- Email: [email protected]
On-chip quantitative detection of pathogen genes by autonomous microfluidic PCR platform
Biosensors and Bioelectronics 74 (2015) 725–730
Contents lists available at ScienceDirect
Biosensors and Bioelectronics journal homepage: www.elsevier.com/locate/bios
On-chip quantitative detection of pathogen genes by autonomous microfluidic PCR platform Hiroaki Tachibana a,b, Masato Saito a,n, Shogo Shibuya b, Koji Tsuji b, Nobuyuki Miyagawa b, Keiichiro Yamanaka a, Eiichi Tamiya a a b
Department of Applied Physics, Graduate School of Engineering, Osaka University, 2-1 Yamadaoka, Suita, Osaka 575-0871, Japan Eco Solutions Company, Panasonic Corporation, 1048 Kadoma, Kadoma, Osaka 571-8686, Japan
art ic l e i nf o
a b s t r a c t
Article history: Received 14 March 2015 Received in revised form 24 June 2015 Accepted 4 July 2015 Available online 9 July 2015
Polymerase chain reaction (PCR)-based genetic testing has become a routine part of clinical diagnoses and food testing. In these fields, rapid, easy-to-use, and cost-efficient PCR chips are expected to be appeared for providing such testing on-site. In this study, a new autonomous disposable plastic microfluidic PCR chip was created, and was utilized for quantitative detection of pathogenic microorganisms. To control the capillary flow of the following solution in the PCR microchannel, a driving microchannel was newly designed behind the PCR microchannel. This allowed the effective PCR by simply dropping the PCR solution onto the inlet without any external pumps. In order to achieve disposability, injection-molded cyclo-olefin polymer (COP) of a cost-competitive plastic was used for the PCR chip. We discovered that coating the microchannel walls with non-ionic surfactant produced a suitable hydrophilic surface for driving the capillary flow through the 1250-mm long microchannel. As a result, quantitative real-time PCR with the lowest initial concentration of human, Escherichia coli (E. coli), and pathogenic E. coli O157 genomic DNA of 4, 0.0019, 0.031 pg/μl, respectively, was successfully achieved in less than 18 min. Our results indicate that the platform presented in this study provided a rapid, easy-to-use, and low-cost real-time PCR system that could be potentially used for on-site gene testing. & 2015 Elsevier B.V. All rights reserved.
Keywords: Continuous-flow polymerase chain reaction Quantitative real-time PCR Microfluidic chip Capillary force On-site pathogen detection
1. Introduction Polymerase chain reaction (PCR) plays a central role in genetic analysis since its invention in the mid-1980s by Kary Mullis (Mullis and Falona, 1987). PCR allows the detection of low levels of pathogens by amplifying their deoxyribo nucleic acid (DNA) with high accuracy and precision. Therefore, PCR is used for diagnosing infectious diseases, for testing bacterial food poisoning, and in biodefense (Christopoulos, 1999). TaqMan-based quantitative realtime PCR allows simultaneous detection of target sequences by using primers and dual-labeled fluorogenic probes. Real-time PCR has been used to detect many bacterial pathogens, such as Escherichia coli (E. coli) (Frydendahl et al., 2001; Ståhl et al., 2011; Silkie et al., 2008), pathogenic E. coli O157 (Call et al., 2001; Lee n
Corresponding author. Fax: þ 81 668797840. E-mail addresses: [email protected] (H. Tachibana), [email protected] (M. Saito), [email protected] (S. Shibuya), [email protected] (K. Tsuji), [email protected] (N. Miyagawa), [email protected] (K. Yamanaka), [email protected] (E. Tamiya). http://dx.doi.org/10.1016/j.bios.2015.07.009 0956-5663/& 2015 Elsevier B.V. All rights reserved.
et al., 2006), Salmonella (Lee et al., 2006), Staphylococcus aureus (Lee et al., 2006), Campylobacter jejuni (Best et al., 2003), and Campylobacter coli (Best et al., 2003). Although PCR is a useful diagnostic tool because of its high sensitivity and accuracy, it is time-consuming when performed using laboratory-scale thermal cyclers. Kopp et al. (1998) and coworkers demonstrated the first continuous-flow PCR (CF-PCR) by using a microfluidic chip for highspeed amplification of target sequences. Three temperature zones, corresponding to denaturation, annealing, and extension temperatures, respectively, were established in a glass-based microfluidic chip by using temperature-fixed heaters. Only the PCR solution underwent thermal cycling in single fluidic microchannel among 3 temperature zones. This allowed the PCR solution to switch rapidly among different temperatures compared with that in conventional thermal cyclers. After this epochal research, some related studies were performed to improve the design and efficacy of microfluidic chips such as tuning the cross-sectional size of the microchannels (Li et al., 2006; Cao et al., 2011),a short cycle length (Crews et al., 2008a, 2008b). Reverse transcription PCR (RT-PCR) targeting an influenza virus (Yamanaka et al., 2011), and integration of PCR chips with fluorescence detection systems (Obeid and Christopoulos, 2003; Nakayama et al., 2006, 2010) were also
726
H. Tachibana et al. / Biosensors and Bioelectronics 74 (2015) 725–730
performed. However, these modifications still required the use of complex external pumps such as syringe pumps to control the flow of PCR solution in the PCR microchannel. To eliminate the use of external pumps, we developed a capillary-driven, self-propelled CF-PCR (SP-CF-PCR) microfluidic chip (Tachibana et al., 2015). The PCR microchannel was prepared using a Si substrate and was covered by a glass substrate. The surface of the microchannel was oxidized to form a hydrophilic film. PCR solution dropped onto the inlet of the microfluidic chip was transported through the microchannel by capillary flow to achieve the desired amplification. The theoretical approach of capillary flow involving more than one temperature was developed based on previous studies (Juncker et al., 2002; Delamarche et al., 1998, 2005; Zhu and Petkovic-Duran, 2010). We used this SP-CF-PCR microfluidic chip to amplify the genomic DNA of some pathogens. However, the amplification was inefficient in the head area of the flow because of the evaporation of PCR solution or adsorption of reagents. Moreover, the use of Si for preparing microfluidic chips is associated with 2 major disadvantages. One of the disadvantages is longer intermediate temperature zone between the high and low temperature zones due to the high thermal conductivity of Si. This intermediate temperature zone increases the microchannel length and decrease PCR efficiency at an unnecessary temperature. Another disadvantage is its high cost. The material used for making microfluidic chips should be cost-competitive to allow mass production for on-site use such as screening of pathogens. From this point of view, plastics are very promising for developing microfluidic chips because they have low thermal conductivity and are inexpensive. However, one of the difficulties of
using plastics for developing SP-CF-PCR chips is the achievement of hydrophilic surface on the microchannel walls because most plastics are hydrophobic. Some dry processes such as plasma (Gervais et al., 2011) or ultraviolet (UV) ozone (Novo et al., 2011, 2013) treatment have been used to activate and deliver hydrophilic groups on the surface of the microchannel walls in polydimethylsiloxane (PDMS) microfluidic chips. However, because the activated surfaces are unstable (Eddington et al., 2006), hydrophilicity disappears during the subsequent fabrication process, and because of time-associated degradation or heating. Therefore, some active hydrophilic layers should be developed on the microchannel walls. In the present study, a plastic-based microfluidic real-time quantitative PCR chip was developed as shown in Fig. 1(a). A driving microchannel, which was placed behind the PCR microchannel, was designed to exclude the head area of the flow from the PCR microchannel and to maintain capillary force for controlling the flow. The microchannel of the SP-CF-PCR chip was formed in a cyclo-olefin polymer (COP) plate by injection molding. Water-soluble polymers and surfactants were examined for producing the hydrophilic layer on the microchannel walls. Moreover, the PCR microfluidic chip was integrated with a fluorescence detection system containing a laser and photomultiplier tube (PMT) to analyze the amplified product within the chip. For verifying our concept, human β-actin, E. coli DNA, and E. coli O157 DNA were amplified using the new SP-CF-PCR chip and the fluorescence detection system.
Fig. 1. Design of the self-propelled continuous-flow PCR (SP-CF-PCR) chip. (a) Concept diagram. The PCR chip is placed on 2 heater blocks having temperatures 95 °C and 60 °C, respectively. The PCR solution dropped onto the inlet hole is loaded into the microchannels autonomously by capillary force. The driving microchannel placed behind the PCR microchannel maintains the capillary force to control the following flow of the solution. (b) Simulated temperature distribution in the COP (filled circle with solid line) and glass/Si (Tachibana et al., 2015) (filled triangle with dashed line) microfluidic chips in planes where the microchannels are formed. The microchannels were formed between a lower and an upper plate. The lower plates were attached to the heater blocks. The origin of displacement was set at the center between the high and low temperature zones on the left (negative direction) and right (positive direction), respectively. The block heaters were fixed with 1-mm gap, and the temperatures at both the ends of the chip were fixed at 99 °C and 60 °C, respectively.
H. Tachibana et al. / Biosensors and Bioelectronics 74 (2015) 725–730
2. Materials and methods 2.1. Thermal simulation analysis The cross-sectional temperature distributions in planes in which the microchannels were formed were simulated using finite element method (FEM) with ANSYS©. The microchannels were formed between upper and lower plates of 20 mm-width. The lower plate was attached on 2 aluminum heater blocks whose width and height were both 15 mm. The blocks were fixed with a 1-mm gap. In the thermal simulations, the blocks were considered as heating element and were heated in order that the temperatures at both the ends of the chip became fixed at 99 °C and 60 °C, respectively. Si, glass and COP plates whose thermal conductivities are 1.5 102, 1.0, 0.60 W/(m K), respectively, were used in the simulations. 2.2. Chip fabrication The PCR chip included two 57 25 mm2 COP plates (Zeon Corp., Tokyo, Japan). One of the plates was 1.5-mm thick and included 150-μm deep and 150-μm wide rectangular cross-sectional microchannels. The other one was a 0.7-mm thick flat cover plate. The PCR chip was fabricated as follows. First, the COP plate with the microchannels was fabricated by injection molding. The pattern of the microchannels was fabricated in Si substrate by using micromachining technologies as a mold. The microchannels were formed by photolithography and deep etching process (Tachibana et al., 2015). Note that the microchannels were convex shape. The injection molding of the COP plate using this Si mold was outsourced (Richel Corp., Toyama, Japan). The inlet and drain holes were punched at both the ends of the microchannels. The plate was then attached to the cover plate by heating and applying pressure by exposing the bonding surfaces of the plates to O2 plasma for surface cleaning and activation. After fabricating the chip structure, a hydrophilic layer was produced on the microchannel walls as follows. The microchannels were filled with a solution containing the hydrophilic medium by using a syringe. Next, the solvent inside the microchannel was evaporated. Three solutions, namely, 1% (w/w) solution of water-soluble polymer carboxymethyl cellulose (CMC) (Wako Pure Chemical Industries, Ltd. Osaka, Japan) in distilled water, 1% (w/w) solution of anion surfactant sodium cocoyl sarcosinate (SCS) (Soypon SCE, Kawaken Fine Chemicals Co. Ltd., Tokyo, Japan) in ethanol, and 1% (w/w) solution of non-ionic surfactant polyoxyethylene (20) sorbitan monolaurate (Tween 20) (Wako Pure Chemical Industries, Ltd. Osaka, Japan) in ethanol were used for coating the microchannel walls in separate chips. 2.3. PCR reagents Genomic DNA of humans, E. coli, and pathogenic E. coli O157 were amplified by performing quantitative PCRs with the SP-CFPCR chip. As a common reagents of all the targets,10 fast buffer I (SpeedSTARs HS DNA Polymerase kit; Takara Bio Inc., Tokyo, Japan), 0.2 mM dNTP mixture (Takara Bio Inc.), 0.15 U/μl SpeedSTAR HS DNA Polymerase (Takara Bio Inc.), 1 μg/μl bovine serum albumin (BSA) were used. The PCR mixture for amplifying human genomic DNA included 300 nM each of forward and reverse primers, and TaqMan fluorescence probe containing fluorescein amidite (FAM) (TaqMan βActin Detection Reagents kit; Life Technologies Inc., Carlsbad, USA). Human genomic DNA concentrations of 4, 40, 400, and 3200 pg/μl (Life Technologies Inc.) were used as the templates. The mixture for amplifying E. coli genomic DNA included 10 fast buffer I, 0.2 mM dNTP mixture, 0.15 U/μl SpeedSTAR HS DNA
727
Polymerase, and 1 μg/μl BSA. The mixture also included 300 nM each of forward (5′-CGGAAGCAACGCGTAAACTC-3′), and reverse (5′-TGAGCGTCGCAGAACATTACA-3′) primers, and 300 nM TaqMan probe (5′-FAM-CGCGTCCGATCACCTGCGTC-BHQ1-3′) targeting uidA (Silkie et al., 2008). Genomic DNA extracted from E. coli DH5α (Takara Bio Inc.) by using PureLink Genomic DNA Mini Kit (Life Technologies Inc.) was used as the template. The mixture for amplifying E. coli O157 genomic DNA included 10 fast buffer I, 0.2 mM dNTP mixture, 0.15 U/μl SpeedSTAR HS DNA Polymerase, and 1 μg/μl BSA. In addition, the mixture included 300 nM each of forward (5′-CAATTTTTCAGGGAATAACATTG-3′), reverse (5′-AAAGTTCAGATCTTGATGACATTG-3′) primers, and 300 nM TaqMan probe (5′-FAM-TCAAGAGTTGCCCATCCTGCAGCAA-BHQ1-3′) targeting eaeA (Call et al., 2001). Genomic DNA of E. coli O157 extracted from GTC 01061 was obtained from Gifu University, School of Medicine, Pathogenic Microorganism Genetic Resource Stock Center (GMGC) and used as the template. 2.4. On chip quantitative PCR The PCR chip was placed on two 15 15 80-mm3 aluminum heater blocks whose temperatures were controlled individually. The aluminum heater blocks were set in parallel with 1-mm gap and included cartridge heaters and thermocouples (Kyushu-Nissho Co. Ltd, Fukuoka, Japan). One of the heaters was controlled at denaturation temperature while the other was controlled at annealing and extension temperatures by using proportional-integral-derivative controller (Shimaden Co. Ltd., Tokyo, Japan). 50 μl of PCR solution was dropped on the inlet hole. Displacement of capillary flow was determined from the time of PCR solution loading by visually monitoring the position of the fluid front until all the microchannels were filled. Once the microchannels were filled, optical component was scanned over the PCR microchannel, and fluorescence of the solution was detected. The detail of the fluorescence detection system and the analysis of the amplification curves with fluorescence are described in Supporting information on page 2–3 and Fig. S1.
3. Results and discussion 3.1. Design of the PCR microfluidic chip One drawback of the SP-CF-PCR is the decrease in PCR efficiency at the fluid front (Tachibana et al., 2015). The fluid front of the PCR solution is continuously in contact with the fresh surface of the microchannel. Therefore, adsorption of PCR reagents such as polymerase or amplified DNA occurs ineluctably (Nakayama et al., 2006). Moreover, evaporation of the PCR solution occurs at the fluid front because it is constantly exposed at high temperature. To overcome these issues, we newly designed a driving microchannel that was placed behind the PCR microchannel (Fig.1). The velocity of capillary flow in CF-PCR is described as (Tachibana et al., 2015)
v (x) =
R h2 Pc (x) x
8 ∫0 η (x‵) dx‵
(1)
According to Eq. (1), the velocity of the capillary flow decreases as the solution flows through the microchannel but never becomes 0. Therefore, the driving microchannel controls the capillary forces required to maintain the flow of the PCR solution even after the PCR microchannel is filled. Because of the driving microchannel, the head area of the flow, with the high initial flow velocity and low PCR efficiency, can be removed to the driving microchannel.
728
H. Tachibana et al. / Biosensors and Bioelectronics 74 (2015) 725–730
Thus, the driving microchannel allows in achieving a stable flow velocity and high PCR efficiency. The SP-CF-PCR chip having meander shaped PCR microchannel was set on 2 aluminum block heaters in order to form 2 temperature zones in the same plane. This results in the ineluctable formation of an intermediate temperature zone between the 2 temperatures zones. A longer intermediate zone not only increases the total microchannel length but also decreases PCR efficiency at an unnecessary temperature. The intermediate temperature zone depends largely on the thermal conductivity of the PCR chips. In this study, intermediate temperature zones were investigated by comparing between a glass/Si chip having high thermal conductivity (Tachibana et al., 2015) and a COP chip having low thermal conductivity. For evaluating the intermediate temperature zones, cross-sectional temperature distributions in planes in which the microchannels were formed were simulated using FEM analysis. Fig. 1(b) shows the results of the comparison between the glass/Si PCR chip and the COP PCR chip. The origin of displacement was set at the center between the high and low temperature zones, which were placed on the left (negative direction) and right (positive direction), respectively. The microchannels of the glass/Si PCR chip were formed between a 0.5-mm thick Si lower plate and a 0.5-mm thick glass upper plate. The microchannels of the COP PCR chip were formed between a 0.7mm thick lower COP plate and a 1.5-mm thick upper plate. High and low temperature zones were defined as 95–99 °C for denaturation and 60–65 °C for annealing and extension, respectively. Hence, the temperature of the intermediate zone was between 65 °C and 95 °C. The intermediate temperature zone in the glass/Si PCR chips was 7.8 mm long, which was twice its length in the COP PCR chips (3.6 mm). Although the formation of the microchannels and hydrophilic surfaces on the microchannel walls by thermal oxidation is easier in the glass/Si PCR chips, heat insulation in these chips is more difficult than that in COP chips because the thermal conductivity of Si is two orders of magnitude higher than that of COP. Materials used to make the PCR chip should have low thermal conductivity to allow the temperature gradient to concentrate in the gap between 2 heaters. According to this simulation, the microchannel length in each turn, which corresponded to half the length of each PCR cycle, was set as 7 mm to achieve sufficient length. The designed microchannels were 820 mm (53 cycles) of PCR microchannel and 440 mm of driving microchannel.
3.2. Capillary flow in the microchannels Hydrophilic layers can be developed on hydrophobic surfaces by using water-soluble polymers (Kargl et al., 2012) and surfactants (Aegerter and Menning, 2004) because they contain both hydrophobic and hydrophilic groups. In this study, we examined suitable surface treatments to develop a hydrophilic layer on the surface of COP plates, such as treatment with CMC, a water-soluble polymer; SCS, an anion surfactant; and Tween 20, a non-ionic surfactant. Contact angle of the hydrophilic layer formed on the COP flat plate was 15° with CMC and below 10° with SCS and Tween 20. The hydrophilicity of all the microchannels was maintained for more than 10 days. The positions of the fluid fronts of the capillary flow in PCR chips whose microchannel walls were covered with CMC, SCS, and Tween 20 are shown in Fig. 2. In the PCR chip whose microchannel walls were covered with CMC, the PCR solution was loaded only several millimeters, indicating that the CMC coating on the microchannel walls was non-homogeneous. This could be because CMC was dissolved in distilled water; therefore, CMC solution was localized in the microchannel during evaporation because the surface of the COP plate repelled water. In contrast, PCR solutions were successfully loaded into PCR chips whose microchannel walls were covered with either SCS or Tween 20 because homogeneous coatings could be obtained. Surfactants can be applied easily on the microchannel walls because they contain both hydrophilic and hydrophobic groups (Aegerter and Menning, 2004). Moreover, the surfactants did not agglomerate during evaporation because they were liquid at room temperature. Tween 20 was found to be suitable for producing the hydrophilic layer in this study because the PCR solution reached the drain in less than 20 min. On the other hand, in the SCS-coated chip, the PCR solution could not reach to the drain. These results indicated that surface hydrophilicity of Tween 20 was higher than that of SCS. This could be because Tween 20 has more hydrophilic groups than SCS and because it is generated by polar ether linkage compared with SCS, which only contains a carboxyl group at the end of a long-chain hydrocarbon. In the Tween 20-coated chip, the PCR solution initially followed the PCR microchannel and filled it within 7 min (Fig. 2). The PCR solution then filled the driving microchannel. The capillary force present at the fluid front in the driving microchannel helped the flow of the PCR solution through the PCR microchannel. The total time for filling the microchannels was 18 min in this study. In this
Fig. 2. Characteristics of capillary flow. (a) Position of the fluid front of the PCR solution with respect to time in the PCR chip whose microchannel walls are treated with polyoxyethylene (20) sorbitan monolaurate (Tween 20), a non-ionic surfactant (filled circle with solid line); sodium cocoyl sarcosinate (SCS), an anion surfactant (filled triangle with dashed line); and carboxymethyl cellulose (CMC), a water-soluble polymer (filled rectangles with dotted line). The dashed line at 820 mm indicates the boundary between the PCR and the driving microchannels. (b) Photographs of capillary flow in the PCR chip whose microchannel walls are treated with Tween 20 by using the new coccine dye. The arrows indicate the position of the fluid front.
H. Tachibana et al. / Biosensors and Bioelectronics 74 (2015) 725–730
experiment, the volume per cycle in the PCR microchannel was 0.35 μl. On the other hand, the volume of the meander-shaped driving microchannel was 9.8 μl, which corresponded to the volume of 28 PCR cycles. Therefore, PCR solution of 28 PCR cycle in the fluid front was eventually moved to the driving microchannel. As a result, effective PCR was initiated when the fluid front was at the positions of the 28th cycle. The average flow velocity during effective PCR was 0.87 mm/s, which corresponded to 16 s per cycle. The time for each cycle was sufficient to achieve an efficient PCR. The possibility of controlling the capillary flow by using the driving microchannel was successfully observed in the present study. Other optimal designs such as capillary pump structure (Zimmermann et al., 2007) will enable us to obtain a faster and steady flow velocity. 3.3. Quantitative PCR detection Quantitative real-time PCR was completed once the driving microchannel was filled. Therefore, the PCR time was 18 min as shown in Fig. 2. The amplification curves of the PCR targeting the human β-actin are shown in Fig. 3(a). Clear amplifications were obtained for the initial human genomic DNA concentrations of 4, 40, 400, and 3200 pg/μl. The amplification curves were shifted to the low cycle as the initial concentration of the human genomic DNA increased. Specific amplification of a 295-bp fragment was achieved as a representative result of gel electrophoresis (Fig. S2). The standard curve (threshold cycles (Ct values) versus initial concentrations of human genomic DNA) of the human genomic DNA is shown in Fig. 3(b). The definition of Ct value is described in
729
the Supporting information on page 1–2. As expected, the Ct values had a linear characteristic with respect to the logarithmic values of the initial DNA concentrations. The obtained Ct value including the reproducibility for initial DNA concentration of 4 pg/ μl was 42.2 7 3.2, which indicates that the corresponding coefficient of variance (CV) was 7.5%. The lowest initial concentration of human genomic DNA was 4 pg/μl, which corresponded to approximately 0.6 genome/μl (Doležel et al., 2003). These results indicated that our PCR chip and detection system were applicable for quantitative DNA analysis with sufficient high sensitivity and reproducibility. Next, we performed the quantitative PCR targeting uidA of E. coli with initial genomic DNA concentrations of 0.0019, 0.019, 0.19, and 1.9 pg/μl. The amplification curves and a result gel electrophoresis are shown in Fig. S3. Standard curve of E. coli genomic DNA were obtained from the Ct values obtained for each initial DNA concentration (Fig. 3(c)). The obtained Ct value including the reproducibility for initial DNA concentration of 0.0019 pg/μl was 45.6 72.9, which indicates that the corresponding coefficient of variance (CV) was 6.3%. The lowest initial concentration of E. coli genomic DNA of 0.0019 pg/μl corresponded to approximately 0.4 genome/μl. Finally, we performed quantitative PCR by using the genomic DNA of a pathogenic microorganism. E. coli O157 is one of the most notorious foodborne pathogens that colonizes the intestinal tract of cattle and is present in beef and other related products. The infectivity of E. coli O157 is extremely high; therefore, infection can be caused by only a few hundred cells (Karmali, 2004). Therefore, rapid, highly accurate, and highly sensitive on-site detection is required to prevent an outbreak of food poisoning. Quantitative
Fig. 3. On-chip quantitative real-time PCR results. (a) Amplification curves with fluorescence targeting human β-actin. (b) Standard curve of PCR targeting human β-actin. (c) Standard curve of PCR targeting uidA of E. coli. (d) Standard curve of quantitative real-time PCR targeting eaeA of E. coli O157.
730
H. Tachibana et al. / Biosensors and Bioelectronics 74 (2015) 725–730
PCR targeting eaeA of E. coli O157 was performed with initial genomic DNA concentrations of0.031, 0.31, 3.1, and 31 pg/μl. The amplification curves and a representative result of gel electrophoresis are shown in Fig. S4. Standard curve of E. coli O157 genomic DNA were obtained from the Ct values calculated from the corresponding amplification curves. The obtained Ct value including the reproducibility for initial DNA concentration of 0.031 pg/μl was 41.772.8, which indicates that the corresponding coefficient of variance (CV) was 6.7%.This standard curve indicated that this technique could be used to quantitatively analyze E. coli O157 genomic DNA for practical purpose such as food testing. Moreover, the lowest initial concentration of E. coli O157 genomic DNA was 0.031 pg/μl in this experiment, which corresponded to approximately 6.9 genome/μl, indicating that this technique could detect DNA from less than 500 cells in a practical sample.
4. Conclusions In this study, we have successfully developed an autonomous microfluidic PCR platform for on-chip quantitative detection of pathogen genes. The newly designed driving microchannel enabled us to maintain the flow of the PCR solution. The excellent thermal controllability was obtained in the COP PCR chip. Moreover, we have achieved optimal surface hydrophilicity of the microchannel walls by detailed investigation of non-ionic surfactants. As the results, the PCR solution was successfully transported through the whole microchannels autonomously and quantitative real-time PCR was successfully performed in less than 18 min. Our results indicated that the SP-CF-PCR chips could be used for rapid, low-cost and easy-to-use gene testing in the medical and environmental fields.
Appendix A. Supplementary material Supplementary data associated with this article can be found in the online version at http://dx.doi.org/10.1016/j.bios.2015.07.009.
References Aegerter, M.A., Menning, M., 2004. Sol–gel technologies for glass producers and users. Kluwer Academic Publishers, Norwell. Best, E.L., Powell, E.J., Swift, C., Grant, K.A., Frost, J.A., 2003. FEMS Microbiol. Lett. 229, 237–241. Call, D.R., Brockman, F.J., Chandler, D.P., 2001. Int. J. Food Microbiol. 67, 71–80. Cao, Q., Kim, M., Klapperich, C.M., 2011. Biotechnol. J. 6, 177–184. Christopoulos, T.K., 1999. Anal. Chem. 71, 425R–438R. Crews, N., Wittwer, C., Gale, B., 2008a. Biomed. Microdev. 10, 187–195. Crews, N., Wittwer, C., Palais, R., Gale, B., 2008b. Lab Chip 8, 919–924. Delamarche, E., Bernard, A., Schmid, H., Bietsch, A., Michel, B., Biebuyck, H., 1998. J. Am. Chem. Soc. 120, 500–508. Delamarche, E., Juncker, D., Schmid, H., 2005. Adv. Mater. 17, 2911–2933. Doležel, J., Bartos, J., Voglmayr, H., Greilhuber, J., 2003. Cytometry A 51, 127–128. Eddington, D.T., Puccinelli, J.P., Beebe, D.J., 2006. Sens. Actuators B 114, 170–172. Frydendahl, K., Imberechts, H., Lehmann, S., 2001. Mol. Cell. Probes 15, 151–160. Gervais, L., Hitzbleck, M., Delamarche, E., 2011. Biosens. Bioelectron. 27, 64–70. Juncker, D., Schmid, H., Drechsler, U., Wolf, H., Wolf, M., Michel, B., Rooij, N., Delamarche, E., 2002. Anal. Chem. 74, 6139–6144. Kargl, R., Mohan, T., Bračič, M., Kulterer, M., Doliška, A., Stana-Kleinschek, K., Ribitsch, V., 2012. Langmuir 28, 11440–11447. Karmali, M.A., 2004. Mol. Biotechnol. 26, 117–122. Kopp, M.U., Mello, A.J., Manz, A., 1998. Science 280, 1046–1048. Lee, D.-Y., Shannon, K., Beaudette, L.A., 2006. J. Microbiol. Method 65, 453–467. Li, S., Fozdar, D.Y., Ali, M.F., Li, H., Shao, D., Vykoukal, D.M., Vykoukal, J., Floriano, P. N., Olsen, M., McDevitt, J.T., Gascoyne, P.R.C., Chen, S., 2006. J. Microelectromech. Syst. 15, 223–236. Mullis, K.B., Falona, F.A., 1987. Methods Enzymol. 155, 335–350. Nakayama, T., Kurosawa, K., Furui, S., Kerman, K., Kobayashi, M., Rao, S.R., Yonezawa, Y., Nakano, K., Hino, A., Takamura, Y., Tamiya, E., 2006. Anal. Bioanal. Chem. 386, 1327–1333. Nakayama, T., Hiep, H.M., Furui, S., Yonezawa, Y., Saito, M., Takamura, Y., Tamiya, E., 2010. Anal. Bioanal. Chem. 396, 457–464. Novo, P., Prazeres, D.M.F., Chu, V., Conde, J.P., 2011. Lab Chip 11, 4063–4071. Novo, P., Volpetti, F., Chu, V., Conde, J.P., 2013. Lab Chip 13, 641–645. Obeid, P.J., Christopoulos, T.K., 2003. Anal. Chim. Acta 494, 1–9. Silkie, S.S., Tolcher, M.P., Nelson, K.L., 2008. J. Microbiol. Method 72, 275–281. Ståhl, M., Kokotovic, B., Hjulsager, C.K., Breum, S.Ø., Angen, Ø., 2011. Vet. Microbiol. 151, 307–314. Tachibana, H., Saito, M., Tsuji, K., Yamanaka, K., Hoa, L.Q., Tamiya, E., 2015. Sens. Actuators B 206, 303–310. Yamanaka, K., Saito, M., Kondoh, K., Hossain, M.M., Koketsu, R., Sasaki, T., Nagatani, N., Ikuta, K., Tamiya, E., 2011. Analyst 136, 2064–2068. Zhu, Y., Petkovic-Duran, K., 2010. Microfluid Nanofluid 8, 275–282. Zimmermann, M., Schmid, H., Hunziker, P., Delamarche, E., 2007. Lab Chip 7, 119–125.
Contents lists available at ScienceDirect
Biosensors and Bioelectronics journal homepage: www.elsevier.com/locate/bios
On-chip quantitative detection of pathogen genes by autonomous microfluidic PCR platform Hiroaki Tachibana a,b, Masato Saito a,n, Shogo Shibuya b, Koji Tsuji b, Nobuyuki Miyagawa b, Keiichiro Yamanaka a, Eiichi Tamiya a a b
Department of Applied Physics, Graduate School of Engineering, Osaka University, 2-1 Yamadaoka, Suita, Osaka 575-0871, Japan Eco Solutions Company, Panasonic Corporation, 1048 Kadoma, Kadoma, Osaka 571-8686, Japan
art ic l e i nf o
a b s t r a c t
Article history: Received 14 March 2015 Received in revised form 24 June 2015 Accepted 4 July 2015 Available online 9 July 2015
Polymerase chain reaction (PCR)-based genetic testing has become a routine part of clinical diagnoses and food testing. In these fields, rapid, easy-to-use, and cost-efficient PCR chips are expected to be appeared for providing such testing on-site. In this study, a new autonomous disposable plastic microfluidic PCR chip was created, and was utilized for quantitative detection of pathogenic microorganisms. To control the capillary flow of the following solution in the PCR microchannel, a driving microchannel was newly designed behind the PCR microchannel. This allowed the effective PCR by simply dropping the PCR solution onto the inlet without any external pumps. In order to achieve disposability, injection-molded cyclo-olefin polymer (COP) of a cost-competitive plastic was used for the PCR chip. We discovered that coating the microchannel walls with non-ionic surfactant produced a suitable hydrophilic surface for driving the capillary flow through the 1250-mm long microchannel. As a result, quantitative real-time PCR with the lowest initial concentration of human, Escherichia coli (E. coli), and pathogenic E. coli O157 genomic DNA of 4, 0.0019, 0.031 pg/μl, respectively, was successfully achieved in less than 18 min. Our results indicate that the platform presented in this study provided a rapid, easy-to-use, and low-cost real-time PCR system that could be potentially used for on-site gene testing. & 2015 Elsevier B.V. All rights reserved.
Keywords: Continuous-flow polymerase chain reaction Quantitative real-time PCR Microfluidic chip Capillary force On-site pathogen detection
1. Introduction Polymerase chain reaction (PCR) plays a central role in genetic analysis since its invention in the mid-1980s by Kary Mullis (Mullis and Falona, 1987). PCR allows the detection of low levels of pathogens by amplifying their deoxyribo nucleic acid (DNA) with high accuracy and precision. Therefore, PCR is used for diagnosing infectious diseases, for testing bacterial food poisoning, and in biodefense (Christopoulos, 1999). TaqMan-based quantitative realtime PCR allows simultaneous detection of target sequences by using primers and dual-labeled fluorogenic probes. Real-time PCR has been used to detect many bacterial pathogens, such as Escherichia coli (E. coli) (Frydendahl et al., 2001; Ståhl et al., 2011; Silkie et al., 2008), pathogenic E. coli O157 (Call et al., 2001; Lee n
Corresponding author. Fax: þ 81 668797840. E-mail addresses: [email protected] (H. Tachibana), [email protected] (M. Saito), [email protected] (S. Shibuya), [email protected] (K. Tsuji), [email protected] (N. Miyagawa), [email protected] (K. Yamanaka), [email protected] (E. Tamiya). http://dx.doi.org/10.1016/j.bios.2015.07.009 0956-5663/& 2015 Elsevier B.V. All rights reserved.
et al., 2006), Salmonella (Lee et al., 2006), Staphylococcus aureus (Lee et al., 2006), Campylobacter jejuni (Best et al., 2003), and Campylobacter coli (Best et al., 2003). Although PCR is a useful diagnostic tool because of its high sensitivity and accuracy, it is time-consuming when performed using laboratory-scale thermal cyclers. Kopp et al. (1998) and coworkers demonstrated the first continuous-flow PCR (CF-PCR) by using a microfluidic chip for highspeed amplification of target sequences. Three temperature zones, corresponding to denaturation, annealing, and extension temperatures, respectively, were established in a glass-based microfluidic chip by using temperature-fixed heaters. Only the PCR solution underwent thermal cycling in single fluidic microchannel among 3 temperature zones. This allowed the PCR solution to switch rapidly among different temperatures compared with that in conventional thermal cyclers. After this epochal research, some related studies were performed to improve the design and efficacy of microfluidic chips such as tuning the cross-sectional size of the microchannels (Li et al., 2006; Cao et al., 2011),a short cycle length (Crews et al., 2008a, 2008b). Reverse transcription PCR (RT-PCR) targeting an influenza virus (Yamanaka et al., 2011), and integration of PCR chips with fluorescence detection systems (Obeid and Christopoulos, 2003; Nakayama et al., 2006, 2010) were also
726
H. Tachibana et al. / Biosensors and Bioelectronics 74 (2015) 725–730
performed. However, these modifications still required the use of complex external pumps such as syringe pumps to control the flow of PCR solution in the PCR microchannel. To eliminate the use of external pumps, we developed a capillary-driven, self-propelled CF-PCR (SP-CF-PCR) microfluidic chip (Tachibana et al., 2015). The PCR microchannel was prepared using a Si substrate and was covered by a glass substrate. The surface of the microchannel was oxidized to form a hydrophilic film. PCR solution dropped onto the inlet of the microfluidic chip was transported through the microchannel by capillary flow to achieve the desired amplification. The theoretical approach of capillary flow involving more than one temperature was developed based on previous studies (Juncker et al., 2002; Delamarche et al., 1998, 2005; Zhu and Petkovic-Duran, 2010). We used this SP-CF-PCR microfluidic chip to amplify the genomic DNA of some pathogens. However, the amplification was inefficient in the head area of the flow because of the evaporation of PCR solution or adsorption of reagents. Moreover, the use of Si for preparing microfluidic chips is associated with 2 major disadvantages. One of the disadvantages is longer intermediate temperature zone between the high and low temperature zones due to the high thermal conductivity of Si. This intermediate temperature zone increases the microchannel length and decrease PCR efficiency at an unnecessary temperature. Another disadvantage is its high cost. The material used for making microfluidic chips should be cost-competitive to allow mass production for on-site use such as screening of pathogens. From this point of view, plastics are very promising for developing microfluidic chips because they have low thermal conductivity and are inexpensive. However, one of the difficulties of
using plastics for developing SP-CF-PCR chips is the achievement of hydrophilic surface on the microchannel walls because most plastics are hydrophobic. Some dry processes such as plasma (Gervais et al., 2011) or ultraviolet (UV) ozone (Novo et al., 2011, 2013) treatment have been used to activate and deliver hydrophilic groups on the surface of the microchannel walls in polydimethylsiloxane (PDMS) microfluidic chips. However, because the activated surfaces are unstable (Eddington et al., 2006), hydrophilicity disappears during the subsequent fabrication process, and because of time-associated degradation or heating. Therefore, some active hydrophilic layers should be developed on the microchannel walls. In the present study, a plastic-based microfluidic real-time quantitative PCR chip was developed as shown in Fig. 1(a). A driving microchannel, which was placed behind the PCR microchannel, was designed to exclude the head area of the flow from the PCR microchannel and to maintain capillary force for controlling the flow. The microchannel of the SP-CF-PCR chip was formed in a cyclo-olefin polymer (COP) plate by injection molding. Water-soluble polymers and surfactants were examined for producing the hydrophilic layer on the microchannel walls. Moreover, the PCR microfluidic chip was integrated with a fluorescence detection system containing a laser and photomultiplier tube (PMT) to analyze the amplified product within the chip. For verifying our concept, human β-actin, E. coli DNA, and E. coli O157 DNA were amplified using the new SP-CF-PCR chip and the fluorescence detection system.
Fig. 1. Design of the self-propelled continuous-flow PCR (SP-CF-PCR) chip. (a) Concept diagram. The PCR chip is placed on 2 heater blocks having temperatures 95 °C and 60 °C, respectively. The PCR solution dropped onto the inlet hole is loaded into the microchannels autonomously by capillary force. The driving microchannel placed behind the PCR microchannel maintains the capillary force to control the following flow of the solution. (b) Simulated temperature distribution in the COP (filled circle with solid line) and glass/Si (Tachibana et al., 2015) (filled triangle with dashed line) microfluidic chips in planes where the microchannels are formed. The microchannels were formed between a lower and an upper plate. The lower plates were attached to the heater blocks. The origin of displacement was set at the center between the high and low temperature zones on the left (negative direction) and right (positive direction), respectively. The block heaters were fixed with 1-mm gap, and the temperatures at both the ends of the chip were fixed at 99 °C and 60 °C, respectively.
H. Tachibana et al. / Biosensors and Bioelectronics 74 (2015) 725–730
2. Materials and methods 2.1. Thermal simulation analysis The cross-sectional temperature distributions in planes in which the microchannels were formed were simulated using finite element method (FEM) with ANSYS©. The microchannels were formed between upper and lower plates of 20 mm-width. The lower plate was attached on 2 aluminum heater blocks whose width and height were both 15 mm. The blocks were fixed with a 1-mm gap. In the thermal simulations, the blocks were considered as heating element and were heated in order that the temperatures at both the ends of the chip became fixed at 99 °C and 60 °C, respectively. Si, glass and COP plates whose thermal conductivities are 1.5 102, 1.0, 0.60 W/(m K), respectively, were used in the simulations. 2.2. Chip fabrication The PCR chip included two 57 25 mm2 COP plates (Zeon Corp., Tokyo, Japan). One of the plates was 1.5-mm thick and included 150-μm deep and 150-μm wide rectangular cross-sectional microchannels. The other one was a 0.7-mm thick flat cover plate. The PCR chip was fabricated as follows. First, the COP plate with the microchannels was fabricated by injection molding. The pattern of the microchannels was fabricated in Si substrate by using micromachining technologies as a mold. The microchannels were formed by photolithography and deep etching process (Tachibana et al., 2015). Note that the microchannels were convex shape. The injection molding of the COP plate using this Si mold was outsourced (Richel Corp., Toyama, Japan). The inlet and drain holes were punched at both the ends of the microchannels. The plate was then attached to the cover plate by heating and applying pressure by exposing the bonding surfaces of the plates to O2 plasma for surface cleaning and activation. After fabricating the chip structure, a hydrophilic layer was produced on the microchannel walls as follows. The microchannels were filled with a solution containing the hydrophilic medium by using a syringe. Next, the solvent inside the microchannel was evaporated. Three solutions, namely, 1% (w/w) solution of water-soluble polymer carboxymethyl cellulose (CMC) (Wako Pure Chemical Industries, Ltd. Osaka, Japan) in distilled water, 1% (w/w) solution of anion surfactant sodium cocoyl sarcosinate (SCS) (Soypon SCE, Kawaken Fine Chemicals Co. Ltd., Tokyo, Japan) in ethanol, and 1% (w/w) solution of non-ionic surfactant polyoxyethylene (20) sorbitan monolaurate (Tween 20) (Wako Pure Chemical Industries, Ltd. Osaka, Japan) in ethanol were used for coating the microchannel walls in separate chips. 2.3. PCR reagents Genomic DNA of humans, E. coli, and pathogenic E. coli O157 were amplified by performing quantitative PCRs with the SP-CFPCR chip. As a common reagents of all the targets,10 fast buffer I (SpeedSTARs HS DNA Polymerase kit; Takara Bio Inc., Tokyo, Japan), 0.2 mM dNTP mixture (Takara Bio Inc.), 0.15 U/μl SpeedSTAR HS DNA Polymerase (Takara Bio Inc.), 1 μg/μl bovine serum albumin (BSA) were used. The PCR mixture for amplifying human genomic DNA included 300 nM each of forward and reverse primers, and TaqMan fluorescence probe containing fluorescein amidite (FAM) (TaqMan βActin Detection Reagents kit; Life Technologies Inc., Carlsbad, USA). Human genomic DNA concentrations of 4, 40, 400, and 3200 pg/μl (Life Technologies Inc.) were used as the templates. The mixture for amplifying E. coli genomic DNA included 10 fast buffer I, 0.2 mM dNTP mixture, 0.15 U/μl SpeedSTAR HS DNA
727
Polymerase, and 1 μg/μl BSA. The mixture also included 300 nM each of forward (5′-CGGAAGCAACGCGTAAACTC-3′), and reverse (5′-TGAGCGTCGCAGAACATTACA-3′) primers, and 300 nM TaqMan probe (5′-FAM-CGCGTCCGATCACCTGCGTC-BHQ1-3′) targeting uidA (Silkie et al., 2008). Genomic DNA extracted from E. coli DH5α (Takara Bio Inc.) by using PureLink Genomic DNA Mini Kit (Life Technologies Inc.) was used as the template. The mixture for amplifying E. coli O157 genomic DNA included 10 fast buffer I, 0.2 mM dNTP mixture, 0.15 U/μl SpeedSTAR HS DNA Polymerase, and 1 μg/μl BSA. In addition, the mixture included 300 nM each of forward (5′-CAATTTTTCAGGGAATAACATTG-3′), reverse (5′-AAAGTTCAGATCTTGATGACATTG-3′) primers, and 300 nM TaqMan probe (5′-FAM-TCAAGAGTTGCCCATCCTGCAGCAA-BHQ1-3′) targeting eaeA (Call et al., 2001). Genomic DNA of E. coli O157 extracted from GTC 01061 was obtained from Gifu University, School of Medicine, Pathogenic Microorganism Genetic Resource Stock Center (GMGC) and used as the template. 2.4. On chip quantitative PCR The PCR chip was placed on two 15 15 80-mm3 aluminum heater blocks whose temperatures were controlled individually. The aluminum heater blocks were set in parallel with 1-mm gap and included cartridge heaters and thermocouples (Kyushu-Nissho Co. Ltd, Fukuoka, Japan). One of the heaters was controlled at denaturation temperature while the other was controlled at annealing and extension temperatures by using proportional-integral-derivative controller (Shimaden Co. Ltd., Tokyo, Japan). 50 μl of PCR solution was dropped on the inlet hole. Displacement of capillary flow was determined from the time of PCR solution loading by visually monitoring the position of the fluid front until all the microchannels were filled. Once the microchannels were filled, optical component was scanned over the PCR microchannel, and fluorescence of the solution was detected. The detail of the fluorescence detection system and the analysis of the amplification curves with fluorescence are described in Supporting information on page 2–3 and Fig. S1.
3. Results and discussion 3.1. Design of the PCR microfluidic chip One drawback of the SP-CF-PCR is the decrease in PCR efficiency at the fluid front (Tachibana et al., 2015). The fluid front of the PCR solution is continuously in contact with the fresh surface of the microchannel. Therefore, adsorption of PCR reagents such as polymerase or amplified DNA occurs ineluctably (Nakayama et al., 2006). Moreover, evaporation of the PCR solution occurs at the fluid front because it is constantly exposed at high temperature. To overcome these issues, we newly designed a driving microchannel that was placed behind the PCR microchannel (Fig.1). The velocity of capillary flow in CF-PCR is described as (Tachibana et al., 2015)
v (x) =
R h2 Pc (x) x
8 ∫0 η (x‵) dx‵
(1)
According to Eq. (1), the velocity of the capillary flow decreases as the solution flows through the microchannel but never becomes 0. Therefore, the driving microchannel controls the capillary forces required to maintain the flow of the PCR solution even after the PCR microchannel is filled. Because of the driving microchannel, the head area of the flow, with the high initial flow velocity and low PCR efficiency, can be removed to the driving microchannel.
728
H. Tachibana et al. / Biosensors and Bioelectronics 74 (2015) 725–730
Thus, the driving microchannel allows in achieving a stable flow velocity and high PCR efficiency. The SP-CF-PCR chip having meander shaped PCR microchannel was set on 2 aluminum block heaters in order to form 2 temperature zones in the same plane. This results in the ineluctable formation of an intermediate temperature zone between the 2 temperatures zones. A longer intermediate zone not only increases the total microchannel length but also decreases PCR efficiency at an unnecessary temperature. The intermediate temperature zone depends largely on the thermal conductivity of the PCR chips. In this study, intermediate temperature zones were investigated by comparing between a glass/Si chip having high thermal conductivity (Tachibana et al., 2015) and a COP chip having low thermal conductivity. For evaluating the intermediate temperature zones, cross-sectional temperature distributions in planes in which the microchannels were formed were simulated using FEM analysis. Fig. 1(b) shows the results of the comparison between the glass/Si PCR chip and the COP PCR chip. The origin of displacement was set at the center between the high and low temperature zones, which were placed on the left (negative direction) and right (positive direction), respectively. The microchannels of the glass/Si PCR chip were formed between a 0.5-mm thick Si lower plate and a 0.5-mm thick glass upper plate. The microchannels of the COP PCR chip were formed between a 0.7mm thick lower COP plate and a 1.5-mm thick upper plate. High and low temperature zones were defined as 95–99 °C for denaturation and 60–65 °C for annealing and extension, respectively. Hence, the temperature of the intermediate zone was between 65 °C and 95 °C. The intermediate temperature zone in the glass/Si PCR chips was 7.8 mm long, which was twice its length in the COP PCR chips (3.6 mm). Although the formation of the microchannels and hydrophilic surfaces on the microchannel walls by thermal oxidation is easier in the glass/Si PCR chips, heat insulation in these chips is more difficult than that in COP chips because the thermal conductivity of Si is two orders of magnitude higher than that of COP. Materials used to make the PCR chip should have low thermal conductivity to allow the temperature gradient to concentrate in the gap between 2 heaters. According to this simulation, the microchannel length in each turn, which corresponded to half the length of each PCR cycle, was set as 7 mm to achieve sufficient length. The designed microchannels were 820 mm (53 cycles) of PCR microchannel and 440 mm of driving microchannel.
3.2. Capillary flow in the microchannels Hydrophilic layers can be developed on hydrophobic surfaces by using water-soluble polymers (Kargl et al., 2012) and surfactants (Aegerter and Menning, 2004) because they contain both hydrophobic and hydrophilic groups. In this study, we examined suitable surface treatments to develop a hydrophilic layer on the surface of COP plates, such as treatment with CMC, a water-soluble polymer; SCS, an anion surfactant; and Tween 20, a non-ionic surfactant. Contact angle of the hydrophilic layer formed on the COP flat plate was 15° with CMC and below 10° with SCS and Tween 20. The hydrophilicity of all the microchannels was maintained for more than 10 days. The positions of the fluid fronts of the capillary flow in PCR chips whose microchannel walls were covered with CMC, SCS, and Tween 20 are shown in Fig. 2. In the PCR chip whose microchannel walls were covered with CMC, the PCR solution was loaded only several millimeters, indicating that the CMC coating on the microchannel walls was non-homogeneous. This could be because CMC was dissolved in distilled water; therefore, CMC solution was localized in the microchannel during evaporation because the surface of the COP plate repelled water. In contrast, PCR solutions were successfully loaded into PCR chips whose microchannel walls were covered with either SCS or Tween 20 because homogeneous coatings could be obtained. Surfactants can be applied easily on the microchannel walls because they contain both hydrophilic and hydrophobic groups (Aegerter and Menning, 2004). Moreover, the surfactants did not agglomerate during evaporation because they were liquid at room temperature. Tween 20 was found to be suitable for producing the hydrophilic layer in this study because the PCR solution reached the drain in less than 20 min. On the other hand, in the SCS-coated chip, the PCR solution could not reach to the drain. These results indicated that surface hydrophilicity of Tween 20 was higher than that of SCS. This could be because Tween 20 has more hydrophilic groups than SCS and because it is generated by polar ether linkage compared with SCS, which only contains a carboxyl group at the end of a long-chain hydrocarbon. In the Tween 20-coated chip, the PCR solution initially followed the PCR microchannel and filled it within 7 min (Fig. 2). The PCR solution then filled the driving microchannel. The capillary force present at the fluid front in the driving microchannel helped the flow of the PCR solution through the PCR microchannel. The total time for filling the microchannels was 18 min in this study. In this
Fig. 2. Characteristics of capillary flow. (a) Position of the fluid front of the PCR solution with respect to time in the PCR chip whose microchannel walls are treated with polyoxyethylene (20) sorbitan monolaurate (Tween 20), a non-ionic surfactant (filled circle with solid line); sodium cocoyl sarcosinate (SCS), an anion surfactant (filled triangle with dashed line); and carboxymethyl cellulose (CMC), a water-soluble polymer (filled rectangles with dotted line). The dashed line at 820 mm indicates the boundary between the PCR and the driving microchannels. (b) Photographs of capillary flow in the PCR chip whose microchannel walls are treated with Tween 20 by using the new coccine dye. The arrows indicate the position of the fluid front.
H. Tachibana et al. / Biosensors and Bioelectronics 74 (2015) 725–730
experiment, the volume per cycle in the PCR microchannel was 0.35 μl. On the other hand, the volume of the meander-shaped driving microchannel was 9.8 μl, which corresponded to the volume of 28 PCR cycles. Therefore, PCR solution of 28 PCR cycle in the fluid front was eventually moved to the driving microchannel. As a result, effective PCR was initiated when the fluid front was at the positions of the 28th cycle. The average flow velocity during effective PCR was 0.87 mm/s, which corresponded to 16 s per cycle. The time for each cycle was sufficient to achieve an efficient PCR. The possibility of controlling the capillary flow by using the driving microchannel was successfully observed in the present study. Other optimal designs such as capillary pump structure (Zimmermann et al., 2007) will enable us to obtain a faster and steady flow velocity. 3.3. Quantitative PCR detection Quantitative real-time PCR was completed once the driving microchannel was filled. Therefore, the PCR time was 18 min as shown in Fig. 2. The amplification curves of the PCR targeting the human β-actin are shown in Fig. 3(a). Clear amplifications were obtained for the initial human genomic DNA concentrations of 4, 40, 400, and 3200 pg/μl. The amplification curves were shifted to the low cycle as the initial concentration of the human genomic DNA increased. Specific amplification of a 295-bp fragment was achieved as a representative result of gel electrophoresis (Fig. S2). The standard curve (threshold cycles (Ct values) versus initial concentrations of human genomic DNA) of the human genomic DNA is shown in Fig. 3(b). The definition of Ct value is described in
729
the Supporting information on page 1–2. As expected, the Ct values had a linear characteristic with respect to the logarithmic values of the initial DNA concentrations. The obtained Ct value including the reproducibility for initial DNA concentration of 4 pg/ μl was 42.2 7 3.2, which indicates that the corresponding coefficient of variance (CV) was 7.5%. The lowest initial concentration of human genomic DNA was 4 pg/μl, which corresponded to approximately 0.6 genome/μl (Doležel et al., 2003). These results indicated that our PCR chip and detection system were applicable for quantitative DNA analysis with sufficient high sensitivity and reproducibility. Next, we performed the quantitative PCR targeting uidA of E. coli with initial genomic DNA concentrations of 0.0019, 0.019, 0.19, and 1.9 pg/μl. The amplification curves and a result gel electrophoresis are shown in Fig. S3. Standard curve of E. coli genomic DNA were obtained from the Ct values obtained for each initial DNA concentration (Fig. 3(c)). The obtained Ct value including the reproducibility for initial DNA concentration of 0.0019 pg/μl was 45.6 72.9, which indicates that the corresponding coefficient of variance (CV) was 6.3%. The lowest initial concentration of E. coli genomic DNA of 0.0019 pg/μl corresponded to approximately 0.4 genome/μl. Finally, we performed quantitative PCR by using the genomic DNA of a pathogenic microorganism. E. coli O157 is one of the most notorious foodborne pathogens that colonizes the intestinal tract of cattle and is present in beef and other related products. The infectivity of E. coli O157 is extremely high; therefore, infection can be caused by only a few hundred cells (Karmali, 2004). Therefore, rapid, highly accurate, and highly sensitive on-site detection is required to prevent an outbreak of food poisoning. Quantitative
Fig. 3. On-chip quantitative real-time PCR results. (a) Amplification curves with fluorescence targeting human β-actin. (b) Standard curve of PCR targeting human β-actin. (c) Standard curve of PCR targeting uidA of E. coli. (d) Standard curve of quantitative real-time PCR targeting eaeA of E. coli O157.
730
H. Tachibana et al. / Biosensors and Bioelectronics 74 (2015) 725–730
PCR targeting eaeA of E. coli O157 was performed with initial genomic DNA concentrations of0.031, 0.31, 3.1, and 31 pg/μl. The amplification curves and a representative result of gel electrophoresis are shown in Fig. S4. Standard curve of E. coli O157 genomic DNA were obtained from the Ct values calculated from the corresponding amplification curves. The obtained Ct value including the reproducibility for initial DNA concentration of 0.031 pg/μl was 41.772.8, which indicates that the corresponding coefficient of variance (CV) was 6.7%.This standard curve indicated that this technique could be used to quantitatively analyze E. coli O157 genomic DNA for practical purpose such as food testing. Moreover, the lowest initial concentration of E. coli O157 genomic DNA was 0.031 pg/μl in this experiment, which corresponded to approximately 6.9 genome/μl, indicating that this technique could detect DNA from less than 500 cells in a practical sample.
4. Conclusions In this study, we have successfully developed an autonomous microfluidic PCR platform for on-chip quantitative detection of pathogen genes. The newly designed driving microchannel enabled us to maintain the flow of the PCR solution. The excellent thermal controllability was obtained in the COP PCR chip. Moreover, we have achieved optimal surface hydrophilicity of the microchannel walls by detailed investigation of non-ionic surfactants. As the results, the PCR solution was successfully transported through the whole microchannels autonomously and quantitative real-time PCR was successfully performed in less than 18 min. Our results indicated that the SP-CF-PCR chips could be used for rapid, low-cost and easy-to-use gene testing in the medical and environmental fields.
Appendix A. Supplementary material Supplementary data associated with this article can be found in the online version at http://dx.doi.org/10.1016/j.bios.2015.07.009.
References Aegerter, M.A., Menning, M., 2004. Sol–gel technologies for glass producers and users. Kluwer Academic Publishers, Norwell. Best, E.L., Powell, E.J., Swift, C., Grant, K.A., Frost, J.A., 2003. FEMS Microbiol. Lett. 229, 237–241. Call, D.R., Brockman, F.J., Chandler, D.P., 2001. Int. J. Food Microbiol. 67, 71–80. Cao, Q., Kim, M., Klapperich, C.M., 2011. Biotechnol. J. 6, 177–184. Christopoulos, T.K., 1999. Anal. Chem. 71, 425R–438R. Crews, N., Wittwer, C., Gale, B., 2008a. Biomed. Microdev. 10, 187–195. Crews, N., Wittwer, C., Palais, R., Gale, B., 2008b. Lab Chip 8, 919–924. Delamarche, E., Bernard, A., Schmid, H., Bietsch, A., Michel, B., Biebuyck, H., 1998. J. Am. Chem. Soc. 120, 500–508. Delamarche, E., Juncker, D., Schmid, H., 2005. Adv. Mater. 17, 2911–2933. Doležel, J., Bartos, J., Voglmayr, H., Greilhuber, J., 2003. Cytometry A 51, 127–128. Eddington, D.T., Puccinelli, J.P., Beebe, D.J., 2006. Sens. Actuators B 114, 170–172. Frydendahl, K., Imberechts, H., Lehmann, S., 2001. Mol. Cell. Probes 15, 151–160. Gervais, L., Hitzbleck, M., Delamarche, E., 2011. Biosens. Bioelectron. 27, 64–70. Juncker, D., Schmid, H., Drechsler, U., Wolf, H., Wolf, M., Michel, B., Rooij, N., Delamarche, E., 2002. Anal. Chem. 74, 6139–6144. Kargl, R., Mohan, T., Bračič, M., Kulterer, M., Doliška, A., Stana-Kleinschek, K., Ribitsch, V., 2012. Langmuir 28, 11440–11447. Karmali, M.A., 2004. Mol. Biotechnol. 26, 117–122. Kopp, M.U., Mello, A.J., Manz, A., 1998. Science 280, 1046–1048. Lee, D.-Y., Shannon, K., Beaudette, L.A., 2006. J. Microbiol. Method 65, 453–467. Li, S., Fozdar, D.Y., Ali, M.F., Li, H., Shao, D., Vykoukal, D.M., Vykoukal, J., Floriano, P. N., Olsen, M., McDevitt, J.T., Gascoyne, P.R.C., Chen, S., 2006. J. Microelectromech. Syst. 15, 223–236. Mullis, K.B., Falona, F.A., 1987. Methods Enzymol. 155, 335–350. Nakayama, T., Kurosawa, K., Furui, S., Kerman, K., Kobayashi, M., Rao, S.R., Yonezawa, Y., Nakano, K., Hino, A., Takamura, Y., Tamiya, E., 2006. Anal. Bioanal. Chem. 386, 1327–1333. Nakayama, T., Hiep, H.M., Furui, S., Yonezawa, Y., Saito, M., Takamura, Y., Tamiya, E., 2010. Anal. Bioanal. Chem. 396, 457–464. Novo, P., Prazeres, D.M.F., Chu, V., Conde, J.P., 2011. Lab Chip 11, 4063–4071. Novo, P., Volpetti, F., Chu, V., Conde, J.P., 2013. Lab Chip 13, 641–645. Obeid, P.J., Christopoulos, T.K., 2003. Anal. Chim. Acta 494, 1–9. Silkie, S.S., Tolcher, M.P., Nelson, K.L., 2008. J. Microbiol. Method 72, 275–281. Ståhl, M., Kokotovic, B., Hjulsager, C.K., Breum, S.Ø., Angen, Ø., 2011. Vet. Microbiol. 151, 307–314. Tachibana, H., Saito, M., Tsuji, K., Yamanaka, K., Hoa, L.Q., Tamiya, E., 2015. Sens. Actuators B 206, 303–310. Yamanaka, K., Saito, M., Kondoh, K., Hossain, M.M., Koketsu, R., Sasaki, T., Nagatani, N., Ikuta, K., Tamiya, E., 2011. Analyst 136, 2064–2068. Zhu, Y., Petkovic-Duran, K., 2010. Microfluid Nanofluid 8, 275–282. Zimmermann, M., Schmid, H., Hunziker, P., Delamarche, E., 2007. Lab Chip 7, 119–125.