Two-dimensional micro-bubble actuator array to ... - Semantic Scholar
Biosensors and Bioelectronics 21 (2006) 1443–1450
Two-dimensional micro-bubble actuator array to enhance the efficiency of molecular beacon based DNA micro-biosensors Peigang Deng a , Yi-Kuen Lee a,∗ , Ping Cheng b a
Department of Mechanical Engineering, Hong Kong University of Science and Technology, Clear Water Bay, Kowloon, Hong Kong b School of Mechanical and Power Engineering, Shanghai Jiaotong University, Shanghai 200030, PR China Received 23 March 2005; received in revised form 14 June 2005; accepted 17 June 2005 Available online 11 August 2005
Abstract Two-dimensional micro-bubble actuator arrays were developed and studied in detail to enhance the hybridization kinetics of a DNA micro-biosensor. The hybridization between a molecular beacon, a kind of oligonucleotide probe, and its complement was investigated in a millimeter-sized PDMS based reaction chamber, where various 2D micro-heater arrays were distributed on the bottom for micro-bubble generation. The hybridization assay without the micro-bubble actuation revealed that the fluorescence increased fast at the beginning and slowed down after that. However, a uniform fluorescence increase was observed when periodic micro-bubble agitation was introduced in the static hybridization solution. A comparison of hybridization assays with and without micro-bubble agitation revealed that the hybridization time could be effectively shortened by 33% with 10 cycles of micro-bubble agitation from a 2 × 1 bubble actuator array, and by 43% with 10 cycles of micro-bubble agitation from a 2 × 2 bubble actuator array. © 2005 Elsevier B.V. All rights reserved. Keywords: Micro-bubble actuator; Molecular beacon; Flow perturbation; DNA hybridization; Micro-biosensor
1. Introduction DNA hybridization is the underlying principle of DNA biosensors. In DNA hybridization, the oligonucleotide probe (DNA probe) specifically recognizes, and binds to, a nucleic acid target (target DNA), which forms a double-stranded hybrid with its nucleic acid complement with high efficiency and specificity. The main types of hybridization used today are solution hybridization, filter hybridization, the polymerase chain reaction, in situ hybridization and hybridization on DNA chips (Tenover, 1993; Anderson, 1999; Freeman et al., 2000; Sosnowski et al., 2002). Hybridization in solution is believed to be a two-step process involving nucleation and zipping up, in which nucleation is the rate-limiting step, and a second-order reaction equation can be used to describe the process (Anderson, 1999).
∗
Corresponding author. Tel.: +852 2358 8663; fax: +852 2358 1543. E-mail address: [email protected] (Y.-K. Lee).
0956-5663/$ – see front matter © 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.bios.2005.06.007
Generally, the hybridization of DNA probes is diffusionlimited, i.e. the signal intensity is determined by the number of target molecules reaching the probe. However, diffusion is such an intrinsically slow process for large molecules that in micro-arrays even after overnight hybridization the system does not reach equilibrium. Furthermore, the slow diffusive transport could result in poor binding efficiency as less than 1% in micro-arrays (Pappaert et al., 2003). The diffusion limitation is a particular problem for a micro-scale biological reaction, where only small amounts of analyte are available and the required reaction volume limits the analyte concentration. The obvious solution to overcome the diffusion limitation of hybridization experiments is to agitate the sample solution. Recently, efforts have been devoted to increase the mixing and diffusion for DNA hybridization in a micro-scale confined space. Nanogen (San Diego, CA, USA) developed microchip-based hybridization arrays (NanoChipTM ) that utilized electric fields as an independent parameter to control DNA transport and enhance hybridization (Edman et al.,
1444
P. Deng et al. / Biosensors and Bioelectronics 21 (2006) 1443–1450
1997). Fixe et al. (2003) employed an electric field pulse to facilitate the DNA probe immobilization and to enhance the DNA hybridization process. Yuen et al. (2003) realized closed-loop fluid circulation and mixing by the rotation of a magnetic stirring bar. Other methods, which were used to enhance the DNA hybridization process, included sheardriven flows or surface acoustic waves (Pappaert et al., 2003). In our previous studies, a micro-thermal bubble actuator capable of generating effective flow perturbations in the bulk fluid was developed (Deng et al., 2003). More recently, transient micro-thermal bubble generation and its subsequent dynamic behavior in a 60-base single-stranded DNA (ssDNA) solution was investigated with particular emphases on the effects of DNA concentration and the total viscosity of the solution on bubble dynamics (Deng et al., 2004). It was found that the DNA macromolecules had a strong retardation effect on bubble dynamics at high concentrations. The bubble nucleation temperature at different DNA concentrations was measured by using the micro-Pt heater as a self-sensing resistive temperature sensor (Deng et al., 2005). The on-set bubble nucleation temperature was around 480 K for low DNA concentrations, and an increase in the DNA concentration could lead to a higher heating rate, a faster bubble nucleation process and a higher on-set nucleation temperature. In this paper, we will use a 2D micro-bubble actuator array to generate effective perturbations in the hybridization solution for the purpose of enhancing the DNA molecule diffusion and consequently the DNA hybridization rate. A type of oligonucleotide probe—molecular beacon (Tyagi and Kramer, 1996), was used as a DNA biosensor in the present study, and the hybridization assay of the molecular beacon with its complement was carried out and compared for the cases with and without bubble actuation.
2. Experimental 2.1. Design of the molecular beacon for DNA hybridization assay Since first reported by Tyagi and Kramer (1996), molecular beacons have been increasingly used in many applications (Poddar, 1999; Liu and Tan, 1999; Mustafa et al., 2004). As illustrated in Fig. 1, molecular beacons are dual-labeled
Fig. 1. Schematic representation of molecular beacon: (a) before hybridization, the molecule beacon remains non-fluorescent, because the fluorophore was quenched by the quencher; (b) after hybridization with the target the molecular beacon becomes fluorescent.
oligonucleotide probes that have a fluorescent dye molecule at one end and a fluorescence quencher at the opposite end. The probe is designed with a target-specific hybridization domain positioned centrally between short sequences that are self-complementary. In the absence of the target, the selfcomplementary domains anneal to form a stem–loop hairpin structure, that results in quenching of the reporter due to the fluorescence resonance energy transfer as shown in Fig. 1(a). In the presence of the target, the central loop domain will hybridize with the complementary target DNA or RNA, forcing the molecule to unfold; reporter and quencher are now physically separated and the fluorescence of the reporter dye will be restored upon excitation, as shown in Fig. 1(b). In this study, the sequence of the loop structure of the molecular beacon was designed according to a portion of the sequence of 16S subunit ribosomal RNA in E. coli (MC41000) and was 17 bases long. The stem part of the molecular beacon was six bases in length with a rich G/C percentage of 66%. The 5 end was labeled with Fluorescein (FAM) and the 3 end was labeled with Dabcyl (4-(4 -dimethylaminophenylazo) benzoic acid) quencher. The molecular beacon was synthesized by IDT (Integrated DNA Technologies Inc., Coralville, IA, USA), and the complete molecular beacon sequence was 5 -/56-FAM/ CAGTCGTATTA ACTTTACTCCCTCGACTG/3Dabcyl/3 , where the underlined nucleotides are the stem sequences. In the present design of the molecular beacon, a socalled “shared-stem” technology was used to make the beacon-target duplex more stable (Tsourkas et al., 2003). This design differs from a conventional molecular beacon in that one arm of the stem (the underlined and italic nucleotides near to the 5 end) participates in either hairpin formation or target hybridization. The beacon complement was also provided by IDT and the sequence was 5 TTTAGGGAGTAAAGTTAATACGACTG-3 . The stability of the molecular beacon-target duplex was estimated by using an IDT online program ‘OligoAnalyzer 3.0’ (IDT). The melting temperature is 51.8 ◦ C at the hybridization conditions of 0.25 M oligo, 50 mM monovalent salt. As to the stability of the self-complementary stem of the molecular beacon, 6 basepair-long G/C rich stems will melt between 60 and 65 ◦ C (Marras et al., 2003). 2.2. Micro-bubble actuator array Two kinds of 2D micro-heater arrays (2 × 1 and 2 × 2) were fabricated as shown in Fig. 2. The area of the slip part of each micro-heater of 2 × 1 array was 10 m × 3 m, and that of the 2 × 2 array was 3 m × 1 m. In our previous study, the effective perturbation region induced by microbubble generation in the bulk liquid was evaluated by putting micro-particles in the working fluid as the fluid tracer. This perturbation region was assumed to be a semi-sphere with its center at the center of the bubble. By examining the recorded dynamic response of the particles, the radius of the perturbation was estimated as six times that of the maximum bubble
P. Deng et al. / Biosensors and Bioelectronics 21 (2006) 1443–1450
1445
Fig. 3. PDMS based reaction chamber on a silicon chip of a micro-bubble actuator array.
Fig. 2. Two-dimensional micro-bubble actuator array: (a) 2 × 1 array; (b) 2 × 2 array; (c) close-up view of the micro-bubble.
radius (Deng et al., 2003). Therefore, the distance between each heater of the micro-bubble actuator array visualized in Fig. 2 was designed such that the effective flow perturbation area of each individual heater could not overlap much with its neighbor. The fabrication of the micro-bubble actuator array followed the conventional MEMS technology. Platinum was chosen as the material for both the heater and the electric pad. A n-type, double-sided polished, and 1 0 0 oriented silicon wafer (4 in. in diameter and 400 m in thickness) was used as the substrate. The wafer was coated with 2 m silicon dioxide on both sides of the wafer by wet thermal oxidation. A 0.02 m thick Ti film and a 0.18 m thick Pt film were sputtered on the wafer and patterned using the liftoff method. The fabricated micro-bubble actuator array was tested in the hybridization buffer. A 1.66-ms wide voltage pulse was imposed on each individual heater of the heater arrays (2 × 1 and 2 × 2). The height of the voltage pulse was increased from zero until micro-bubble was observed at a certain voltage, which is called “onset bubble nucleation”. The control circuit for the voltage pulse generation and the experimental set up for micro-bubble visualization were introduced in detail elsewhere (Deng et al., 2003, 2004). For the onset bubble nucleation, a stable single bubble could be generated on each individual heater of each heater array (2 × 1 and 2 × 2) simultaneously, as shown in Fig. 2. The micro-bubble grew and collapsed just on the heater surface (no detachment) due to the highly localized near homogeneous boiling mechanism for bubble nucleation (Deng et al., 2003). The bubble growth period was very similar to the heating pulse width, i.e. 1.66 ms in the present study, and the maximum bubble diameter was comparable to the size of the slim part of the micro-heater. The bubble began to collapse after the heating process stopped at the end of the 1.66 ms voltage pulse, at which time the bubble reached its maximum size. The bubble collapse process was much longer compared to the bubble growth process, which was mainly dominated by the vapor condensation in the subcooled liquid. The bub-
ble lifetime on heaters of a 2 × 1 array (10 m × 3 m) was around 320 ms, and it was around 45 ms on heaters of a 2 × 2 array (3 m × 1 m). The bubble nucleation temperature on the micro-heaters was evaluated at around 485 K from our previous study (Deng et al., 2003, 2005), and the superheated region (over 373 K) in the liquid was the highly localized area adjacent to the micro-heater surface (12 m × 3 m × 4 m for the 10 m × 3 m heater, and 3 m × 1 m × 1 m for the 3 m × 1 m heater). Note that for the present transient micro-bubble generation, the superheat region of the liquid not only was restricted to a very local area but also lasted for a short period. A series of fundamental research studies on micro-bubble generation under pulse heating in water, DNA buffers and DNA solutions have been carried out in our previous studies (Deng et al., 2003, 2004, 2005) in order to design an optimized array of micro-bubble actuators for DNA hybridization. 2.3. Fabrication of the PDMS reaction chamber Fig. 3 shows the picture of a micro-bubble actuator chip, where the reaction chamber was made of PDMS (Sylgard184, Dow Corning, USA). To precisely control the amount of the molecular beacons and the beacon complements in the hybridization solution, the dimension of the reaction chamber was designed at millimeter level, and thus the conventional pipette operation was possible. The PDMS reaction chamber was of a cylindrical shape with a diameter of 8.5 mm and a height of 1.8 mm. The total volume of the reaction chamber was about 100 L. To avoid direct exposure of the DNA hybridization buffer with the micro-Pt heater, a 60 nm thick parylene N (diX-NTM , Daisan Kasei Co., Japan) layer was coated on the packaged chip, which served as an electrical and chemical insulating layer. The procedure for the Parylene N coating was described in our previous paper (Deng et al., 2004). 2.4. DNA sample preparation The lyophilized molecular beacon and beacon complement were first centrifuged after keeping them at room
1446
P. Deng et al. / Biosensors and Bioelectronics 21 (2006) 1443–1450
temperature for 1 h. Then they were dissolved completely in distilled water (UltraPure and DNase/RNase free, Invitrogen, USA) so as to obtain three stock concentrations: 100, 10 and 1 M. Before hybridization assay, the concentrated oligonucleotides were further diluted with the distilled water to the working concentration.
3. Results and discussions 3.1. Molecular beacon hybridization without micro-bubble actuation The hybridization assay in the PDMS reaction chamber was first operated without the micro-bubble actuation. To start the hybridization assay, 75.2 L of hybridization buffer was loaded in the PDMS reaction chamber, and then 1.6 L of molecular beacon and 3.2 L of beacon complement were added to the hybridization buffer. Thus, the total volume of the hybridization solution was 80 L, where the concentrations of the molecular beacon and the beacon complement were both 100 nM, and the hybridization solution contained 50 mM KCl, 5 mM MgCl2 , 10 mM Tris–HCl, pH value = 7.5. Note that there was no mechanical mixing of the molecular beacon and the beacon complement involved. A 0.9 mm thick PDMS cover was put on the reaction chamber to avoid evaporization of the liquid. The liquid loss due to the permeability of the PDMS was tested by filling the reaction chamber with the hybridization buffer, covering it with the PDMS cover and then putting the whole system onto an electric balance. It was found that the loss of the weight was less than 0.1% after 24 h, which means the liquid loss due to the permeability of the PDMS is negligible. After loading the molecular beacon and the beacon complement in the hybridization buffer, the whole system was then put on the stage of an Olympus BX41 microscope, and the hybridization solution was excited with a 475 nm laser and the emission filter was set at 510 nm. The image of the fluorescence emission at different times was recorded by a CCD camera (7.2 Color Mosaic, Diagnostic Instruments Inc., USA), where the exposure time was fixed at 500 ms. In all the fluorescence intensity measurements the microscope was always focused on the micro-heater array at the bottom of the reaction chamber to minimize any potential experimental error. After the 12-bit, 800 × 600 pixels, grayscale image was obtained, as shown in Fig. 4, it was analyzed by a MatlabTM program. To calculate the effective fluorescence intensity from the image, firstly a rectangle two or three times larger than the area of the micro-heater array was defined, as indicated by the red rectangle box in Fig. 4. Then, the intensity value at each pixel inside this rectangle was accumulated and then averaged by the total pixels to get an averaged intensity value, which was regarded as the intensity of the fluorescence emission of the hybridization solution at the time when the image was recorded. Errors due to the focusing and setting the florescence calculation area could result in an average rel-
Fig. 4. Typical image of fluorescence emission of molecular beacon hybridization in a PDMS reaction chamber.
ative error of ±5% in the measured value of the fluorescence intensity. The hybridization assay was repeated five times, and Fig. 5 shows the variation of the fluorescence intensity with respect to time as the hybridization assay occurred in the PDMS reaction chamber without the micro-bubble actuation. Obviously, after loading the molecular beacon and the beacon complement in the hybridization buffer, as time elapsed, more and more molecular beacons could meet the beacon complements and hybridize with them. Thus the intensity of the fluorescence emission increased gradually, as shown in Fig. 5. The gradually increasing fluorescence intensity finally reached a saturated value (maximum value), which means most of the molecular beacons have successfully hybridized with their complements and a dynamic equilibrium has reached. It is important to examine the change of the slope of the fluorescence intensity increase with respect to time in Fig. 5, which shows that the slope is much sharper in the first 2 h, while it flattened after that. This actually implies that the hybridization of the molecular beacons and the beacon complements was a fast process in the first 2 h, while slowed down after that. We know that as time elapsed, the free molecular beacons and beacon complements became less and less because of the formation of the hybridized duplexes, which also increased the diffusion obstacle. As a result, the diffusion process slowed down gradually and thus reduced the
Fig. 5. Fluorescence intensity as a function of time for molecular beacon hybridization without micro-bubble actuation. The relative standard deviation for the experimental data is 12.5%.
P. Deng et al. / Biosensors and Bioelectronics 21 (2006) 1443–1450
accessibility of the molecular beacons and the beacon complements and subsequently the hybridization rate. In this paper, the hybridization time is defined as the time required for the fluorescence intensity of the hybridization solution to reach the saturated value (maximum value). From Fig. 5, it required about 4–6 h for the fluorescence intensity to reach a saturated value. Thus, the hybridization time was approximately 4–6 h in this case, which is, of the same order of magnitude, as that deduced from the plot of the re-association curve for double-stranded nucleic acids with various nucleotide pairs (Britten and Kohne, 1968). Note that detailed simulation to predict the hybridization time is possible by including the coupled mass, momentum, thermal transport and the DNA hybridization kinetics (Erickson et al., 2003). However, this is not the major objective of this paper. 3.2. Assessment of the molecular beacon damage due to bubble generation Although thermal bubble-jet printing technology has been successfully employed to eject DNA segments onto a glass surface for DNA micro-array (Okamoto et al., 2000), caution must be applied when generating thermal bubbles in the DNA solutions. From previous studies (Deng et al., 2003, 2005), we know that accompanying with the vapor bubble generation, it would induce an effective flow perturbation and a transient high temperature region in a very localized area in the liquid. Both of these might lead to the “opening” of the chemical bond of the base-paired “stem part” of the molecular beacon even when there is no beacon complement in the solution. This type of malfunction of the molecular beacon, if it happens, would result in fluorescence increase in the hybridization solution and should be deducted from the fluorescence obtained in the bubble enhanced hybridization assay. Therefore, before introducing the micro-bubble actuation in the hybridization solution, it is necessary to evaluate whether this process would induce any damage to the molecular beacons. Two assays were designed to test this damage issue. In the first assay, 75.2 L of hybridization buffer was loaded in one reaction chamber, and then 1.6 L of molecular beacon and 3.2 L of double-distilled H2 O were added into the hybridization buffer. The concentration of the molecular beacon was 100 nM, and the hybridization buffer contained 50 mM KCl, 5 mM MgCl2 , and 10 mM Tris–HCl, pH value = 7.5. In the second assay, the same solution was prepared in another reaction chamber and then micro-bubbles were generated simultaneously from the four heaters of a 2 × 2 heater array. A total of 10 heating pulses were employed and a 10 s waiting pulse was inserted between two consecutive heating pulses. In the experiment, the hybridization buffer was loaded to the two reaction chambers at the same time, and then bubble actuation was introduced in one of them. After that, the fluorescence emission from the hybridization buffer in two reaction chambers were recorded and compared.
1447
Fig. 6. Fluorescence emission from the solution containing molecular beacon with and without micro-bubble agitation. The data were from eight sets of experiments, and the relative standard deviation is 5%.
The two assays with and without bubble actuation were repeated eight times and the results are shown in Fig. 6. Basically, there is no distinct difference in the fluorescence emission for each case. In six of the total of eight cases the average fluorescence emission from solutions without bubble agitation was even a little bit larger than that with bubble agitation. For the other two cases, however, the situation was reversed and the fluorescence emission was a little bit larger from solutions with bubble agitation. Note that the differences for all the cases are still within the range of the error bar as shown in Fig. 6. As the temperature rises, the hairpin stem of the molecular beacon will unravel into a fluorescent randomly coiled oligonucleotide (Bonnet et al., 1999). As discussed in Section 2, the melting temperature of the self-complementary stem of the molecular beacon, 6 basepair-long and G/C rich in this study, is between 60 and 65 ◦ C (Marras et al., 2003). When vapor bubbles were generated in the solution containing molecular beacons, the localized shear force and high temperature might damage the molecular beacon by breaking the chemical bonds. As a result, the stem part of the molecular beacon would be cleaved and the molecular beacon would become a fluorescent randomly coiled oligonucleotide. However, if the molecular beacons were not broken into small pieces of oligonucleotides, the cleaved stem part of the molecular beacon could re-associate again as long as the flow perturbation stopped and the high temperature went back to the room temperature. This might be the reason that there was no distinct fluorescence increase from the solution where molecular beacons were subjected to micro-bubble agitations. It implies that the micro-bubble actuation in the hybridization solution would not lead to detectable malfunction of the molecular beacon. 3.3. Micro-bubble actuation for DNA hybridization enhancement The molecular beacon hybridization with micro-bubble actuation was conducted in the PDMS reaction chamber.
1448
P. Deng et al. / Biosensors and Bioelectronics 21 (2006) 1443–1450
Fig. 7. Flow perturbation induced by micro-bubble actuation.
The hybridization assay followed exactly the same procedure as described in Section 3.1. After the preparation of the hybridization solution, micro-bubbles were generated in the static hybridization solution to produce flow perturbations for the purpose of increasing the accessibility of the molecular beacons to the beacon complements, and thus enhancing the DNA hybridization process. As discussed in Section 2.2, two kinds of 2D micro-bubble actuator arrays, 2 × 1 and 2 × 2, were employed for bubble agitation in the hybridization solution. A 1.66-ms wide voltage pulse was imposed on each individual heater of each heater array for onset bubble nucleation. A 10 s waiting pulse was inserted in between two consecutive heating pulses. Thus, as the micro-bubble actuator worked, micro-single bubbles were generated on each individual heater of a heater array simultaneously every 10 s. Fig. 7 shows the visualization of the flow perturbation as a micro-bubble was generated on a micro-heater of 10 m × 3 m, where 0.96 m particles were placed in the liquid as a tracer. Because of the asymmetry of the bubble expansion and collapse process, a net displacement of the particles was achieved, indicating an effective flow perturbation by micro-bubble agitation. A total of 10 cycles of the micro-bubble actuation was introduced in the hybridization solution. For each cycle, the bubble agitation occurred at a short time (less than one second) and after that was a 10 s relaxation period until another cycle began. After the generation of the micro-bubbles, the fluorescence intensity of the solution was recorded with an
Fig. 8. Variation of fluorescence intensity as a function of time for molecular beacon hybridization with micro-bubble agitation: (a) micro-bubbles were generated form 2 × 1 heater array; (b) micro-bubbles were generated form 2 × 2 heater array. The relative standard deviation for the experimental data is 6%.
Olympus BX41 microscope at different times. The hybridization assay was repeated five times, and the results are shown in Fig. 8, where (a) corresponds to micro-bubble generation from the 2 × 1 bubble actuator array and (b) micro-bubble generation from the 2 × 2 bubble actuator array. Figs. 5 and 8 can be compared to explore the effects of the bubble agitation to the hybridization process because all the procedures for these two assays were the same except that Fig. 5 corresponds to the hybridization without bubble agitation, while Fig. 8 with bubble agitation. It is pertinent to note that the starting point “0” in the horizontal axis of Fig. 8 lags 15–20 min behind that of Fig. 5. The reason is that for Fig. 5, the fluorescence was read immediately after adding the molecular beacon and the beacon complement to the hybridization solution. While for Fig. 8, after adding the molecular beacon and the beacon complement to the hybridization solution it took some time, 15–20 min, for the operation of bubble generation and then the fluorescence recording began. The primary distinction between Figs. 5 and 8 is the slope of the fluorescence increase with time. For Fig. 5, as discussed in Section 3.1, the slope is larger at the first 2 h and becomes flattened after that, which is due to the gradually slowing down diffusion process of the molecular beacons
P. Deng et al. / Biosensors and Bioelectronics 21 (2006) 1443–1450
and the beacon complements. For Fig. 8, however, the fluorescence emission from the hybridization solution shows a uniform increase with time until reaching its maximum value. This implies that after micro-bubble agitation, the molecular beacons and the beacon complements had a relatively uniform distribution in the hybridization solution. Therefore, the hybridization process was enhanced by increasing the accessibility of the molecular beacons to the beacon complements. In addition to the increase in the fluorescence slope, the enhanced hybridization process with the introduction of micro-bubble actuation can also be verified from the reduced hybridization time. Fig. 5 shows that the hybridization time was approximately 4–6 h without bubble agitation, and it was shortened to about 3 h after 10 cycles of bubble generation from the 2 × 1 array as shown in Fig. 8(a), and it was further shortened to about 2–3 h after 10 cycles of bubble generation from the 2 × 2 array as shown in Fig. 8(b). Taking into account the 20 min delay for the operation of the micro-bubble actuator, the hybridization time, compared with that without bubble actuation, was distinctively shortened by approximately 33% with bubble agitation from the 2 × 1 bubble actuator array and 43% with bubble agitation from the 2 × 2 bubble actuator array. The reduced hybridization time can be attributed to the periodic flow perturbations as the micro-bubble generated in the hybridization solution, which effectively enhanced the diffusion of the DNA molecules and facilitated the hybridization process. In fact, this situation is similar to a traditional laboratory action—mixing the solution with a stirrer. The difference is that this time it was the micro-bubbles that acted as a virtual micro-stirrer. We know that all fluorescent dyes bleach over time under high intensity light illumination. The mechanism of the photobleaching phenomenon is still not well understood, but is due in part to the photochemical reactions induced by the light used for excitation (Picciolo and Kaplan, 1984). The interaction of the fluorophore with a combination of light and oxygen would produce some oxygen radicals, which could in turn react with the fluorophores and lead to a permanent destruction of them. It has been proved by test that the reduction of the oxygen in the test sample can decrease the photobleaching rate (Kishino and Yanagida, 1988; Smith, 1999) by scavenging oxygen with 50 mg/mL glucose oxidase, 0.1% glucose, and 10 mg/mL catalase. In this paper, the photobleaching might become serious when vapor bubbles were generated in the hybridization solution. This is because the transient high temperature accompanying the vapor bubble generation might accelerate the photodynamic process. For example, a decay of the fluorescence emission was observed in Fig. 8, where micro-bubble agitation was introduced in the hybridization solution. However, for the cases without bubble generation, the fluorescence intensity can be stable for several hours (see Fig. 5). Of course, anti-fade agents, such as the Fluorescence Antifade Kit (Molecular Probes Inc., Eugene, OR, USA), may help to reduce this kind of side effect in our study.
1449
4. Conclusions The experimental study of using 2D micro-bubble actuator array to generate flow perturbations in the DNA hybridization solution for the purpose of enhancing the DNA microbiosensor was reported for the first time in this paper. Two types of micro-heater arrays: 2 × 1 and 2 × 2 were distributed on the bottom of a millimeter-sized PDMS based reaction chamber for micro-bubble generation. The hybridization between molecular beacon and its complement was investigated. The periodic micro-bubble agitation in the static hybridization solution could effectively mix the molecular beacon and the beacon complements. As a result, the fluorescence emission from the hybridization solution shows a uniform increase with time. Compared to the hybridization time for hybridization without bubble agitation, it was distinctively shortened by 33% with 10 cycles of micro-bubble agitation from a 2 × 1 bubble actuator array, and by 43% with 10 cycles of micro-bubble agitation from a 2 × 2 bubble actuator array. The possible damage of the shear force and high temperature, accompanying with the vapor bubble formation, to the molecular beacon was experimentally studied by generating vapor bubbles in solution containing molecular beacons. No detectable malfunction of the molecular beacons was observed.
Acknowledgements This work was supported by the Hong Kong Research Grant Council (HKUST6014/02E and HKUST6134/04E).
References Anderson, M.L.M., 1999. Nucleic Acid Hybridization. Bios Scientific Publishers, New York. Bonnet, G., Tyagi, S., Libchaber, A., Kramer, F.R., 1999. Thermodynamic basis of the enhanced specificity of structured DNA probes. In: Proceedings of the National Academy of Sciences of the United States, vol. 96, pp. 6171–6176. Britten, R.J., Kohne, D.E., 1968. Repeated sequences in DNA. Science 161 (3841), 529–540. Deng, P., Lee, Y.-K., Cheng, P., 2003. The growth and collapse of a micro-bubble under pulse heating. Int. J. Heat Mass Trans. 46 (21), 4041–4050. Deng, P., Lee, Y.-K., Cheng, P., 2004. Micro-bubble dynamics in DNA solutions. J. Micromech. Microeng. 14 (5), 693–701. Deng, P., Lee, Y.-K., Cheng, P., 2005. Measurements of micro-bubble nucleation temperature in DNA solutions. J. Micromech. Microeng. 15, 564–574. Edman, C.F., et al., 1997. Electric field directed nucleic acid hybridization on microchips. Nucl. Acid Res. 25, 4907–4914. Erickson, D., Li, D., Krull, U.J., 2003. Modeling of DNA hybridization kinetics for spatially resolved biochips. Anal. Biochem. 317, 186–200. Fixe, F., Prazeres, D.M.F., Chu, V., Conde, J.P., 2003. Electric field pulse assisted covalent immobilization and hybridization of DNA in the nanosecond time scale for genetic information analysis. In: Transducers ’03 the 12th International Conference on Solid-State Sensors, Actuators and Microsystems, June 8–12, Boston, pp. 690–693.
1450
P. Deng et al. / Biosensors and Bioelectronics 21 (2006) 1443–1450
Freeman, W.M., Robertson, D.J., Vrana, K.E., 2000. Fundamentals of DNA hybridization arrays for gene expression analysis. Biotechniques 29 (5), 1042–1055. IDT online program 2004. OligoAnalyzer 3.0, Integrated DNA Technologies Inc., Coralville, IA, USA (http://207.32.43.70/biotools/oligocalc/ oligocalc.asp). Kishino, A., Yanagida, T., 1988. Force measurements by micromanipulation of a single actin filament by glass needles. Science 334, 74–76. Liu, X.J., Tan, W., 1999. A fiber-optic evanescent wave DNA biosensor based on novel molecular beacons. Anal. Chem. 71, 5054–5059. Marras, S., Kramer, F., Tyagi, S., 2003. Genotyping single nucleotide polymorphisms with molecular beacons. In: Kwok, P.Y. (Ed.), Single Nucleotide Polymorphisms: Methods and Protocols, 212. The Humana Press Inc., pp. 111–128. Mustafa, C., David, L.S., Guy, D.G., Tuan, V.D., 2004. Application of a miniature biochip using the molecular beacon probe in breast cancer gene BRCA1 detection. Biosens. Bioelectron. 19, 1007–1012. Okamoto, T., Suzuki, T., Yamamoto, N., 2000. Microarray fabrication with covalent attachment of DNA using bubble jet technology. Nat. Biotechnol. 18, 438–441. Pappaert, K., et al., 2003. Enhancement of DNA micro-array analysis using a shear-driven micro-channel flow system. J. Chromatogr. A 1014, 1–9.
Picciolo, G.L., Kaplan, D.S., 1984. Reduction of fading of fluorescent reaction product for microphotometric quantitation. Adv. Appl. Microbiol. 30, 197–234. Poddar, S.K., 1999. Detection of adenovirus using PCR and molecular beacon. J. Virol. Meth. 82, 19–26. Smith, D.E., 1999. Polymer physics experiments with single DNA molecules, PhD Dissertation, Department of Applied Physics, Stanford University, Stanford, CA, USA. Sosnowski, R., Heller, M.J., Tu, E., Forster, A.H., Radtkey, R., 2002. Active microelectronic array system for DNA hybridization, genotyping and pharmacogenomic applications. Psychiat. Genet. 12 (4), 181–192. Tenover, F.C., 1993. DNA hybridization techniques and their application to the diagnosis of infectious diseases. Infect. Disease Clin. North Am. 7 (2), 171–181. Tsourkas, A., Behlke, M.A., Rose, S.D., Bao, G., 2003. Hybridization kinetics and thermodynamics of molecular beacons. Nucl. Acids Res. 31 (4), 1319–1330. Tyagi, S., Kramer, F.R., 1996. Molecular beacons: probes that fluoresce upon hybridization. Nat. Biotechnol. 14, 303–308. Yuen, P.K., Li, G., Bao, Y., Muller, U.R., 2003. Microfluidic devices for fluidic circulation and mixing improve hybridization signal intensity on DNA arrays. Lab. Chip. 3 (1), 46–50.
Two-dimensional micro-bubble actuator array to enhance the efficiency of molecular beacon based DNA micro-biosensors Peigang Deng a , Yi-Kuen Lee a,∗ , Ping Cheng b a
Department of Mechanical Engineering, Hong Kong University of Science and Technology, Clear Water Bay, Kowloon, Hong Kong b School of Mechanical and Power Engineering, Shanghai Jiaotong University, Shanghai 200030, PR China Received 23 March 2005; received in revised form 14 June 2005; accepted 17 June 2005 Available online 11 August 2005
Abstract Two-dimensional micro-bubble actuator arrays were developed and studied in detail to enhance the hybridization kinetics of a DNA micro-biosensor. The hybridization between a molecular beacon, a kind of oligonucleotide probe, and its complement was investigated in a millimeter-sized PDMS based reaction chamber, where various 2D micro-heater arrays were distributed on the bottom for micro-bubble generation. The hybridization assay without the micro-bubble actuation revealed that the fluorescence increased fast at the beginning and slowed down after that. However, a uniform fluorescence increase was observed when periodic micro-bubble agitation was introduced in the static hybridization solution. A comparison of hybridization assays with and without micro-bubble agitation revealed that the hybridization time could be effectively shortened by 33% with 10 cycles of micro-bubble agitation from a 2 × 1 bubble actuator array, and by 43% with 10 cycles of micro-bubble agitation from a 2 × 2 bubble actuator array. © 2005 Elsevier B.V. All rights reserved. Keywords: Micro-bubble actuator; Molecular beacon; Flow perturbation; DNA hybridization; Micro-biosensor
1. Introduction DNA hybridization is the underlying principle of DNA biosensors. In DNA hybridization, the oligonucleotide probe (DNA probe) specifically recognizes, and binds to, a nucleic acid target (target DNA), which forms a double-stranded hybrid with its nucleic acid complement with high efficiency and specificity. The main types of hybridization used today are solution hybridization, filter hybridization, the polymerase chain reaction, in situ hybridization and hybridization on DNA chips (Tenover, 1993; Anderson, 1999; Freeman et al., 2000; Sosnowski et al., 2002). Hybridization in solution is believed to be a two-step process involving nucleation and zipping up, in which nucleation is the rate-limiting step, and a second-order reaction equation can be used to describe the process (Anderson, 1999).
∗
Corresponding author. Tel.: +852 2358 8663; fax: +852 2358 1543. E-mail address: [email protected] (Y.-K. Lee).
0956-5663/$ – see front matter © 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.bios.2005.06.007
Generally, the hybridization of DNA probes is diffusionlimited, i.e. the signal intensity is determined by the number of target molecules reaching the probe. However, diffusion is such an intrinsically slow process for large molecules that in micro-arrays even after overnight hybridization the system does not reach equilibrium. Furthermore, the slow diffusive transport could result in poor binding efficiency as less than 1% in micro-arrays (Pappaert et al., 2003). The diffusion limitation is a particular problem for a micro-scale biological reaction, where only small amounts of analyte are available and the required reaction volume limits the analyte concentration. The obvious solution to overcome the diffusion limitation of hybridization experiments is to agitate the sample solution. Recently, efforts have been devoted to increase the mixing and diffusion for DNA hybridization in a micro-scale confined space. Nanogen (San Diego, CA, USA) developed microchip-based hybridization arrays (NanoChipTM ) that utilized electric fields as an independent parameter to control DNA transport and enhance hybridization (Edman et al.,
1444
P. Deng et al. / Biosensors and Bioelectronics 21 (2006) 1443–1450
1997). Fixe et al. (2003) employed an electric field pulse to facilitate the DNA probe immobilization and to enhance the DNA hybridization process. Yuen et al. (2003) realized closed-loop fluid circulation and mixing by the rotation of a magnetic stirring bar. Other methods, which were used to enhance the DNA hybridization process, included sheardriven flows or surface acoustic waves (Pappaert et al., 2003). In our previous studies, a micro-thermal bubble actuator capable of generating effective flow perturbations in the bulk fluid was developed (Deng et al., 2003). More recently, transient micro-thermal bubble generation and its subsequent dynamic behavior in a 60-base single-stranded DNA (ssDNA) solution was investigated with particular emphases on the effects of DNA concentration and the total viscosity of the solution on bubble dynamics (Deng et al., 2004). It was found that the DNA macromolecules had a strong retardation effect on bubble dynamics at high concentrations. The bubble nucleation temperature at different DNA concentrations was measured by using the micro-Pt heater as a self-sensing resistive temperature sensor (Deng et al., 2005). The on-set bubble nucleation temperature was around 480 K for low DNA concentrations, and an increase in the DNA concentration could lead to a higher heating rate, a faster bubble nucleation process and a higher on-set nucleation temperature. In this paper, we will use a 2D micro-bubble actuator array to generate effective perturbations in the hybridization solution for the purpose of enhancing the DNA molecule diffusion and consequently the DNA hybridization rate. A type of oligonucleotide probe—molecular beacon (Tyagi and Kramer, 1996), was used as a DNA biosensor in the present study, and the hybridization assay of the molecular beacon with its complement was carried out and compared for the cases with and without bubble actuation.
2. Experimental 2.1. Design of the molecular beacon for DNA hybridization assay Since first reported by Tyagi and Kramer (1996), molecular beacons have been increasingly used in many applications (Poddar, 1999; Liu and Tan, 1999; Mustafa et al., 2004). As illustrated in Fig. 1, molecular beacons are dual-labeled
Fig. 1. Schematic representation of molecular beacon: (a) before hybridization, the molecule beacon remains non-fluorescent, because the fluorophore was quenched by the quencher; (b) after hybridization with the target the molecular beacon becomes fluorescent.
oligonucleotide probes that have a fluorescent dye molecule at one end and a fluorescence quencher at the opposite end. The probe is designed with a target-specific hybridization domain positioned centrally between short sequences that are self-complementary. In the absence of the target, the selfcomplementary domains anneal to form a stem–loop hairpin structure, that results in quenching of the reporter due to the fluorescence resonance energy transfer as shown in Fig. 1(a). In the presence of the target, the central loop domain will hybridize with the complementary target DNA or RNA, forcing the molecule to unfold; reporter and quencher are now physically separated and the fluorescence of the reporter dye will be restored upon excitation, as shown in Fig. 1(b). In this study, the sequence of the loop structure of the molecular beacon was designed according to a portion of the sequence of 16S subunit ribosomal RNA in E. coli (MC41000) and was 17 bases long. The stem part of the molecular beacon was six bases in length with a rich G/C percentage of 66%. The 5 end was labeled with Fluorescein (FAM) and the 3 end was labeled with Dabcyl (4-(4 -dimethylaminophenylazo) benzoic acid) quencher. The molecular beacon was synthesized by IDT (Integrated DNA Technologies Inc., Coralville, IA, USA), and the complete molecular beacon sequence was 5 -/56-FAM/ CAGTCGTATTA ACTTTACTCCCTCGACTG/3Dabcyl/3 , where the underlined nucleotides are the stem sequences. In the present design of the molecular beacon, a socalled “shared-stem” technology was used to make the beacon-target duplex more stable (Tsourkas et al., 2003). This design differs from a conventional molecular beacon in that one arm of the stem (the underlined and italic nucleotides near to the 5 end) participates in either hairpin formation or target hybridization. The beacon complement was also provided by IDT and the sequence was 5 TTTAGGGAGTAAAGTTAATACGACTG-3 . The stability of the molecular beacon-target duplex was estimated by using an IDT online program ‘OligoAnalyzer 3.0’ (IDT). The melting temperature is 51.8 ◦ C at the hybridization conditions of 0.25 M oligo, 50 mM monovalent salt. As to the stability of the self-complementary stem of the molecular beacon, 6 basepair-long G/C rich stems will melt between 60 and 65 ◦ C (Marras et al., 2003). 2.2. Micro-bubble actuator array Two kinds of 2D micro-heater arrays (2 × 1 and 2 × 2) were fabricated as shown in Fig. 2. The area of the slip part of each micro-heater of 2 × 1 array was 10 m × 3 m, and that of the 2 × 2 array was 3 m × 1 m. In our previous study, the effective perturbation region induced by microbubble generation in the bulk liquid was evaluated by putting micro-particles in the working fluid as the fluid tracer. This perturbation region was assumed to be a semi-sphere with its center at the center of the bubble. By examining the recorded dynamic response of the particles, the radius of the perturbation was estimated as six times that of the maximum bubble
P. Deng et al. / Biosensors and Bioelectronics 21 (2006) 1443–1450
1445
Fig. 3. PDMS based reaction chamber on a silicon chip of a micro-bubble actuator array.
Fig. 2. Two-dimensional micro-bubble actuator array: (a) 2 × 1 array; (b) 2 × 2 array; (c) close-up view of the micro-bubble.
radius (Deng et al., 2003). Therefore, the distance between each heater of the micro-bubble actuator array visualized in Fig. 2 was designed such that the effective flow perturbation area of each individual heater could not overlap much with its neighbor. The fabrication of the micro-bubble actuator array followed the conventional MEMS technology. Platinum was chosen as the material for both the heater and the electric pad. A n-type, double-sided polished, and 1 0 0 oriented silicon wafer (4 in. in diameter and 400 m in thickness) was used as the substrate. The wafer was coated with 2 m silicon dioxide on both sides of the wafer by wet thermal oxidation. A 0.02 m thick Ti film and a 0.18 m thick Pt film were sputtered on the wafer and patterned using the liftoff method. The fabricated micro-bubble actuator array was tested in the hybridization buffer. A 1.66-ms wide voltage pulse was imposed on each individual heater of the heater arrays (2 × 1 and 2 × 2). The height of the voltage pulse was increased from zero until micro-bubble was observed at a certain voltage, which is called “onset bubble nucleation”. The control circuit for the voltage pulse generation and the experimental set up for micro-bubble visualization were introduced in detail elsewhere (Deng et al., 2003, 2004). For the onset bubble nucleation, a stable single bubble could be generated on each individual heater of each heater array (2 × 1 and 2 × 2) simultaneously, as shown in Fig. 2. The micro-bubble grew and collapsed just on the heater surface (no detachment) due to the highly localized near homogeneous boiling mechanism for bubble nucleation (Deng et al., 2003). The bubble growth period was very similar to the heating pulse width, i.e. 1.66 ms in the present study, and the maximum bubble diameter was comparable to the size of the slim part of the micro-heater. The bubble began to collapse after the heating process stopped at the end of the 1.66 ms voltage pulse, at which time the bubble reached its maximum size. The bubble collapse process was much longer compared to the bubble growth process, which was mainly dominated by the vapor condensation in the subcooled liquid. The bub-
ble lifetime on heaters of a 2 × 1 array (10 m × 3 m) was around 320 ms, and it was around 45 ms on heaters of a 2 × 2 array (3 m × 1 m). The bubble nucleation temperature on the micro-heaters was evaluated at around 485 K from our previous study (Deng et al., 2003, 2005), and the superheated region (over 373 K) in the liquid was the highly localized area adjacent to the micro-heater surface (12 m × 3 m × 4 m for the 10 m × 3 m heater, and 3 m × 1 m × 1 m for the 3 m × 1 m heater). Note that for the present transient micro-bubble generation, the superheat region of the liquid not only was restricted to a very local area but also lasted for a short period. A series of fundamental research studies on micro-bubble generation under pulse heating in water, DNA buffers and DNA solutions have been carried out in our previous studies (Deng et al., 2003, 2004, 2005) in order to design an optimized array of micro-bubble actuators for DNA hybridization. 2.3. Fabrication of the PDMS reaction chamber Fig. 3 shows the picture of a micro-bubble actuator chip, where the reaction chamber was made of PDMS (Sylgard184, Dow Corning, USA). To precisely control the amount of the molecular beacons and the beacon complements in the hybridization solution, the dimension of the reaction chamber was designed at millimeter level, and thus the conventional pipette operation was possible. The PDMS reaction chamber was of a cylindrical shape with a diameter of 8.5 mm and a height of 1.8 mm. The total volume of the reaction chamber was about 100 L. To avoid direct exposure of the DNA hybridization buffer with the micro-Pt heater, a 60 nm thick parylene N (diX-NTM , Daisan Kasei Co., Japan) layer was coated on the packaged chip, which served as an electrical and chemical insulating layer. The procedure for the Parylene N coating was described in our previous paper (Deng et al., 2004). 2.4. DNA sample preparation The lyophilized molecular beacon and beacon complement were first centrifuged after keeping them at room
1446
P. Deng et al. / Biosensors and Bioelectronics 21 (2006) 1443–1450
temperature for 1 h. Then they were dissolved completely in distilled water (UltraPure and DNase/RNase free, Invitrogen, USA) so as to obtain three stock concentrations: 100, 10 and 1 M. Before hybridization assay, the concentrated oligonucleotides were further diluted with the distilled water to the working concentration.
3. Results and discussions 3.1. Molecular beacon hybridization without micro-bubble actuation The hybridization assay in the PDMS reaction chamber was first operated without the micro-bubble actuation. To start the hybridization assay, 75.2 L of hybridization buffer was loaded in the PDMS reaction chamber, and then 1.6 L of molecular beacon and 3.2 L of beacon complement were added to the hybridization buffer. Thus, the total volume of the hybridization solution was 80 L, where the concentrations of the molecular beacon and the beacon complement were both 100 nM, and the hybridization solution contained 50 mM KCl, 5 mM MgCl2 , 10 mM Tris–HCl, pH value = 7.5. Note that there was no mechanical mixing of the molecular beacon and the beacon complement involved. A 0.9 mm thick PDMS cover was put on the reaction chamber to avoid evaporization of the liquid. The liquid loss due to the permeability of the PDMS was tested by filling the reaction chamber with the hybridization buffer, covering it with the PDMS cover and then putting the whole system onto an electric balance. It was found that the loss of the weight was less than 0.1% after 24 h, which means the liquid loss due to the permeability of the PDMS is negligible. After loading the molecular beacon and the beacon complement in the hybridization buffer, the whole system was then put on the stage of an Olympus BX41 microscope, and the hybridization solution was excited with a 475 nm laser and the emission filter was set at 510 nm. The image of the fluorescence emission at different times was recorded by a CCD camera (7.2 Color Mosaic, Diagnostic Instruments Inc., USA), where the exposure time was fixed at 500 ms. In all the fluorescence intensity measurements the microscope was always focused on the micro-heater array at the bottom of the reaction chamber to minimize any potential experimental error. After the 12-bit, 800 × 600 pixels, grayscale image was obtained, as shown in Fig. 4, it was analyzed by a MatlabTM program. To calculate the effective fluorescence intensity from the image, firstly a rectangle two or three times larger than the area of the micro-heater array was defined, as indicated by the red rectangle box in Fig. 4. Then, the intensity value at each pixel inside this rectangle was accumulated and then averaged by the total pixels to get an averaged intensity value, which was regarded as the intensity of the fluorescence emission of the hybridization solution at the time when the image was recorded. Errors due to the focusing and setting the florescence calculation area could result in an average rel-
Fig. 4. Typical image of fluorescence emission of molecular beacon hybridization in a PDMS reaction chamber.
ative error of ±5% in the measured value of the fluorescence intensity. The hybridization assay was repeated five times, and Fig. 5 shows the variation of the fluorescence intensity with respect to time as the hybridization assay occurred in the PDMS reaction chamber without the micro-bubble actuation. Obviously, after loading the molecular beacon and the beacon complement in the hybridization buffer, as time elapsed, more and more molecular beacons could meet the beacon complements and hybridize with them. Thus the intensity of the fluorescence emission increased gradually, as shown in Fig. 5. The gradually increasing fluorescence intensity finally reached a saturated value (maximum value), which means most of the molecular beacons have successfully hybridized with their complements and a dynamic equilibrium has reached. It is important to examine the change of the slope of the fluorescence intensity increase with respect to time in Fig. 5, which shows that the slope is much sharper in the first 2 h, while it flattened after that. This actually implies that the hybridization of the molecular beacons and the beacon complements was a fast process in the first 2 h, while slowed down after that. We know that as time elapsed, the free molecular beacons and beacon complements became less and less because of the formation of the hybridized duplexes, which also increased the diffusion obstacle. As a result, the diffusion process slowed down gradually and thus reduced the
Fig. 5. Fluorescence intensity as a function of time for molecular beacon hybridization without micro-bubble actuation. The relative standard deviation for the experimental data is 12.5%.
P. Deng et al. / Biosensors and Bioelectronics 21 (2006) 1443–1450
accessibility of the molecular beacons and the beacon complements and subsequently the hybridization rate. In this paper, the hybridization time is defined as the time required for the fluorescence intensity of the hybridization solution to reach the saturated value (maximum value). From Fig. 5, it required about 4–6 h for the fluorescence intensity to reach a saturated value. Thus, the hybridization time was approximately 4–6 h in this case, which is, of the same order of magnitude, as that deduced from the plot of the re-association curve for double-stranded nucleic acids with various nucleotide pairs (Britten and Kohne, 1968). Note that detailed simulation to predict the hybridization time is possible by including the coupled mass, momentum, thermal transport and the DNA hybridization kinetics (Erickson et al., 2003). However, this is not the major objective of this paper. 3.2. Assessment of the molecular beacon damage due to bubble generation Although thermal bubble-jet printing technology has been successfully employed to eject DNA segments onto a glass surface for DNA micro-array (Okamoto et al., 2000), caution must be applied when generating thermal bubbles in the DNA solutions. From previous studies (Deng et al., 2003, 2005), we know that accompanying with the vapor bubble generation, it would induce an effective flow perturbation and a transient high temperature region in a very localized area in the liquid. Both of these might lead to the “opening” of the chemical bond of the base-paired “stem part” of the molecular beacon even when there is no beacon complement in the solution. This type of malfunction of the molecular beacon, if it happens, would result in fluorescence increase in the hybridization solution and should be deducted from the fluorescence obtained in the bubble enhanced hybridization assay. Therefore, before introducing the micro-bubble actuation in the hybridization solution, it is necessary to evaluate whether this process would induce any damage to the molecular beacons. Two assays were designed to test this damage issue. In the first assay, 75.2 L of hybridization buffer was loaded in one reaction chamber, and then 1.6 L of molecular beacon and 3.2 L of double-distilled H2 O were added into the hybridization buffer. The concentration of the molecular beacon was 100 nM, and the hybridization buffer contained 50 mM KCl, 5 mM MgCl2 , and 10 mM Tris–HCl, pH value = 7.5. In the second assay, the same solution was prepared in another reaction chamber and then micro-bubbles were generated simultaneously from the four heaters of a 2 × 2 heater array. A total of 10 heating pulses were employed and a 10 s waiting pulse was inserted between two consecutive heating pulses. In the experiment, the hybridization buffer was loaded to the two reaction chambers at the same time, and then bubble actuation was introduced in one of them. After that, the fluorescence emission from the hybridization buffer in two reaction chambers were recorded and compared.
1447
Fig. 6. Fluorescence emission from the solution containing molecular beacon with and without micro-bubble agitation. The data were from eight sets of experiments, and the relative standard deviation is 5%.
The two assays with and without bubble actuation were repeated eight times and the results are shown in Fig. 6. Basically, there is no distinct difference in the fluorescence emission for each case. In six of the total of eight cases the average fluorescence emission from solutions without bubble agitation was even a little bit larger than that with bubble agitation. For the other two cases, however, the situation was reversed and the fluorescence emission was a little bit larger from solutions with bubble agitation. Note that the differences for all the cases are still within the range of the error bar as shown in Fig. 6. As the temperature rises, the hairpin stem of the molecular beacon will unravel into a fluorescent randomly coiled oligonucleotide (Bonnet et al., 1999). As discussed in Section 2, the melting temperature of the self-complementary stem of the molecular beacon, 6 basepair-long and G/C rich in this study, is between 60 and 65 ◦ C (Marras et al., 2003). When vapor bubbles were generated in the solution containing molecular beacons, the localized shear force and high temperature might damage the molecular beacon by breaking the chemical bonds. As a result, the stem part of the molecular beacon would be cleaved and the molecular beacon would become a fluorescent randomly coiled oligonucleotide. However, if the molecular beacons were not broken into small pieces of oligonucleotides, the cleaved stem part of the molecular beacon could re-associate again as long as the flow perturbation stopped and the high temperature went back to the room temperature. This might be the reason that there was no distinct fluorescence increase from the solution where molecular beacons were subjected to micro-bubble agitations. It implies that the micro-bubble actuation in the hybridization solution would not lead to detectable malfunction of the molecular beacon. 3.3. Micro-bubble actuation for DNA hybridization enhancement The molecular beacon hybridization with micro-bubble actuation was conducted in the PDMS reaction chamber.
1448
P. Deng et al. / Biosensors and Bioelectronics 21 (2006) 1443–1450
Fig. 7. Flow perturbation induced by micro-bubble actuation.
The hybridization assay followed exactly the same procedure as described in Section 3.1. After the preparation of the hybridization solution, micro-bubbles were generated in the static hybridization solution to produce flow perturbations for the purpose of increasing the accessibility of the molecular beacons to the beacon complements, and thus enhancing the DNA hybridization process. As discussed in Section 2.2, two kinds of 2D micro-bubble actuator arrays, 2 × 1 and 2 × 2, were employed for bubble agitation in the hybridization solution. A 1.66-ms wide voltage pulse was imposed on each individual heater of each heater array for onset bubble nucleation. A 10 s waiting pulse was inserted in between two consecutive heating pulses. Thus, as the micro-bubble actuator worked, micro-single bubbles were generated on each individual heater of a heater array simultaneously every 10 s. Fig. 7 shows the visualization of the flow perturbation as a micro-bubble was generated on a micro-heater of 10 m × 3 m, where 0.96 m particles were placed in the liquid as a tracer. Because of the asymmetry of the bubble expansion and collapse process, a net displacement of the particles was achieved, indicating an effective flow perturbation by micro-bubble agitation. A total of 10 cycles of the micro-bubble actuation was introduced in the hybridization solution. For each cycle, the bubble agitation occurred at a short time (less than one second) and after that was a 10 s relaxation period until another cycle began. After the generation of the micro-bubbles, the fluorescence intensity of the solution was recorded with an
Fig. 8. Variation of fluorescence intensity as a function of time for molecular beacon hybridization with micro-bubble agitation: (a) micro-bubbles were generated form 2 × 1 heater array; (b) micro-bubbles were generated form 2 × 2 heater array. The relative standard deviation for the experimental data is 6%.
Olympus BX41 microscope at different times. The hybridization assay was repeated five times, and the results are shown in Fig. 8, where (a) corresponds to micro-bubble generation from the 2 × 1 bubble actuator array and (b) micro-bubble generation from the 2 × 2 bubble actuator array. Figs. 5 and 8 can be compared to explore the effects of the bubble agitation to the hybridization process because all the procedures for these two assays were the same except that Fig. 5 corresponds to the hybridization without bubble agitation, while Fig. 8 with bubble agitation. It is pertinent to note that the starting point “0” in the horizontal axis of Fig. 8 lags 15–20 min behind that of Fig. 5. The reason is that for Fig. 5, the fluorescence was read immediately after adding the molecular beacon and the beacon complement to the hybridization solution. While for Fig. 8, after adding the molecular beacon and the beacon complement to the hybridization solution it took some time, 15–20 min, for the operation of bubble generation and then the fluorescence recording began. The primary distinction between Figs. 5 and 8 is the slope of the fluorescence increase with time. For Fig. 5, as discussed in Section 3.1, the slope is larger at the first 2 h and becomes flattened after that, which is due to the gradually slowing down diffusion process of the molecular beacons
P. Deng et al. / Biosensors and Bioelectronics 21 (2006) 1443–1450
and the beacon complements. For Fig. 8, however, the fluorescence emission from the hybridization solution shows a uniform increase with time until reaching its maximum value. This implies that after micro-bubble agitation, the molecular beacons and the beacon complements had a relatively uniform distribution in the hybridization solution. Therefore, the hybridization process was enhanced by increasing the accessibility of the molecular beacons to the beacon complements. In addition to the increase in the fluorescence slope, the enhanced hybridization process with the introduction of micro-bubble actuation can also be verified from the reduced hybridization time. Fig. 5 shows that the hybridization time was approximately 4–6 h without bubble agitation, and it was shortened to about 3 h after 10 cycles of bubble generation from the 2 × 1 array as shown in Fig. 8(a), and it was further shortened to about 2–3 h after 10 cycles of bubble generation from the 2 × 2 array as shown in Fig. 8(b). Taking into account the 20 min delay for the operation of the micro-bubble actuator, the hybridization time, compared with that without bubble actuation, was distinctively shortened by approximately 33% with bubble agitation from the 2 × 1 bubble actuator array and 43% with bubble agitation from the 2 × 2 bubble actuator array. The reduced hybridization time can be attributed to the periodic flow perturbations as the micro-bubble generated in the hybridization solution, which effectively enhanced the diffusion of the DNA molecules and facilitated the hybridization process. In fact, this situation is similar to a traditional laboratory action—mixing the solution with a stirrer. The difference is that this time it was the micro-bubbles that acted as a virtual micro-stirrer. We know that all fluorescent dyes bleach over time under high intensity light illumination. The mechanism of the photobleaching phenomenon is still not well understood, but is due in part to the photochemical reactions induced by the light used for excitation (Picciolo and Kaplan, 1984). The interaction of the fluorophore with a combination of light and oxygen would produce some oxygen radicals, which could in turn react with the fluorophores and lead to a permanent destruction of them. It has been proved by test that the reduction of the oxygen in the test sample can decrease the photobleaching rate (Kishino and Yanagida, 1988; Smith, 1999) by scavenging oxygen with 50 mg/mL glucose oxidase, 0.1% glucose, and 10 mg/mL catalase. In this paper, the photobleaching might become serious when vapor bubbles were generated in the hybridization solution. This is because the transient high temperature accompanying the vapor bubble generation might accelerate the photodynamic process. For example, a decay of the fluorescence emission was observed in Fig. 8, where micro-bubble agitation was introduced in the hybridization solution. However, for the cases without bubble generation, the fluorescence intensity can be stable for several hours (see Fig. 5). Of course, anti-fade agents, such as the Fluorescence Antifade Kit (Molecular Probes Inc., Eugene, OR, USA), may help to reduce this kind of side effect in our study.
1449
4. Conclusions The experimental study of using 2D micro-bubble actuator array to generate flow perturbations in the DNA hybridization solution for the purpose of enhancing the DNA microbiosensor was reported for the first time in this paper. Two types of micro-heater arrays: 2 × 1 and 2 × 2 were distributed on the bottom of a millimeter-sized PDMS based reaction chamber for micro-bubble generation. The hybridization between molecular beacon and its complement was investigated. The periodic micro-bubble agitation in the static hybridization solution could effectively mix the molecular beacon and the beacon complements. As a result, the fluorescence emission from the hybridization solution shows a uniform increase with time. Compared to the hybridization time for hybridization without bubble agitation, it was distinctively shortened by 33% with 10 cycles of micro-bubble agitation from a 2 × 1 bubble actuator array, and by 43% with 10 cycles of micro-bubble agitation from a 2 × 2 bubble actuator array. The possible damage of the shear force and high temperature, accompanying with the vapor bubble formation, to the molecular beacon was experimentally studied by generating vapor bubbles in solution containing molecular beacons. No detectable malfunction of the molecular beacons was observed.
Acknowledgements This work was supported by the Hong Kong Research Grant Council (HKUST6014/02E and HKUST6134/04E).
References Anderson, M.L.M., 1999. Nucleic Acid Hybridization. Bios Scientific Publishers, New York. Bonnet, G., Tyagi, S., Libchaber, A., Kramer, F.R., 1999. Thermodynamic basis of the enhanced specificity of structured DNA probes. In: Proceedings of the National Academy of Sciences of the United States, vol. 96, pp. 6171–6176. Britten, R.J., Kohne, D.E., 1968. Repeated sequences in DNA. Science 161 (3841), 529–540. Deng, P., Lee, Y.-K., Cheng, P., 2003. The growth and collapse of a micro-bubble under pulse heating. Int. J. Heat Mass Trans. 46 (21), 4041–4050. Deng, P., Lee, Y.-K., Cheng, P., 2004. Micro-bubble dynamics in DNA solutions. J. Micromech. Microeng. 14 (5), 693–701. Deng, P., Lee, Y.-K., Cheng, P., 2005. Measurements of micro-bubble nucleation temperature in DNA solutions. J. Micromech. Microeng. 15, 564–574. Edman, C.F., et al., 1997. Electric field directed nucleic acid hybridization on microchips. Nucl. Acid Res. 25, 4907–4914. Erickson, D., Li, D., Krull, U.J., 2003. Modeling of DNA hybridization kinetics for spatially resolved biochips. Anal. Biochem. 317, 186–200. Fixe, F., Prazeres, D.M.F., Chu, V., Conde, J.P., 2003. Electric field pulse assisted covalent immobilization and hybridization of DNA in the nanosecond time scale for genetic information analysis. In: Transducers ’03 the 12th International Conference on Solid-State Sensors, Actuators and Microsystems, June 8–12, Boston, pp. 690–693.
1450
P. Deng et al. / Biosensors and Bioelectronics 21 (2006) 1443–1450
Freeman, W.M., Robertson, D.J., Vrana, K.E., 2000. Fundamentals of DNA hybridization arrays for gene expression analysis. Biotechniques 29 (5), 1042–1055. IDT online program 2004. OligoAnalyzer 3.0, Integrated DNA Technologies Inc., Coralville, IA, USA (http://207.32.43.70/biotools/oligocalc/ oligocalc.asp). Kishino, A., Yanagida, T., 1988. Force measurements by micromanipulation of a single actin filament by glass needles. Science 334, 74–76. Liu, X.J., Tan, W., 1999. A fiber-optic evanescent wave DNA biosensor based on novel molecular beacons. Anal. Chem. 71, 5054–5059. Marras, S., Kramer, F., Tyagi, S., 2003. Genotyping single nucleotide polymorphisms with molecular beacons. In: Kwok, P.Y. (Ed.), Single Nucleotide Polymorphisms: Methods and Protocols, 212. The Humana Press Inc., pp. 111–128. Mustafa, C., David, L.S., Guy, D.G., Tuan, V.D., 2004. Application of a miniature biochip using the molecular beacon probe in breast cancer gene BRCA1 detection. Biosens. Bioelectron. 19, 1007–1012. Okamoto, T., Suzuki, T., Yamamoto, N., 2000. Microarray fabrication with covalent attachment of DNA using bubble jet technology. Nat. Biotechnol. 18, 438–441. Pappaert, K., et al., 2003. Enhancement of DNA micro-array analysis using a shear-driven micro-channel flow system. J. Chromatogr. A 1014, 1–9.
Picciolo, G.L., Kaplan, D.S., 1984. Reduction of fading of fluorescent reaction product for microphotometric quantitation. Adv. Appl. Microbiol. 30, 197–234. Poddar, S.K., 1999. Detection of adenovirus using PCR and molecular beacon. J. Virol. Meth. 82, 19–26. Smith, D.E., 1999. Polymer physics experiments with single DNA molecules, PhD Dissertation, Department of Applied Physics, Stanford University, Stanford, CA, USA. Sosnowski, R., Heller, M.J., Tu, E., Forster, A.H., Radtkey, R., 2002. Active microelectronic array system for DNA hybridization, genotyping and pharmacogenomic applications. Psychiat. Genet. 12 (4), 181–192. Tenover, F.C., 1993. DNA hybridization techniques and their application to the diagnosis of infectious diseases. Infect. Disease Clin. North Am. 7 (2), 171–181. Tsourkas, A., Behlke, M.A., Rose, S.D., Bao, G., 2003. Hybridization kinetics and thermodynamics of molecular beacons. Nucl. Acids Res. 31 (4), 1319–1330. Tyagi, S., Kramer, F.R., 1996. Molecular beacons: probes that fluoresce upon hybridization. Nat. Biotechnol. 14, 303–308. Yuen, P.K., Li, G., Bao, Y., Muller, U.R., 2003. Microfluidic devices for fluidic circulation and mixing improve hybridization signal intensity on DNA arrays. Lab. Chip. 3 (1), 46–50.
