High-Efficiency Single-Cell Entrapment and Fluorescence in Situ

Jun 7, 2008 - When cells were introduced into the microfluidic device integrated with the .... Michelle M. Packard , Maxim Shusteff , Evangelyn C. Alo...
0 downloads 0 Views 602KB Size
Download PDF
Anal. Chem. 2008, 80, 5139–5145

High-Efficiency Single-Cell Entrapment and Fluorescence in Situ Hybridization Analysis Using a Poly(dimethylsiloxane) Microfluidic Device Integrated with a Black Poly(ethylene terephthalate) Micromesh Tadashi Matsunaga,* Masahito Hosokawa, Atsushi Arakaki, Tomoyuki Taguchi, Tetsushi Mori, Tsuyoshi Tanaka, and Haruko Takeyama Department of Biotechnology, Tokyo University of Agriculture and Technology, 2-24-16 Naka-cho, Koganei, Tokyo 184-8588, Japan Here, we report a high-efficiency single-cell entrapment system with a poly(dimethylsiloxane) (PDMS) microfluidic device integrated with a micromesh, and its application to single-cell fluorescence in situ hybridization (FISH) analysis. A micromesh comprising of 10 × 10 microcavities was fabricated on a black poly(ethylene terephthalate) (PET) substrate by laser ablation. The cavity was approximately 2 µm in diameter. Mammalian cells were driven and trapped onto the microcavities by applying negative pressure. Trapped cells were uniformly arrayed on the micromesh, enabling high-throughput microscopic analysis. Furthermore, we developed a method of PDMS surface modification by using air plasma and the copolymer Pluronic F-127 to prevent nonspecific adsorption on the PDMS microchannel. This method decreased the nonspecific adsorption of cells onto the microchannel to less than 1%. When cells were introduced into the microfluidic device integrated with the black PET micromesh, approximately 70-80% of the introduced cells were successfully trapped. Moreover, for mRNA expression analysis, on-chip fluorescence in situ hybridization (e.g., membrane permeabilization, hybridization, washing) can be performed in a microfluidic assay on an integrated device. This microfluidic device has been employed for the detection of β-actin mRNA expression in individual Raji cells. Differences in the levels of β-actin mRNA expression were observed in serum-supplied or serumstarved cell populations. Single-cell gene expression analysis has attracted a great deal of attention as the technology that can reveal the heterogeneity of individual cells in living organisms.1–5 In single-cell mRNA expression analysis, single cells have been collected by capillary * Corresponding author. Fax: +81-42-385-7713. Phone: +81-42-388-7020. E-mail: [email protected]. (1) Cai, L.; Friedman, N.; Xie, X. S. Nature 2006, 440, 358–362. (2) Golding, I.; Paulsson, J.; Zawilski, S. M.; Cox, E. C. Cell 2005, 123, 1025– 1036. (3) Yu, J.; Xiao, J.; Ren, X.; Lao, K.; Xie, X. S. Science 2006, 311, 1600–1603. (4) Newman, J. R.; Ghaemmaghami, S.; Ihmels, J.; Breslow, D. K.; Noble, M.; DeRisi, J. L.; Weissman, J. S. Nature 2006, 441, 840–846. 10.1021/ac800352j CCC: $40.75  2008 American Chemical Society Published on Web 06/07/2008

microaspiration6 or laser capture microdissection7 followed by mRNA extraction and amplification. However, the quantity of mRNA harvested from a single cell is estimated to be approximately 0.1-1.0 pg, and this is not sufficient for standard mRNA extraction procedures.8,9 Hence, mRNA isolation from individual cells has been difficult. The loss of mRNA due to RNase contamination and mRNA damage due to uncontrolled environments are some significant problems faced during single-cell reverse transcription polymerase chain reaction (RT-PCR). Fluorescence in situ hybridization (FISH) is an alternate method that allows the detection and precise localization of mRNAs in cells and subcellular domains without the need of mRNA isolation.10,11 Recently, monitoring of the expression of several mRNAs in an individual living cell has been made possible by using molecular beacons and microinjection techniques.12 In addition, a highly sensitive FISH method that enables mRNA enumeration at single-molecule resolution in individual cells has also been reported, indicating that FISH enables the detection of stochastic mRNA expression in mammalian cells.13 Therefore, FISH is an effective method to measure the mRNA expression levels in individual cells, which would then further reveal cell-to-cell gene expression variations and information obscured in the population average. In conjunction to the interest and attention in single-cell analysis research, microfluidic devices have also been utilized and have demonstrated their great potential and significance. Some (5) Warren, L.; Bryder, D.; Weissman, I. L.; Quake, S. R. Proc. Natl. Acad. Sci. U.S.A. 2006, 103, 17807–17812. (6) Hemby, S. E.; Ginsberg, S. D.; Brunk, B.; Arnold, S. E.; Trojanowski, J. Q.; Eberwine, J. H. Arch. Gen. Psychiatry 2002, 59, 631–640. (7) Schutze, K.; Lahr, G. Nat. Biotechnol. 1998, 16, 737–742. (8) Ginsberg, S. D. Methods 2005, 37, 229–237. (9) Nashimoto, Y.; Takahashi, Y.; Yamakawa, T.; Torisawa, Y. S.; Yasukawa, T.; Ito-Sasaki, T.; Yokoo, M.; Abe, H.; Shiku, H.; Kambara, H.; Matsue, T. Anal. Chem. 2007, 79, 6823–6830. (10) Femino, A. M.; Fay, F. S.; Fogarty, K.; Singer, R. H. Science 1998, 280, 585–590. (11) Levsky, J. M.; Shenoy, S. M.; Pezo, R. C.; Singer, R. H. Science 2002, 297, 836–840. (12) Medley, C. D.; Drake, T. J.; Tomasini, J. M.; Rogers, R. J.; Tan, W. Anal. Chem. 2005, 77, 4713–4718. (13) Raj, A.; Peskin, C. S.; Tranchina, D.; Vargas, D. Y.; Tyagi, S. PLoS Biol. 2006, 4, e309.

Analytical Chemistry, Vol. 80, No. 13, July 1, 2008

5139

of the efforts introduced to date include the scale-down of microchannels in microfluidic devices to accommodate the size of most cells and decreasing both the consumption of analytes and the duration of analyses to allow high time-resolved singlecell analysis. Single-cell entrapment techniques that employ microfluidic techniques such as mechanical trapping,14–16 optical tweezers,17 and electrochemical18 and dielectrophoretic19 techniques have also been widely studied. Thus far, cell analyses have mainly focused on techniques to improve cell lysis17,20,21 and consequent intracellular reactions such as protein extraction,22,23 PCR amplification,24 and intracellular enzyme reaction,25 but few have focused on single-cell manipulation or the analysis of gene expression levels. A microfluidic device with integrated pneumatic valves capable of isolating single cells and lysing them using a chemical lysis buffer and capable of extracting and recovering mRNA from a single cell26,27 is an example of the latter technique. However, current microfluidic single-cell analysis devices face challenges such as the small number of trapped cells, inefficient trapping due to long entrapment time, high cost, and structural and design complexity. In particular, inefficient cell entrapment proved to be a significant problem as it causes the loss of rare cells and low-throughput cell analysis. A more efficient method of trapping single cells from a designated cell population is required to address these problems. Furthermore, it is necessary to establish a cell array platform that can confine a large number of cells, allowing simultaneous analysis. Ideally, a single-cell analysis system should enable the entrapment, stimulation, and measurement of cells in a fast, high-throughput, and highly parallel manner. In a previous study, we developed a microfluidic device integrated with SUS micromesh for high-throughput detection of the protozoan parasite, Cryptosporidium parvum oocyst.28 The microfluidic device exploits a geometry that enabled entrapment of oocysts on microcavities of the micromesh, allowing their rapid enumeration under a microscope. Furthermore, the device introduces a series of reagents and washes through the microfluidic structure. Thus, a simple yet specific single-cell detection system was successfully developed. (14) Li, X.; Li, P. C. Anal. Chem. 2005, 77, 4315–4322. (15) Ionescu-Zanetti, C.; Shaw, R. M.; Seo, J.; Jan, Y. N.; Jan, L. Y.; Lee, L. P. Proc. Natl. Acad. Sci. U.S.A. 2005, 102, 9112–9117. (16) Di Carlo, D.; Aghdam, N.; Lee, L. P. Anal. Chem. 2006, 78, 4925–4930. (17) Munce, N. R.; Li, J.; Herman, P. R.; Lilge, L. Anal. Chem. 2004, 76, 4983– 4989. (18) Toriello, N. M.; Douglas, E. S.; Mathies, R. A. Anal. Chem. 2005, 77, 6935– 6941. (19) Voldman, J.; Gray, M. L.; Toner, M.; Schmidt, M. A. Anal. Chem. 2002, 74, 3984–3990. (20) McClain, M. A.; Culbertson, C. T.; Jacobson, S. C.; Allbritton, N. L.; Sims, C. E.; Ramsey, J. M. Anal. Chem. 2003, 75, 5646–5655. (21) Wu, H.; Wheeler, A.; Zare, R. N. Proc. Natl. Acad. Sci. U.S.A. 2004, 101, 12809–12813. (22) Schilling, E. A.; Kamholz, A. E.; Yager, P. Anal. Chem. 2002, 74, 1798– 1804. (23) Gao, J.; Yin, X. F.; Fang, Z. L. Lab Chip 2004, 4, 47–52. (24) Waters, L. C.; Jacobson, S. C.; Kroutchinina, N.; Khandurina, J.; Foote, R. S.; Ramsey, J. M. Anal. Chem. 1998, 70, 158–162. (25) Heo, J.; Thomas, K. J.; Seong, G. H.; Crooks, R. M. Anal. Chem. 2003, 75, 22–26. (26) Hong, J. W.; Studer, V.; Hang, G.; Anderson, W. F.; Quake, S. R. Nat. Biotechnol. 2004, 22, 435–439. (27) Marcus, J. S.; Anderson, W. F.; Quake, S. R. Anal. Chem. 2006, 78, 3084– 3089. (28) Taguchi, T.; Arakaki, A.; Takeyama, H.; Haraguchi, S.; Yoshino, M.; Kaneko, M.; Ishimori, Y.; Matsunaga, T. Biotechnol. Bioeng. 2007, 96, 272–280.

5140

Analytical Chemistry, Vol. 80, No. 13, July 1, 2008

Figure 1. Schematic diagram of the PDMS microfluidic device integrated with a micromesh for entrapment of mammalian cells: (A) the PDMS microfluidic device integrated with micromesh for entrapment of mammalian cells; (B) the black PET micromesh.

In this paper, mammalian cells were trapped onto single microcavities of the micromesh in a microfluidic device, and mRNA in individual cells was detected by FISH. EXPERIMENTAL SECTION Micromesh Fabrication. A micromesh was microfabricated according to a previously described method.28 Stainless steel type 304 (special use stainless, SUS 304 (20 mm × 20 mm, thickness 5 µm)) and poly(ethylene terephthalate) (PET) plates (20 mm × 20 mm, thickness 38 µm) were used as substrate to fabricate the micromesh. PET plates used in this study were either clear or black, in which the black PET plate used was of commercial grade and was dispersed with black pigments. Microcavities were drilled through the substrate with an excimer laser (Optec Micro-Master System, Optec S. A., Belgium) at a wavelength of 248 nm with a frequency of 150 Hz, to produce conical apertures with diameters of 2, 5, and 10 µm at the top surface (Figure 1B). The distance between each microcavity was 30 µm, with a total of 100 cavities arranged in each 10 × 10 array. Fluorescence images of the microcavities on the micromesh were captured using a confocal fluorescence microscope (FV1000, Olympus Corp., Tokyo, Japan) with an excitation wavelength of 559 nm and a band-pass filter within the range of 575-675 nm. Intensity data was stored at 12-bit gray scale levels, and fluorescence intensity data was analyzed accordingly. Microfluidic Device Fabrication. Poly(dimethylsiloxane) (PDMS) microfluidic devices were designed for two independent analyses. The first was to assess the nonspecific adsorption of mammalian cells and oligonucleotides on PDMS surfaces. To this end, a PDMS microfluidic device comprising of a single straight fluidic microchannel, 500 µm in length and depth, was prepared. In the second analysis, to examine the entrapment of lymphoma on the micromesh, a microfluidic device comprising of two substrates was designed as described below (Figure 1A). The upper substrate consists of a microchannel, sample inlet, and outlet, and a circular drain attached to the micromesh was fabricated in the middle of the microchannel. A vacuum line was fabricated in the lower substrate beneath the micromesh to produce negative pressure, enabling cell entrapment. Master mold substrates comprising poly(methyl methacrylate) (PMMA) were

prepared by a computer-aided modeling machine (PNC-300, Roland DG Corp., Shizuoka, Japan), and silicone tubes (i.d. 2 mm) were then connected to the inlet, outlet, and vacuum lines of the respective molds. Both the upper and lower PDMS layers were then fabricated by pouring a mixture of Sylpot 184 silicone elastomer (Dow Corning Asia Ltd., Tokyo, Japan) and curing agent (10:1) onto either the master molds or a blank wafer, followed by curing for at least 20 min at 85 °C. The cured PDMS substrates were carefully peeled off the molds and exposed to air plasma (condition: 40 mL/min) for 20 s. The two pieces of PDMS substrates were immediately bonded together after plasma treatment. Cell Culture and Labeling. Raji Burkitt’s lymphoma cells were cultured in RPMI 1640 medium, containing L-glutamine (Sigma-Aldrich, Irvine, U.K.), 10% (v/v) fetal bovine serum (FBS, Invitrogen Corp., Carlsbad, CA), and 1% (v/v) penicillin/streptomycin (Invitrogen Corp.) for 3-4 days at 37 °C with 5% carbon dioxide supplementation. Immediately prior to each experiment, the cells were washed with phosphate-buffered saline (PBS, pH 7.4) three times and resuspended in the same buffer. Cells were labeled with CellTracker red CMTPX (Molecular Probes, Eugene, OR), in which labeling was achieved by incubating the cells with the tracking dye (5 µM) for 30 min. Cells were pelleted by centrifugation (200g for 5 min), the supernatant was decanted, and cells were washed twice with PBS to remove any excess dye. For the cell viability assay, cells were incubated for 30 min with 5 µM calcein-AM (Molecular Probes) and the excess dye was removed by washing with PBS. The actual number of cells in suspension was determined as follows: 5 µL of cell suspension, which is equal to the volume injected into the microfluidic device, was transferred onto a glass slide where cells were counted under a fluorescence microscope. Counting was repeated three times to determine the number of cells in suspension. Surface Modification of the Microchannel. The surface of the single straight microchannel fabricated within the microfluidic device was modified to prevent nonspecific adsorption of cells. The microchannel was filled with various concentrations of Pluronic F-127 (Sigma-Aldrich, St. Louis, MO) solution for 2 h to modify or treat the PDMS surfaces. The surfactant was then flushed out with PBS at a flow rate of 100 µL/min for 10 min. Equilibrium contact angles were also measured to prove the effectiveness of surface modification with the surfactant. Ultrapure water (10 µL) was dropped onto the surface-modified PDMS, its lateral images were acquired with a digital microscope (VHX-100, KEYENCE Corp., Osaka, Japan), and the contact angles were calculated by analyzing the images with the specified software. An experiment to evaluate surface modification was performed using the single straight fluidic microchannel. Raji cells labeled with CellTracker red were resuspended in fresh PBS at a concentration of approximately 1 × 105 cells/mL. The cell suspension was introduced into a microchannel previously filled with PBS with a syringe pump at a flow rate of 15 µL/min for 1 min and incubated for 20 min at room temperature. Subsequently, 200 µL of PBS was loaded into the microchannel at a flow rate of 20 µL/min to flush out the cells from the microchannel. The device was mounted on the stage of a fluorescence microscope (BX-51, Olympus Corp.), and fluorescence images of adsorbed

cells within the microchannels were captured using a WIG filter set (excitation, 530-560 nm; emission, >575 nm) and a cooled digital camera (DP-70, Olympus Corp.). Entrapment of Lymphoma on the Micromesh and Cell Viability Test after Trapping. Raji cells labeled with CellTracker red were resuspended in PBS at a concentration of approximately 1 × 104 cells/mL. Five microliters of cell suspension was introduced into a PBS-filled microfluidic device equipped with SUS and black PET micromesh from the inlet of the upper layer. By applying negative pressure to the micromesh using the vacuum line, cells were driven toward and trapped onto the micromesh. The fluorescence images of the trapped cells were captured using a fluorescence microscope equipped with a cooled digital camera under a WIG filter set. To estimate the effect of entrapment on cell viability, cell viability was measured using calcein-AM and propidium iodide (PI).29,30 Cells that were prelabeled with calcein-AM were introduced into the microfluidic device and trapped onto the micromesh, followed by the introduction of 2 µg/mL PI solution via a syringe pump at a flow rate of 2 µL/min for 15 min. Fresh PBS was then loaded into the microchannel at a flow rate of 20 µL/ min for 10 min to wash out excess dyes. The fluorescence images of viable or damaged cells, visualized as green and red fluorescence on the micromesh, respectively, were obtained using a fluorescence microscope equipped with a digital camera under a WIG filter set for PI and an NIBA filter set (excitation, 470-490 nm; emission, 515-550 nm) for calcein-AM. Fluorescence in Situ Hybridization on the Micromesh. For single-cell analysis by FISH, Raji cells were fixed with 4% paraformaldehyde at 4 °C for 1 h. After washing with PBS, fixed cells were resuspended in PBS at a concentration of approximately 1 × 104 cells/mL. Five microliters of the cell suspension was introduced into the microfluidic device and trapped onto the micromesh by applying 1.5 kPa of negative pressure. After cell entrapment, prewarmed permeabilization solution (50% (v/v) ethanol/PBS) was loaded into the microfluidic device to allow penetration of mRNA-specific oligonucleotide probes into the cells. The microfluidic device was mounted on a hot plate and incubated at 60 °C for 45 min. To wash the permeabilization solution, 200 µL of PBS was loaded into the microchannel at a flow rate of 20 µL/min. Fluorescence-labeled oligonucleotide probes for FISH were designed to target human β-actin mRNA, as reported previously.31 The oligonucleotide probe with the following sequence was used: 5′-GAA GCT GTA GCC GCG CTC GGT GAG GAT CTT CAT GAG GTA GTC AGT CAG-3′. The probe was purchased from Operon Biotechnologies Inc., Tokyo, Japan and labeled with Cy3 at the 5′-end. To perform probe hybridization, a hybridization buffer (5× SSC, 0.1% Tween-20) containing the probe at a final concentration of 1 µM was loaded through the microchannel and allowed to react with the permeabilized cells. Both the inlet and outlet of the microfluidic device were sealed, and the device was set in a humidified chamber. Probe-cell hybridization was carried out at (29) De Clerck, L. S.; Bridts, C. H.; Mertens, A. M.; Moens, M. M.; Stevens, W. J. J. Immunol. Methods 1994, 172, 115–124. (30) Papadopoulos, N. G.; Dedoussis, G. V.; Spanakos, G.; Gritzapis, A. D.; Baxevanis, C. N.; Papamichail, M. J. Immunol. Methods 1994, 177, 101– 111. (31) Holt, D. J.; Bachus, S. E.; Hyde, T. M.; Wittie, M.; Herman, M. M.; Vangel, M.; Saper, C. B.; Kleinman, J. E. Biol. Psychiatry 2005, 58, 408–416.

Analytical Chemistry, Vol. 80, No. 13, July 1, 2008

5141

Figure 2. Confocal fluorescence microscopic images of PET micromesh: (A) clear PET; (B) black PET. X-1, confocal image; X-2, fluorescence intensity profile on the yellow line. Confocal images were captured using a confocal fluorescence microscope with an excitation wavelength of 559 nm and a band-pass filter within the range of 575-675 nm. Scale bars are 30 µm.

42 °C for 2 h. After hybridization, 500 µL of prewarmed hybridization buffer was loaded into the microchannel at a flow rate of 20 µL/min to wash out excess probes. The fluorescence images of the labeled cells were captured using a cooled EM-CCD camera (C9100-02, Hamamatsu Photonics K. K., Shizuoka, Japan) under a specific filter set for Cy3 with a 488 ms or 1.65 s exposure. Intensity data from the EM-CCD was stored at 14-bit gray scale levels in the original NAF format. The obtained single-cell intensity data were analyzed using the AQUACOSMOS software (Hamamatsu Photonics K. K.), and the cutoff/threshold value was set by taking into account background signals from the substrate. As the control, cells were treated with RNase (DNase free, Wako Pure Chemical Industries, Ltd., Osaka, Japan) at a concentration of 100 µg/mL in PBS before hybridization. For serum starvation experiments, Raji cells maintained with RPMI 1640 medium containing 10% FBS were pelleted by centrifugation, washed with PBS, and subsequently incubated with medium containing 0.5% FBS for 24 h. Serum-supplied cells were initially stained with calcein-AM as described earlier and mixed with the serum-starved Raji cells at equal amounts. In contrast to the serum-supplied cells, the serum-starved cells were not stained with calcein-AM. The cell suspension mixture was introduced into the microfluidic device and trapped onto the micromesh. The trapped cells were stained by FISH and the fluorescence intensity of each cell was estimated by image analysis. All reagents were commercially available and purchased as analytical- or laboratory-grade materials. Ultrapure water, incubated with diethyl pyrocarbonate (DEPC) for 12 h and autoclaved to inactivate RNase, was used in all the FISH experiments. RESULTS Entrapment of Mammalian Cells on Various Micromeshes. In a previous report, an SUS micromesh was used for entrapment of protozoan oocysts.28 In this report, a PET plate was used as the alternate substrate for micromesh fabrication (Supporting Information Figure S-1). Nontreated PET is known to exhibit autofluorescence. The edges of the microcavities fabricated against a clear PET substrate produced fluorescence under most filter sets (Supporting Information Figure S-2). As an alternative, black PET, a substrate that is dispersed with black pigment, was employed. As illustrated in Figure 2, the black PET substrate exhibited no fluorescence as compared to the clear PET substrate. 5142

Analytical Chemistry, Vol. 80, No. 13, July 1, 2008

We investigated the optimum amount of negative pressure applied through the vacuum line. The black PET micromesh (diameter of the microcavity, 5 µm) was integrated with a vacuum line, and 5 µL of cell suspension was directly dropped onto the upper surface of the micromesh. Cells were driven toward and trapped onto the microcavities of the micromesh when negative pressure was applied through the vacuum line. When over 10 kPa of negative pressure was applied, most of the cells not only ruptured through the micromesh, but some cells were observed to have been damaged, resulting in cell disruption. When negative pressure of lower than 10 kPa was applied, cells were successfully trapped and retained on the micromesh. Furthermore, various microcavity sizes were fabricated to achieve high cell-entrapment efficiency. The diameters of the microcavities varied from 2, 5, and 10 µm while the distance between each microcavity remained constant at 30 µm. The difference between the diameters of the microcavities resulted in a large difference in the number of trapped cells (Supporting Information Figure S-3). Over 83% (N ) 65) of the introduced cells were trapped on the 2 µm microcavity micromesh, and cells were not observed to pass through the microcavities. When the diameter of the microcavities increased over 2 µm, some of the cells passed through the microcavities, resulting in a decrease in the number of trapped cells. As a result, 63% of the cells were trapped on the 5 µm microcavity micromesh, whereas only 35% of the cells were trapped on the 10 µm microcavity micromesh. Prevention of Nonspecific Cell Adsorption on the Microchannel in the Microfluidic Device. PDMS has a hydrophobic property that strongly interacts with the hydrophobic domain of introduced cells, causing significant loss in the trapping efficiency during separation processes. To prevent the nonspecific adsorption of the cells onto the microchannel, the PDMS surface was modified. A microchannel, comprising a single straight channel, was designed and used for this purpose. The surface of PDMS was modified from hydrophobic to hydrophilic by treatment with air plasma and surfactant. Pluronic F-127, a nonionic triblock copolymer surfactant, which consisted of two hydrophilic polyethylene oxides and one hydrophobic polypropylene oxide, was used for this treatment in which the PDMS surface was treated with the surfactant immediately after plasma treatment. The contact angle of the polymer-coated PDMS surface was subsequently measured to estimate its surface property. The contact angle of the untreated PDMS surface was 105.8° ± 0.9°. Surface treatment with Pluronic F-127 was performed after air plasma treatment decreased the contact angle to less than 30°. When cells were introduced into the air plasma and Pluronic F-127 treated microchannel, the amount of cells adsorbed on the surface of the microchannel was markedly decreased without the observation of cell lysis (Supporting Information Figure S-4). A relationship between the concentration of Pluronic F-127 and the number of adsorbed cells is illustrated in Figure 3. The concentration was varied between 0.1% and 10%. When the PDMS surface was treated with greater than 1% Pluronic F-127, the number of cells that adsorbed onto the microchannel dramatically decreased. Nonspecific cell adsorption onto the microchannel was decreased to less than 1%, when the PDMS surface was treated with 10% Pluronic F-127. On the basis of these results, treatments with air

Figure 3. Number of Raji cells adsorbed on the PDMS microchannel and its contact angle after treatment with Pluronic F-127 after air plasma treatment. A total of 1500 cells was introduced into the PDMS microchannel.

Figure 4. Evaluation of cell trapping efficiency using the microdevice equipped with micromesh; Raji cells were trapped by applying 1.5 kPa of negative pressure: b, black PET micromesh; O, SUS micromesh; dotted line, 100% cell trapping efficiency.

plasma and 10% Pluronic F-127 solution for 2 h were used for the subsequent experiments. Entrapment of Mammalian Cells on the Micromesh in the Microfluidic Device. The PDMS microchannel integrated with the SUS or black PET micromesh pretreated with 10% Pluronic F-127 solution was employed for the entrapment of mammalian cells. A suspension of Raji cells was introduced into the microchannel from the inlet, negative pressure was applied from the vacuum line to entrap cells onto the micromesh, and the number of trapped cells was counted using a fluorescence microscope. Cells were successfully trapped onto the SUS or black PET micromesh within approximately 3 min. Figure 4 shows the relationships between the number of cell injected and trapped cells on the SUS and black PET micromeshes. When the cells were trapped using the microfluidic device integrated with the SUS micromesh (diameter of the microcavity: 2 µm), the maximum entrapment efficiency was 58%. In contrast, when the cells were trapped using the microfluidic device integrated with the black PET micromesh (diameter of the microcavity: 2 µm), the maximum entrapment efficiency was 79%. It is also understood that

the microcavities of the SUS micromesh are nonuniform in size, resulting in the possibility of cells to pass through the microcavities of the SUS micromesh. To estimate and analyze the effects of pressure and its damage to the cells trapped on the micromesh, a cell viability test was performed. In this study, a population of Raji cells with cell viability of 90% ± 3%, which was measured with calcein-AM and PI, was trapped onto the black PET micromesh, and cell viability dropped to 73% ± 2% after staining with PI. This result indicated that the cell-entrapment technique introduced in this work may have inflicted slight damage to the cell membrane of the trapped cells. Fluorescence in Situ Hybridization on the Micromesh in the Microfluidic Device. For single-cell mRNA expression analysis, Cy3-labeled oligonucleotide probes specific to the β-actin mRNA were hybridized to β-actin mRNA in Raji cells in the microtubes or on the microfluidic device integrated with the black PET micromesh. The fluorescence intensity of each hybridized cell was compared in order to optimize the FISH conditions against the cells trapped onto the micromesh. For establishment of optimal hybridization conditions of the probes to the target mRNA in the cells, the permeabilization condition of trapped cells needed to be addressed. First, permeabilization conditions were examined in the microtubes for efficient probe penetration. Without heat treatment, fluorescence intensity of the stained cells was similar to nontreated cells. Fluorescence intensities of cells that were treated in 50% ethanol/PBS for 45 min at 60 °C exhibited significant increase in comparison to non-heat-treated cells. Furthermore, upon RNase treatment, no fluorescence signal was observed in the heat-treated cells. This result indicated that the fluorescence signal was derived from the target mRNA. Upon optimization of probe-cell hybridization in the microtubes, similar conditions have been employed for FISH against cells trapped onto the micromesh. Raji cells treated with 4% paraformaldehyde were introduced into the microfluidic device, trapped onto the black PET micromesh, and treated with 50% ethanol/PBS at 60 °C for 45 min to allow penetration of probes into the trapped cells. Permeabilized cells were then incubated with fluorescence-labeled probes for β-actin mRNA. Fluorescence intensity of cells on the micromesh was sufficient for the observation of successful probe hybridization based on optimized conditions (Figure 5A). In contrast, no significant signal was observed from cells that were not treated for permeabilization (Figure 5B). This result indicated the successful visualization of β-actin mRNA in individual cells by the microfluidic FISH method (Supporting Information Figure S-5). To demonstrate that this method can contribute to single-cell analysis, human β-actin mRNA expression levels of serum-supplied cells were measured against those of serum-starved cells. The gene encoding for the β-actin protein is a serum-responsive gene, and its expression is induced by serum supply.10,11,32 In this study, evaluation of the β-actin gene expression levels in serum-supplied and serum-starved Raji cells was demonstrated in the microfluidic device. A solution containing an equal number of serum-supplied cells stained with calcein-AM and nonstained serum-starved cells were prepared and loaded into the microfluidic device. Positions of both cells entrapped on the micromesh were determined by detecting fluorescence from calcein-AM. After the permeabilization (32) Herschman, H. R. Annu. Rev. Biochem. 1991, 60, 281–319.

Analytical Chemistry, Vol. 80, No. 13, July 1, 2008

5143

Figure 5. Fluorescence microphotographs of β-actin mRNA in Raji cells trapped on the black PET micromesh. Raji cells were introduced into the microfluidic device and trapped on the black PET micromesh, and on-chip FISH was performed. (A) FISH was performed after permeabilization treatment of cells. Permeabilization on the microfluidic device was carried out by introducing 50% ethanol/PBS to the microchannel by setting the device on a hot plate prior to hybridization. (B) FISH was performed without permeabilization treatment. Scale bars are 50 µm.

mRNA expression was induced in serum-supplied cells where a broader distribution and higher fluorescence intensity was observed. In contrast, the distribution of histogram peaks at lower fluorescence intensity was observed from the serum-starved cells. The calculated average fluorescence intensity of both types of cells further supports that β-actin mRNA expression was higher in serum-supplied cells than in serum-starved cells.

Figure 6. Evaluation of the β-actin gene expression levels in serumsupplied and serum-starved cells. Raji cells suspension (mixture of serum-supplied (n ) 28) and serum-starved (n ) 38) cells) was introduced into the microfluidic device and trapped on the black PET micromesh. Subsequently, FISH for human β-actin mRNA was performed on the microfluidic device. Following this, the fluorescence intensity of single cells was measured. The histograms present the number of individual cells expressing β-actin mRNA within the indicated fluorescence intensity range. (A) Serum-supplied cells. The inset illustrates a fluorescence microscopic image of a serum-supplied single cell exhibiting a fluorescence intensity of 10 537. (B) Serumstarved cells. The inset illustrates an image of a serum-starved single cell exhibiting a fluorescence intensity of 4790. Scale bars are 20 µm.

treatment and hybridization, fluorescence intensities from Cy3labeled probes in individual cells were analyzed. Figure 6 shows the histogram of the fluorescence intensity of serum-supplied and serum-starved Raji cells hybridized with fluorescence-labeled probes for β-actin mRNA. This result demonstrates that β-actin 5144

Analytical Chemistry, Vol. 80, No. 13, July 1, 2008

DISCUSSION Array-based single-cell analysis systems based on hydrodynamic, microwell,33–35 and obstacle entrapment16 techniques have been reported. These systems enable the entrapment and analysis of a large number of cells on a single array with the aid of a microscope. However, due to their low cell-entrapment efficiencies, these systems have required the introduction of an excessive number of cells in order to completely fill the arrays with single cells. In contrast, our system indicated high entrapment efficiencies of 79% by applying negative pressure to the black PET micromesh in the device. Furthermore, cell entrapment was achieved in approximately 3 min. The design of the microfluidic device, which allows the visualization of the entire micromesh under a single microscopic observation, demonstrated that each of the trapped cells occupied a single microcavity. Although slight cell damage was observed, many cells retained their cellular integrity and were applicable in subsequent analyses. We demonstrated the ability to stain trapped cells by loading fluorescent dye in the microchannel as shown in the cell viability test in which trapped Raji cells were stained with PI to determine the survival of cells on the micromesh. This suggests that trapped cells can be exposed to various reagents by introducing the reagents into the microchannel, and various analyses, such as the determination of cell activity in individual cells via the measurements of calcium flux using an ion indicator, can be performed.36 Furthermore, high efficient entrapment of single cells onto the micromesh also shows (33) Rettig, J. R.; Folch, A. Anal. Chem. 2005, 77, 5628–5634. (34) Yamamura, S.; Kishi, H.; Tokimitsu, Y.; Kondo, S.; Honda, R.; Rao, S. R.; Omori, M.; Tamiya, E.; Muraguchi, A. Anal. Chem. 2005, 77, 8050–8056. (35) Love, J. C.; Ronan, J. L.; Grotenbreg, G. M.; van der Veen, A. G.; Ploegh, H. L. Nat. Biotechnol. 2006, 24, 703–707. (36) Wheeler, A. R.; Throndset, W. R.; Whelan, R. J.; Leach, A. M.; Zare, R. N.; Liao, Y. H.; Farrell, K.; Manger, I. D.; Daridon, A. Anal. Chem. 2003, 75, 3581–3586.

promising applications in target or rare cell analysis using fluorescence labeling by specific antibodies against cell surface markers. This entrapment method offers a platform for highthroughput single-cell analysis such as drug screening, genetic heterogeneity, and cell-cell signaling. FISH for human β-actin mRNA detection in cells was demonstrated using the microfluidic device equipped with a black PET micromesh. Fluorescence intensity of cells on the micromesh was sufficient for the observation of successful probe hybridization based on the optimized conditions. The β-actin gene used in this work is a housekeeping gene commonly used as a reference to determine the expression level of specific genes. The successful FISH analysis of the β-actin gene with this system further demonstrates that by performing concomitant analysis of specific genes with the β-actin gene, relative quantification can also be performed. Moreover, cell-to-cell variation was investigated in serumresponsive β-actin mRNA expression by using this microfluidic FISH system. Serum starvation suppresses cell proliferation and synchronizes cell growth cycle at the G0 phase.37 For this reason, serum-starved cells showed a narrow distribution at low fluorescence intensities. On the other hand, the population of the serumsupplied cells includes cells in various phases. Hence, the β-actin mRNA expression of serum-supplied cells showed a broader distribution as compared to serum-starved cells. Thus, cell-to-cell variation in β-actin mRNA expression was successfully demonstrated using this microfluidic device. Entrapment of single cells, permeabilization, hybridization, washing, and measurement of mRNA expression levels were implemented in a microfluidic assay within one integrated device. Therefore, it is possible to automate these steps and develop a system that can measure mRNA expression levels in individual cells from a cell population. Furthermore, simultaneous measurement of mRNA expression levels of up to 100 single cells and analysis of the cell-to-cell variation of gene expression levels are (37) Zhou, J.; Chau, C. M.; Deng, Z.; Shiekhattar, R.; Spindler, M. P.; Schepers, A.; Lieberman, P. M. EMBO J. 2005, 24, 1406–1417.

possible by this microfluidic FISH system using this novel microfluidic device. CONCLUSIONS In this study, we have developed a microfluidic device equipped with black PET micromesh for the entrapment of mammalian cells. Cell adsorption was prevented by treating the PDMS surface of the microchannel with air plasma and Pluronic F-127 while each microcavity was fabricated to ensure that only one cell is trapped on a single microcavity, and cells introduced into the microfluidic device were trapped onto the black PET micromesh with high efficiency. In addition, FISH could be directly performed against cells that have been trapped onto the black PET micromesh, and we could successfully identify cell-to-cell variations in β-actin mRNA expression of serum-supplied and serum-starved cells. On the basis of these specific features of our microfluidic device, a simple yet specific trapping technique for the high-throughput measurement of specific mRNA expression levels was introduced. It is believed that the development of this system would widely contribute to single-cell analysis. ACKNOWLEDGMENT This work was supported by a Grant-in-Aid for Scientific Research on Priority Areas “Lifesurveyor”, No. 17066002 from the Ministry of Education, Culture, Sports, Science and Technology of Japan and by the Human Resource Development Program for Scientific Powerhouse from the Ministry of Education, Culture, Sports, Science and Technology of Japan granted to Masahito Hosokawa. SUPPORTING INFORMATION AVAILABLE Additional information is available as noted in text. This material is available free of charge via the Internet at http://pubs. acs.org. Received for review February 19, 2008. Accepted April 29, 2008. AC800352J

Analytical Chemistry, Vol. 80, No. 13, July 1, 2008

5145