Coates, G. W.; Craighead, H. G. Journal of Vacuum Science & Technology B 2001, 19, 2842-2845. (30) Wang, K.; Yue, S.; Wang, L.; Jin, A.; Gu, C.; Wang, P.; Wang, H.; Xu, X.; Wang, Y.; Niu, H. Iee Proceedings-Nanobiotechnology 2006, 153, 11-15. (31) Kim, M.; Ha, D.; Kim, T. Nat Commun 2015, 6. (32) Unger, M. A.; Chou, H. P.; Thorsen, T.; Scherer, A.; Quake, S. R. Science 2000, 288, 113-116. (33) Choi, W. S.; Kim, M.; Park, S.; Lee, S. K.; Kim, T. Biomaterials 2012, 33, 624-633. (34) Guo, Q.; Izumisawa, S.; Jhon, M. S.; Hsia, Y. T. Ieee T Magn 2004, 40, 3177-3179. (35) Fuerstman, M. J.; Lai, A.; Thurlow, M. E.; Shevkoplyas, S. S.; Stone, H. A.; Whitesides, G. M. Lab Chip 2007, 7, 1479-1489. (36) Kim, S.; Kim, H. J.; Jeon, N. L. Integr Biol (Camb) 2010, 2, 584-603. (37) Abbott, D. A.; Suir, E.; van Maris, A. J.; Pronk, J. T. Appl Environ Microbiol 2008, 74, 57595768. (38) de Kievit, T. R.; Iglewski, B. H. Infect Immun 2000, 68, 4839-4849.


Page 20 of 28

Page 21 of 28

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figures

Figure 1. Mass transport control through a nanoscale hydrodynamic film (NHF). a, Schematic diagram describing the overall process required to form NHFs for compartmentalization of microparticles and diffusive chemical transport control. The microfluidic device consisted of a duallevel main microchannel (loading channel and compartmentalized chamber) and a pressure-control channel (i). Pressurization at the control channel deformed the subjacent main channels, leading to the compartmentalization of microparticles in the chamber and formation of NHFs between the PDMS and glass surfaces (ii). The NHFs suppressed any convective shear stress by fluid flows but allowed chemical transport by pure diffusion, enabling continuous chemical exchanges (iii). b, Diffusive transport was visualized using 1-µm spheres and red- and green-colored food dyes. The microspheres statically remained in the chamber while the food dyes were rapidly adapted according to the chemical conditions of the loading channel.



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 2. Ccompartmentalization control in various pressurization and flow conditions. a,b, Particle velocities at the compartmentalized chamber were characterized under various pressurization and flow conditions, which revealed convection or diffusion dominant particle transport depending on the applied pressures. c, Characterization of the compartmentalization uniformity in a massively parallelized chamber array. The number of 1 µm spheres was almost uniform in a randomly selected chamber. The error bars were obtained by triplicated experiments.


Page 22 of 28

Page 23 of 28

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 3. Characterization of diffusive mass transport through NHFs in various pressurization conditions and shallow channel widths. a, Fluorescence intensities remained constant when the chamber was surrounded by an oil solution (control experiment). b, c, Diffusion through the NHFs was quantified using fluorescent molecules (e.g., FITC). The NHFs allowed diffusive molecular transport between the compartmentalized chamber and an aqueous solution surrounding the chamber. The diffusive transport rates were controlled by the applied pressure, which determined the height of NHFs. Under all the pressurized conditions, 1 µm spheres were confined inside the chamber without directed motions. d, Effects of the shallow channel width on diffusive transport rates. As the shallow channel width increased, the diffusive transport rate decreased at the given pressurization condition (70 kPa) owing to the increased diffusion resistance. When the shallow channel width was narrower than 5 µm, the diffusive transport rate was fairly unstable and sometimes out of control at the high pressurized conditions.



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 4. Chemostat-like continuous bacterial proliferation. a, A constant chemical environment achieved by using NHFs enabled continuous bacterial culture from a single cell to ultra-high populations. Nutrient media such as LB, TB, and M9-Glu showed different growth rates while the whole chamber was fully packed with bacterial cells at the end by continuous proliferation. b, Fedbatch cultures showed no additional bacterial growth because the loading channel was filled with an oil that blocked additional mass transport from/to the culture chamber. c, Quantification of bacterial growth under various culture media. The black and red dashed lines indicate the calibrated fluorescence intensities of OD600 = 10 and OD600 = 100, respectively. d, Bacterial growth curve according to the initial seeding number. The initial bacterial number affected lag-phase periods. The error bars were obtained by triplicated experiments. e, High-throughput and continuous bacterial cultures in a 500 compartmentalized chamber array, showing high-density, uniform bacterial populations and rapid growth rates over time in the device. Multiple fluorescent images were taken at 12 h after the compartmentalization and stitched together. 24 ACS Paragon Plus Environment

Page 24 of 28

Page 25 of 28

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 5. Tight sealing of the compartmentalized chambers for selective lysis of biofilms in the loading channels. a,b, Selective bacterial lysis at far higher pressurization (e.g., 250 kPa) was the one required for compartmentalization and diffusive transport control (e.g., 50 kPa). NHFs disappeared at the high pressurization, cutting off diffusive transport of lysis chemicals toward the culture chamber. c, Quantification of bacterial viability inside (upper four legends) and outside (lower four legends) of the chambers. Bacterial colonies and biofilms at the loading channels were selectively lysed while the bacterial cells in the culture chambers remained intact during the lysisbuffer-based channel clean-up. 25 ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 6. Continuous chemical induction of synthetically engineered bacterial cells. a, b, Characterization of continuous bacterial gene expression under various chemical inducers, such as 4 mM arabinose for RFPs and 50 nM tetracycline for GFPs. Bacterial cells containing the tzBRG plasmid were engineered to regulate a genetic circuit for expression RFPs or GFPs according to the chemical inducer. c, d, Control groups without loading chemical inducers showed no GFP or RFP expression, although the bacterial cells were fully proliferated in the chamber due to continuous supplementation of a nutrient medium (1× TB). e, Quantification of the fluorescent intensity over observation time (15 h). The error bars were obtained by triplicated experiments.


Page 26 of 28

Page 27 of 28

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 7. Selective growth of a target bacterial strain from a cocultured bacterial mixture. a, b, Two bacterial strains, GFP-expressing MG1655 and RFP-expressing DH10B, were initially loaded into the same culture chamber with a large population difference. c, Microscopic observation of the cocultured strains under continuous supplementation with TB with 1× kanamycin (50 µg/mL). Kanamycin-resistant DH10B survived and continuously grew in the chemostat-like environment, resulting in gradual increases in RFP signals. By contrast, GFP intensity from MG1655 continued to decrease owing to lack of antibiotic resistance.



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

For TOC only


Page 28 of 28

Nanoscale Hydrodynamic Film for Diffusive Mass Transport Control in

Recommend Documents