- Email: [email protected]
Manipulation and extraction of genomic DNA from cell lysate by functionalized magnetic particles for lab on a chip applications
Biosensors and Bioelectronics 21 (2006) 989–997
Manipulation and extraction of genomic DNA from cell lysate by functionalized magnetic particles for lab on a chip applications Siu Wai Yeung, I-Ming Hsing ∗ Department of Chemical Engineering, Hong Kong University of Science and Technology, Clear Water Bay, Kowloon, Hong Kong Received 18 January 2005; received in revised form 9 March 2005; accepted 11 March 2005 Available online 6 April 2005
Abstract A novel approach for extracting living cells’ genomic DNA materials utilizing functionalized magnetic particles (MPs) is reported in this investigation. This strategy is amenable to handle bio-samples in a miniaturized environment and it offers a possibility to separate and purify DNA from other cell lysate mixtures “on-chip”, which is known to be a bottle-neck step in an integrated micro-total-analysis-system (TAS). “Species-specific” genomic DNA of interest is captured by the MPs based on the hybridization interaction between the biotinylated probes modified MPs and a complementary region of the targeted genome. The genome DNA anchored on the particles can be separated from the rest of cellular mixtures by a simple buffer washing upon the exertion of external magnetic force. Surface modifications of MPs and hybridization conditions affecting the genome capturing efficiency are investigated. Extraction of genomic DNA from E. coli is demonstrated in a silicon/glass-based micro-reactor patterned with a platinum heater and sensors. On-chip extraction and manipulation of genomic DNAs illustrated in this study is a step forward toward a total integrated bioanalytical microsystem for crude cells/sample analysis. © 2005 Elsevier B.V. All rights reserved. Keywords: Magnetic particle; DNA hybridization; Genome extraction; Microchip
1. Introduction Polymerase chain reaction (PCR), DNA sequencing techniques, together with fluorescence-based or electrochemicalbased detection methods, have been developed to analyze genetic materials (e.g. DNAs). However, to directly access the genomic materials contained within the cell envelope, tedious procedures of cell lysis, DNA extraction and purification are often performed. Most of the commercial kits for genome extraction (e.g. the silica membranes or conventional solvent processes) involve laborious routine steps of pipetting, vortexing and centrifugation in order to remove cell debris and to retrieve highly purified genomic DNAs. These procedures are complex and not amenable for the portable, lab-on-a-chip applications such as environmental monitoring, point-of-care diagnostics, and prevention of bio-terrorism. In our view, to realize the advantages of a portable bioanalytical microsys∗
Corresponding author. Tel.: +852 23587131; fax: +852 23580054. E-mail address: [email protected] (I.-M. Hsing).
0956-5663/$ – see front matter © 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.bios.2005.03.008
tem, a DNA-based micro-total-analysis-system (TAS), also called lab on a chip, should have the ability to perform sample preparation, DNA amplification, and sequence-specific detection (either based on electrochemical or fluorescent reporters) on chip. Chip-based techniques for cell lysis (Han et al., 2003; Taylor et al., 2001; Carlo et al., 2003), PCR (Northrup et al., 1993; Poser et al., 1997; Lee et al., 2000; Schneegaß et al., 2001) and capillary electrophoresis (Lagally et al., 2000; Nicole and Brazill, 2003; Huang et al., 2004) were largely demonstrated. Among them, the on-chip sample preparation technique (e.g. separating genetic materials from crude cell samples) is known to be the most challenging task for an integrated lab-on-a-chip system. Although novel DNA extraction methods utilizing adsorptive silica beads (Breadmore et al., 2003) and dielectrophoretic forces (Prinz et al., 2002; Gao et al., 2004; Gunasekera et al., 2004) were reported, on-chip integration issues were largely not addressed. One of chip-amenable approaches would be the use of magnetic particles (MPs), which
990
S.W. Yeung, I.-M. Hsing / Biosensors and Bioelectronics 21 (2006) 989–997
were demonstrated for the protein separation (Bucak et al., 2003; Hubbuch and Owen, 2002), the bacterial cell accumulation (Gu et al., 2003), electrical detection of DNA (Wang et al., 2003; Weizmann et al., 2003) and DNA extraction from crude blood samples by polyvinyl–alcohol-based (Oster et al., 2001) and hyperbranched polyamidoamine dendrimer modified (Yoza et al., 2003) MPs. Most of the published methods relied on the electrostatic interaction of the genomic materials and functionalized surfaces of magnetic particles for the genome extraction. Their approaches did not provide “species-specific” capturing of genomic DNAs and interferences of extraction by other cellular components (e.g. RNA and protein) are hardly avoidable. The DNA hybridization process can be used for the sequence-specific recognition and in recent years this method has been applied for the extraction and purification of DNAs. A few studies were previously reported for the capturing of mRNA via polyA–polyT interaction (Carlo and Lee, 2002) and for the extraction of plasmid DNA (Matteo et al., 2003). Moreover, the collection of trace amounts of DNA/mRNA using MPs functionalized by molecular beacons was recently demonstrated (Zhao et al., 2003). Magnetic particles modified by species-specific probes opens a possibility to capture/extract specific genome/DNA on chip. The genomic DNAs anchored on the MPs can be easily separated from the rest of cellular components by a simple buffer washing when an external magnetic force is applied. In addition, the release of the captured DNA from MPs can be effectively controlled by the on-chip heater, instead of using running buffers at different ionic strengths. In our work, the direct extraction/capturing of genomic DNA from cultured E. coli samples is demonstrated in a silicon-based micro-reactor with functionalized MP capturers. Factors, such as probe concentration and density, the sequences of probes and the states of hybridization that may impact the genome capturing efficiency are investigated. Ag- and Au-based sequence-specific electrochemical detection of PCR amplicons (Li et al., 2004; Cai et al., 2004; Lee et al., 2004) using the genome template extracted by MPs is studied. Finally, the cell lysis and genome capturing using MPs are demonstrated on integrated silicon-based microchip where the on-chip heater and sensors are used to control the temperatures for the cell lysing and genome probe hybridization.
operon (acc. no.: X07850) with the following primers: ECfwd (5 -GAC-AAG-AAA-ATC-TCC-AAC-ATC-C-3 ) and, EC-rev (5 -ACA -ACA-CGT-TTA-GCC-TGA-CC-3 ). The sequences of DNA probes used to demonstrate the capturing of E. coli genome via hybridization were 5 end biotinylated forward primers (biotin-EC-fwd) and a biotinylated random nucleotide sequences (biotin-R12, 5 -biotin-NNNNNN-NNN-NNN-3 ). LB agar and LB broth were purchased from Amersham Bioscience. DNA purification kit was ordered from Promega (catalog no. A1120). Oligonucleotides bearing a pyrrole group at 5 end (pyrrole-reverse primer, pEC-rev) for electrochemical detection were ordered from APIBIO (Cedex, France). Streptavidin/colloidal gold (5 nm) and sodium nitrate were obtained from Sigma (St. Louis, MO). Silver nitrate was ordered from Fisher Chemicals. Lithium hyperchlorate was obtained from Fluka. Sodium chloride, trisodium citrate dehydrate, disodium hydrogen phosphate, pyrrole monomer and HEPES (N-[2-Hydroxyethyl] piperazine-N -[2-ethanesulfonic acid] were purchased from Sigma-Aldrich (Germany). All solutions were prepared with ultrapure water from a Millipore Milli-Q system. 2.2. Instrumentation PCR was carried out in an Eppendorf® Mastercycler. Ultraviolet (UV) transillumination of agarose gel electrophoresis was performed by a Gel Documentation System (Alpha Innotech Corp.). DNA concentrations were measured by an Eppendorf® BioPhotometer. The control system for chip-based genome extraction consisted of signal conditioning board (SC-2042-RTD, National Instruments), a data acquisition DAQ card (PCI-MIO-16-1, National Instruments), a power supply unit (HP6629A), and a GPIB card (HP82350A). They were connected together for the computational control by a LabVIEW program (National Instruments, Texas, USA). The electrochemical detection methods: cyclic voltammetry (CV), chronoamperometry (CA) and potentiometric stripping analysis (PSA) were performed with a 16-channel VMP Multichannel Potentiostats/Galvanostats (PerkinElmer Instruments, USA) controlled by the EC-Lab (V6.70) software (Bio-Logic-Science Instruments). 2.3. Chip fabrication
2. Experimental 2.1. Reagents The avidin-coated MPs used in the experiment were purchased from Spherotech Inc. (catalog no. VMS-30-20). They are paramagnetic magnetic particles with average diameter of 3.0 m. The binding capacity of these MPs is 0.103 nmol of biotin-FITC to 1 mg of particles. PCR primers and biotinylated DNA probes were purchased from Invitrogen. During PCR, a 300 bps product is amplified from the E. coli groE
The silicon chip consists of one reaction chamber (∼8 L in volume) having the specific dimensions of 5 mm × 5 mm × 300 m (Fig. 1, inset I) and a patterned platinum heater and temperature sensors. The dimensions of heater line and sensor line are L: 3.6 cm × W: 600 m × H: 0.1 m and L: 0.49 cm × W: 20 m × H: 0.1 m, respectively (Fig. 1, inset II). After the chip was cleaned with acetone and deionized water in a ultrasonic bath (5 min for each washing step), a closed chamber was formed by bonding the chip with an indium tin oxide (ITO) glass via UV glue (Type UV-69,
S.W. Yeung, I.-M. Hsing / Biosensors and Bioelectronics 21 (2006) 989–997
991
Fig. 1. Chip temperature during the DNA extraction process (cell lysis and genome denaturation at 85 ◦ C, hybridization between denatured genome and the MP linked probe at 50 ◦ C). Insets: (I) Front side of silicon chip showing a square reaction chamber. (II) Back side of silicon chip showing the fabricated heater and temperature sensors. The silicon chip and glass were bonded together with UV glue.
Summers Optical, UV illumination: 90 min, vacuum oven: 2 h at 60 ◦ C). The ITO glass was drilled with holes for sample input and output and cleaned with deionized water before binding with the silicon chip. 2.4. Procedure 2.4.1. Functionalized MPs preparation To prepare the avidin-coated MPs for genome extraction via DNA–DNA pairing, 10 L of avidin-coated MPs were first extracted from the stock solution into Eppendorf tube and washed with 0.5 × SSC buffer. After removing the supernatant via centrifugation and pipetting, 10 L of biotinylated probes (biotin-EC-fwd) at different concentrations were added to the residual MPs and the well mixed mixture was then overnight incubated to achieve various probe densities on the particle surface. Finally, the DNA probes functionalized MPs were washed with 0.5 × SSC to remove the excess DNA-oligos and stored in 10 L 0.5 × SSC buffer at room temperature. 2.4.2. Tube-based genome extraction Two MPs-based genome extraction strategies were studied: solid phase hybridization approach and liquid phase hybridization approach. To start with, E. coli genome purified with DNA purification kit was adopted as the target DNA to illustrate the two proposed extraction strategies. In solid phase hybridization approach, base-pairing takes place at the MPs surface, as shown in Fig. 2a. In brief, each Eppendorf tube containing 15 L of 0.5 × SSC buffer, 1.0 L of probe-modified MPs, and 4.0 L of purified E. coli genome (1.0 ng/L) was first heated at 94 ◦ C for 5 min to denature the genome, followed by the incubation at 60 ◦ C for 10 min
to allow the hydrogen bonds formation between the immobilized probes and its complementary sequence on the denatured genomes to achieve DNA capturing. MPs-based genome capturing via liquid phase hybridization, as shown in Fig. 2b, was achieved by first mixing the genome samples (4 ng) with 1 L biotinylated probes (biotinEC-fwd, 1 nM) and 15 L of 0.5 × SSC buffer. After thermal denaturation (94 ◦ C for 5 min) and hybridization (60 ◦ C for 10 min), MPs were introduced to the samples and incubated for another 10 min to link the biotinylated probe-carrying genomic DNA to the particle surface via biotin–avidin interaction. In each approach, the DNA-attached particles were then washed with 0.5 × SSC buffer and separated from the washing buffer by an external magnet at the bottom of Eppendorf tube. Finally, PCR reagents without DNA template were added to the MPs to amplify a selected region of the extracted genome after they were released from the MPs during the initial denaturation step in PCR. 2.4.3. PCR protocols PCR reagents and Taq polymerase were purchased from Bioline Ltd. (catalog no. BIO-21040). For E. coli PCR experiment, reaction mixture contained 5 L of 10× PCR buffer, 1.0 L of 10 mM dNTP, 1.0 L of 50 mM MgCl2 , 2.0 L of forward and reverse primers (10 M), 0.5 L of 5 units/L Taq polymerase and 34.5 L of autoclaved double-deionized water. In the positive control, 4 ng of E. coli genome purified from commercial kit was added. In the tested samples, in which genomic DNA-captured MPs presented, ddH2 O was used instead to make up 50 L for PCR experiment. The master mix was subjected to the following thermal cycling profile: initial denaturation at 94 ◦ C for 5 min, 30 cycles at
992
S.W. Yeung, I.-M. Hsing / Biosensors and Bioelectronics 21 (2006) 989–997
Fig. 2. (a) Schematic diagram showing genome extraction by solid phase hybridization approach. Step 1: probe immobilization on MPs; Step 2: thermal denaturation of genome and its hybridization on the probes modified-MPs. (b) Schematic diagram showing genome extraction by liquid phase hybridization approach. Step 1: thermal denaturation of genome and genome-biotinylated probes hybridization; Step 2: addition of MPs for genome capturing via biotin–avidin interaction.
94 ◦ C for 30 s, at 60 ◦ C for 30 s, at 72 ◦ C for 30 s, and a final extension at 72 ◦ C for 5 min. PCR products were loaded in a 2% agarose gel (USB Corp. catalog no. 75817) prepared with 1 × TAE buffer for gel electrophoresis at 100 V for 1 h. The gel was immersed in ethidium bromide solution and visualized by UV transillumination. Asymmetric PCR was performed to produce singlestranded amplicons for the hybridization step in electrochemical detection. During amplification, concentration of EC-rev was adjusted to 0.1 M and the biotin-EC-fwd used is 10 M. It was cycled 40 times with the same temperature profile mentioned above. 2.4.4. On-chip genome extraction To demonstrate the compatibility of MPs-based genomic DNA capturing in the micro-environment, liquid phase hybridization approach was adopted and the 10 L sample was prepared with 0.5 L biotin-fwd probes (1 nM), 2.0 L of purified E. coli genomic DNA (1 ng/L) or living E. coli cells in LB (cultured overnight at 37 ◦ C), 7.5 L 0.5 × SSC buffer, 0.5 L BSA (20 mg/mL), 0.5 L dimethylformamide (DMF). Then, 8 L of well mixed sample was injected (bubble free) into the reaction chamber via the drilled hole. The injection holes are then sealed with a plastic tape (3M 5419,
3M Corporation) to prevent evaporation during heating and an insulation tape. Here, DMF was added to lower the denaturation temperature to 85 ◦ C for power consumption reduction. Once the chip was electrically connected, the chamber was heated to 85 ◦ C for 5 min and then 50 ◦ C for 10 min under computational control. MPs (1.0 L) were then injected into the chip via the injection hole. The MPs can be well dispersed by the internal mixing induced by the random movement of an external magnet in the chip area. The magnet was then removed for the incubation process. After 10 min incubation, genome-anchored MPs were washed three times by pipetting 0.5 × SSC buffer in and out of the chamber to wash out the unlinked genome and other cell components in the chamber. During washing, an external magnet was held under the chip to keep the MPs retained in the chamber. In the final wash, magnet was removed so that MPs were transported out of the chip for PCR amplification in a thermocycler. 2.4.5. Electrochemical detection 2.4.5.1. Probe immobilization and DNA hybridization. Electrochemical reaction was carried out on a glass chip electrode patterned with four Indium Tin Oxide (ITO) working electrodes, one platinum counter electrode and reference electrode. After the electrode-patterned glass
S.W. Yeung, I.-M. Hsing / Biosensors and Bioelectronics 21 (2006) 989–997
993
was cleaned with water and electrically connected to the VMP system, 10 L of freshly prepared reaction mixture (60 mM pyrrole, 0.1 M LiClO4 , and 20 M pEC-rev) was added to the electrodes surface for cyclic voltammetry to co-electropolymerize the oligonucleotide-linked pyrrole (pEC-rev) and pyrrole monomer (potential sweeping between −0.5 and 0.65 V at a scan rate of 50 mV/s for three cycles). The chip was then cleaned with water and dried by compressed nitrogen gas before the loading of asymmetric PCR products (0.5 L) on any two of four working electrodes for 2 h-incubation. Finally, chip is washed with 2 × SSC buffer to remove excess, nonspecifically bound PCR products. 2.4.5.2. Gold nanoparticle labeling. Gold nanoparticles were bound to the PCR amplicons by exposing the electrode to 0.5 L streptavidin-gold solution (the stock was 10 times diluted with HEPES buffer: 0.05 M HEPES and 0.2 M NaCl) at room temperature for 35 min. Any unbound gold nanoparticle labels were removed by rinsing with phosphate-buffered nitrate solution (PBN, 0.3 M NaNO3 /10 mM sodium phosphate, pH 7.0). 2.4.5.3. Silver electro-deposition and electrochemical stripping. Two cyclic voltammograms were scanned simultaneously from +0.2 to −1.0 V (scan rate: 100 mV/s) in 1.0 M KNO3 solution containing 1.0 mM AgNO3 on the probetarget electrode and probe only electrode. Because of the catalytic effect of gold nanoparticle (Lee et al., 2004), an appropriate silver electrodeposition potential was chosen where metallic silver only deposited on the probe-target electrode surface. After electrodeposition of silver onto the gold nanoparticle surface under the chosen potential for 20 s, the amount of silver deposited was determined by measuring the oxidative silver dissolution response in a potentiometric stripping analysis (PSA) scan. All the measurements were conducted with an applied anodic current of 10 A in the same silver nitrate solution at room temperature. 3. Result and discussion To miniaturize the sample preparation step of bio-analysis on a micro-fabricated chip, most of the current fluidic control schemes involved in washing will increase the complexity of the chip design and fabrication. On the other hand, the compatibility of MPs and the proposed genomic DNA extraction strategy to PCR simplify the DNA retrieval procedure such that the captured DNA can be released directly to the PCR mixture by breaking the hydrogen bonds between the probe and the genome during the initial denaturation step of PCR. In this paper, the amount of resulting PCR amplicons produced using the released genomic DNA from the MPs as the initial template in PCR is used to measure the extraction/capturing efficiency of the genome using MPs.
Fig. 3. (a) Gel electrophoresis photograph showing the influence of probe densities on the MPs on the genome capturing performance based on the solid phase hybridization approach. Lane 1: negative control (sample without DNA template); Lane 2: probe density (nmol probe/mg MPs) = 0.1; Lane 3: probe density, 0.01; Lane 4: probe density, 0.001; Lane 5: probe density, 1e-4; Lane 6: probe density, 1e-5; Lane 7: probe density, 0; Lane 8, PCR positive control. (b) Bar chart showing the relative band intensities in gel electrophoresis with the lane numbers corresponding to (a) to compare the influence of probe densities on MPs on the genome capturing performance based on the solid phase hybridization approach. The band intensity of positive control equals to 100%.
3.1. Characterization of probe modified magnetic particle for genomic DNA extraction To capture genome at a solid surface utilizing the hybridization process between a specific region of the genome and the complementary immobilized probe, the coverage of the DNA probes (i.e. probe density) on MPs is a deciding factor because of the anticipated hindrance effect of the large size genome and the strong electrostatic repulsion between negatively charged genome molecules. As it can be seen from Fig. 3a, the relative band intensity data of the gel electrophoresis results show that the genome extraction efficiency increases with the decrease of the probe density at the
994
S.W. Yeung, I.-M. Hsing / Biosensors and Bioelectronics 21 (2006) 989–997
range of 1.0 nmol/mg MP ∼ 1.0 × 10−3 nmol/mg MP. High probe coverage on the MPs could inhibit the probes to access the specific complementary binding site of the genome. As the probe density/coverage decreases, the large genome and the immobilized probes are given opportunities to rearrange among themselves to establish the hydrogen binding. (Fig. 3a, Lanes 2–4). With a further decrease of probe density on MPs, non-specific adsorption of genome could occur (Fig. 3a, Lanes 5–7) without improving the capturing process. This finding suggests that a sufficient amount of DNA probes should be immobilized to build up a resistant layer against non-specific adsorption. The genome capturing process based on the hybridization interaction between the genome and immobilized probe on MPs (the so-called solid phase hybridization approach) clearly requires a further improvement, as revealed by Fig. 3b. The optimal capturing efficiency relative to the positive control stays at around 30%. The gigantic circular bacterial genome with a complicated entanglement structure after thermal denaturation could explain the difficulty of the capturing process relying on the hybridization between a huge genome and tiny MP linked probes. 3.2. Capture genome by liquid phase approach To address the steric effect caused by the large genome structure and to improve the performance of MPs-based genome extraction, the liquid phase hybridization approach is adopted where the biotinylated probes are first introduced and hybridized with genomic DNA before being immobilized on MPs. MPs were pretreated to build up an oligo blocking layer (0.01 nmol/mg MPs) on the surface. This oligo layer can serve as a blocking layer to reduce the non-specific DNA adsorption as evidenced by a control experiment where the fluorescent signal can be seen from the complementary probe modified MPs, but not on the non-complementary probe modified MPs (Figure not shown). The fluorescein-labeled target DNA contributes to the fluorescent image of the complementary probe MPs after the hybridization. The absence of a fluorescent signal from the non-complementary probe MPs indicates the elimination of non-specific adsorption of DNAs by the oligo blocking layer. The relative band intensities of PCR products after the gel electrophoresis (Fig. 4) using the solid and liquid phase hybridization approaches suggest that when the liquid phase hybridization approach is employed, a much higher genome capturing efficiency can be achieved (∼75% of the relative to the positive PCR control) as compared to the solid phase one. However, in the liquid phase hybridization step the excessive use of capturing probes might also lower the DNA extraction performance. As displayed in Fig. 5, the amount of the PCR products amplified from the extracted genome increases with the continuous decrease of the capturing probe concentration from 10 M to 1.0 nM. Although, a high probe
Fig. 4. Gel electrophoresis photograph showing the effect of solid and liquid phase hybridization methods and the use of different DNA probes. Lanes 1 and 2: sequence-specific probe; Lanes 3 and 4: degenerated probe; Lanes 1 and 3: solid phase hybridization; Lanes 2 and 4 liquid phase hybridization; Lane 5: positive control.
concentration should increase the chance for these probes to anneal on the specific site of genomic DNA, nonetheless these biotinylated probes will also compete with the biotin reporters bound with hybridized genomes for the avidin sites of the MPs and thus lead to the reduction of the capturing efficiency. To characterize the hybridization process in the MP-based DNA extraction approach, biotin-R12 are allowed to capture E. coli genome according to the base pairing reaction between the degenerated sequences and the corresponding complements at the genome so that the one genome may carry several biotins for the extraction. Based on the respective PCR results, the relative band intensity of the gel electrophoresis as shown in Fig. 4, the difference in genome capturing
Fig. 5. Gel electrophoresis photograph showing the effect of the concentrations of liquid phase probes on the genome capture efficiency during the liquid phase hybridization process. Lane 1: probe concentration, 10 M; Lane 2: probe concentration, 1 M; Lane 3: probe concentration, 1 nM.
S.W. Yeung, I.-M. Hsing / Biosensors and Bioelectronics 21 (2006) 989–997
995
performances between degenerated probes and sequencespecific probes is not as obvious as expected. Therefore, it is not conclusive to state that the randomized hybridization process using degenerated probes provides any noticeable improvement in the DNA extraction process. One possible explanation is that the hybridized genome with multiple biotin labels resulted from the random hybridization might take away more than one avidin binding sites, thus decreasing the available binding sites of the MPs. It is possible that large surface areas when smaller MPs are used could contribute more biotin binding sites and thus improve the extraction process. 3.3. On-chip genome extraction from biological sample The realization of a fully integrated DNA-based bioanalytical microsystem requires the various bioprocessing modules for the sample preparation, target amplification and product detection to be integrated on chip. In this study, on-chip genomic DNA extraction using MPs is to be demonstrated in a silicon/glass-based microdevice previously developed in our group for the simultaneous DNA amplification (PCR) and electrochemistry-based (EC) sequence-specific detection (Lee et al., 2003). This PCR-EC microdevice has an integrated platinum-based thin film heaters and sensors patterned on the silicon surface and the working and reference electrodes are coated on the glass side of the device for the electrochemical detection. Unlike the previous experiments that utilized on-chip Pt heaters and sensors to control the real time temperatures of the PCR thermal cycles, the present investigation uses the chip heater and sensors for the cell lysis and genomic DNA denaturation (∼85 ◦ C) and the hybridization reaction between a specific region of the genomic DNA and its complementary probe linked to the MPs (∼50 ◦ C). As shown in Fig. 1, the chip temperature closely resembles to the desired temperature set in the LabVIEW. The successful collection of E. coli genome from the diluted samples of cell culture medium (O.D.600nm = 0.678) by probe modified MPs under the thermal manipulation of silicon chip is shown in Fig. 6. The image of gel electrophoresis reveals a clear band at 300 bps of PCR products amplified from the extracted genome. From this result, the presence of other cellular components such as enzymes, proteins and RNA did not inhibit the PCR process, which can be possibly explained by the deactivation or degradation of those biomolecules during the high temperature cell lysis step and the flush out of those inhibiting components during the washing step. The capturing efficiency of on-chip genome extraction is examined by the amplified PCR products after the extraction process. The chip-based extraction (∼75% of the positive control) efficiency is comparable to the extraction done in the Eppendorf® tube (figure not shown). The ability of MPs to collect genomic DNA via liquid phase hybridization and to be retained by an external magnetic field for washing presented suggests its potential applications in mi-
Fig. 6. Gel electrophoresis image of the PCR products showing on-chip genomic DNA extraction from crude E. coli cells by MPs in the microfabricated silicon chip and in Eppendorf tube. Lane 1: Eppendorf tube-based extraction, Lane 2: silicon chip-based extraction.
crochip environment as an active medium in sample preparation. 3.4. Electrochemical detection of PCR amplicons from extracted genome In addition to the gel electrophoresis analysis, a silver and gold-based electrochemical detection approach (Li et al., 2004; Cai et al., 2004; Lee et al., in press) was used to exanimate the compatibility of MPs to the EC-detection system. The chronopotentiometric stripping analysis as shown in Fig. 7a reveals a much higher electrochemical signal for the PCR amplicons than that of the probe only surface, indirectly confirming the success of capturing E. coli genome using MPs and the presence of probe-modified MPs does not contribute significantly to the background noise during ECdetection. The electrochemistry-based detection assays were repeated on a number of E. coli culture mediums at different dilution conditions. As shown by the PSA data of Fig. 7b, silver dissolution signal increases as a function of increasing number of the E. coli cells in the sample solution. Reproducible detection signals were obtained from the samples in the range of 2700 ∼ 2.7 × 106 cell/mL (stepwise diluted from E. coli culture medium: O.D.600nm = 0.678; 1.0 O.D.600nm = 1.0 × 108 E. coli cell/mL), suggesting that genomic DNAs of E. coli can be successfully extracted, amplified and electrochemically detected when the number of the bacteria cell is in the range of 102 ∼ 103 .
996
S.W. Yeung, I.-M. Hsing / Biosensors and Bioelectronics 21 (2006) 989–997
compatible and simple platform for the miniaturization of sample preparation step in micro-system. In the future work, specificity of the MPs-based genomic DNA extraction, together with the on-chip PCR reaction and sequence-specific electrochemistry-based detection will be demonstrated on the same device for the on-spot monitoring of food and water quality.
Acknowledgements The work described in this study was substantially supported by a grant from the Research Grants Council of the Hong Kong Special Administrative Region, China (project no. 601504). The laboratory facilities provided by the Bioengineering Laboratory and Microelectronics Fabrication Facility (MFF) of HKUST are also acknowledged. The authors thank Dr. Cai Hong for conducting the electrochemical detection experiments.
References
Fig. 7. (a) Chronopotentiometric stripping responses of (A) the probe-only and (B) the probe-target surfaces, in 1.0 mM AgNO3 /1.0 M KNO3 solution at a constant current of 10 A (1 h for hybridization with PCR product amplified from extracted genome, 35 min for gold nanoparticles labeling, 20 s for silver electrodeposition at −0.18 V). (b) The calibration curve of the silver dissolution signal with different concentrations of E. coli cells for genome extraction using functionalized MPs (PCR amplification and electrochemical detection conditions are the same as (a)).
4. Conclusions A novel strategy to capture genomic DNA directly from bacterial cell culture by functionalized MPs through DNA–DNA hybridization in a silicon-based microchip has been reported. The silicon chip is fabricated with a heater and temperature sensors for thermal cell lysis and genomic DNA denaturation to provide an accessible DNA template for MPs to capture through the biotinylated DNA probes. After washing under an external magnetic field, the genomic DNA anchored on MPs are ready for PCR and electrochemical detection by gold labeling and silver deposition/stripping. This on-chip genomic DNA extraction technique offers a
Breadmore, M.C., Wolfe, K.A., Arcibal, I.G., Leung, W.K., Dickson, D., Giordano, B.C., Power, M.E., Ferrance, J.P., Feldman, S.H., Norris, P.M., Landers, J.P., 2003. Microchip-based purification of DNA from biological samples. Anal. Chem. 75, 1880–1886. Bucak, S., Jones, D.A., Laibinis, P.E., Hatton, T.A., 2003. Protein separations using colloidal magnetic nanoparticles. Biotechnol. Prog. 19, 477–484. Cai, H., Shang, C., Hsing, I.M., 2004. Sequence-specific electrochemical recognition of multiple species using nanoparticle labels. Anal. Chim. Acta 523, 61–68. Carlo, D.D., Lee, L.P., 2002. Mechanical cell lysis results of a sample preparation module for functional genomics. In: Second Annual International IEEE-EMBS Special Topic Conference on Microtechnologies in Medicine and Biology, pp. 527–530. Carlo, D.D., Jeong, K.H., Lee, L.P., 2003. Reagentless mechanical cell lysis by nanoscale barbs in microchannels for sample preparation. Lab on a Chip 3, 287–291. Gao, J., Yin, X.F., Fang, Z.L., 2004. Integration of single cell injection, cell lysis, separation and detection of intracellular constituents on a microfluidic chip. Lab on a Chip 4, 47–52. Gu, H.W., Ho, P.L., Tsang, K.W.T., Wang, L., Xu, B., 2003. Using biofunctional magnetic nanoparticles to capture vancomycii-resistant enterococci and other gram-positive bacteria at ultralow concentration. J. Am. Chem. Soc. 125, 15702–15703. Gunasekera, N., Olson, K.J., Musier-Forsyth, K., Arriaga, E.A., 2004. Capillary electrophoretic separation of nuclei released from single cells. Anal. Chem. 76, 655–662. Han, F., Wang, Y., Sims, C.E., Bachman, M., Chang, R., Li, G.P., Allbritton, N.L., 2003. Fast electrical lysis of cells for capillary electrophoresis. Anal. Chem. 75, 3688–3696. Huang, M.F., Kuo, Y.C., Huang, C.C., Chang, H.T., 2004. Separation of long double-stranded DNA by nanoparticle-filled capillary electrophoresis. Anal. Chem. 76, 192–196. Hubbuch, J.J., Owen, R.T.T., 2002. High-gradient magnetic affinity separation of trypsin from porcine pancreatin. Biotechnol. Bioeng. 79, 301–313. Lagally, E.T., Simpson, P.C., Mathies, R.A., 2000. Monolithic integrated DNA amplification and capillary electrophoresis analysis system. Sens. Actuators B 63, 138–146.
S.W. Yeung, I.-M. Hsing / Biosensors and Bioelectronics 21 (2006) 989–997 Lee, T.M.H., Hsing, I.M., Lao, A.I.K., Carles, M.C., 2000. A miniaturized DNA amplifier: its application in traditional Chinese medicine. Anal. Chem. 72, 4242–4274. Lee, T.M.H., Carles, M.C., Hsing, I.M., 2003. Microfabricated PCRelectrochemical device for simultaneous DNA amplification and detection. Lab on a Chip 3, 100–105. Lee, T.M.H., Cai, H., Hsing, I.M., 2004. Gold nanoparticle-catalyzed silver electrodeposition on an indium tin oxide electrode and its application in DNA hybridization transduction. Electroanalysis 16, 1628–1631. Li, L.L., Cai, H., Lee, T.M.H., Barford, J., Hsing, I.M., 2004. Electrochemical detection of PCR amplicons using electroconductive polymer modified electrode and multiple nanoparticle labels. Electroanalysis 16, 81–87. Matteo, D.C., Igor, F., Fr´ed´eric, G.F., Frank, H., Ruth, F., 2003. DNA purification by triple-helix affinity precipitation. Biotechnol. Bioeng. 81, 535–545. Nicole, E.H., Brazill, S.A., 2003. Microchip capillary gel electrophoresis with electrochemical detection for the analysis of known SNPs. Lab on a Chip 4, 241–247. Northrup, M.A., Ching, M.T., White, R.M., Watson, R.T., 1993. DNA amplification with a microfabricated reaction chamber. In: Transducers, Seventh International Conference of Solid State Sensors and Actuators, vol. 34, pp. 924–926. Oster, J., Parker, J., Brassard, L., 2001. Polyvinyl–alcohol-based magnetic beads for rapid and efficient separation of specific or un-
997
specific nucleic acid sequences. J. Magn. Magn. Mater. 225, 145– 150. Poser, S., Schulz, T., Dillner, U., Baier, V., Kohler, J.M., Schimkat, D., Mayer, G., Siebert, A., 1997. Chip elements for fast thermocycling. Sens. Actuators A 62, 672–675. Prinz, C., Tegenfeldt, J.O., Austin, R.H., Cox, E.C., Sturm, J.C., 2002. Bacterial chromosome extraction and isolation. Lab on a Chip 2, 207–212. Schneegaß, I., Brautig¨am, R., K¨ohler, J.M., 2001. Miniaturized flowthrough PCR with different template types in a silicon chip thermocycler. Lab on a Chip 1, 42–49. Taylor, M.T., Belgrader, P., Furman, B.J., Pourahmadi, F., Kovacs, G.T.A., Northrup, M.A., 2001. Lysing bacterial spores by sonication through a flexible interface in a microfluidic system. Anal. Chem. 73, 492– 496. Wang, J., Polsky, R., Merkoci, A., Turner, K.L., 2003. Electroactive beads for ultrasensitive DNA detection. Langmuir 19, 989–991. Weizmann, Y., Patolsky, F., Katz, E., Willner, I., 2003. Amplified DNA sensing and immunosensing by the rotation of functional magnetic particles. J. Am. Chem. Soc. 125, 3452–3454. Yoza, B., Matsumoto, M., Matsunaga, T., 2003. DNA extraction using bacterial magnetic particles modified with hyperbranched polyamidoamine dendrimer. J. Biotechnol. 101, 219–228. Zhao, X., Tapec-Dytioco, R., Wang, K., Tan, W., 2003. Collection of trace amounts of DNA/mRNA molecules using genomagnetic nanocapturers. Anal. Chem. 75, 3476–3483.
Manipulation and extraction of genomic DNA from cell lysate by functionalized magnetic particles for lab on a chip applications Siu Wai Yeung, I-Ming Hsing ∗ Department of Chemical Engineering, Hong Kong University of Science and Technology, Clear Water Bay, Kowloon, Hong Kong Received 18 January 2005; received in revised form 9 March 2005; accepted 11 March 2005 Available online 6 April 2005
Abstract A novel approach for extracting living cells’ genomic DNA materials utilizing functionalized magnetic particles (MPs) is reported in this investigation. This strategy is amenable to handle bio-samples in a miniaturized environment and it offers a possibility to separate and purify DNA from other cell lysate mixtures “on-chip”, which is known to be a bottle-neck step in an integrated micro-total-analysis-system (TAS). “Species-specific” genomic DNA of interest is captured by the MPs based on the hybridization interaction between the biotinylated probes modified MPs and a complementary region of the targeted genome. The genome DNA anchored on the particles can be separated from the rest of cellular mixtures by a simple buffer washing upon the exertion of external magnetic force. Surface modifications of MPs and hybridization conditions affecting the genome capturing efficiency are investigated. Extraction of genomic DNA from E. coli is demonstrated in a silicon/glass-based micro-reactor patterned with a platinum heater and sensors. On-chip extraction and manipulation of genomic DNAs illustrated in this study is a step forward toward a total integrated bioanalytical microsystem for crude cells/sample analysis. © 2005 Elsevier B.V. All rights reserved. Keywords: Magnetic particle; DNA hybridization; Genome extraction; Microchip
1. Introduction Polymerase chain reaction (PCR), DNA sequencing techniques, together with fluorescence-based or electrochemicalbased detection methods, have been developed to analyze genetic materials (e.g. DNAs). However, to directly access the genomic materials contained within the cell envelope, tedious procedures of cell lysis, DNA extraction and purification are often performed. Most of the commercial kits for genome extraction (e.g. the silica membranes or conventional solvent processes) involve laborious routine steps of pipetting, vortexing and centrifugation in order to remove cell debris and to retrieve highly purified genomic DNAs. These procedures are complex and not amenable for the portable, lab-on-a-chip applications such as environmental monitoring, point-of-care diagnostics, and prevention of bio-terrorism. In our view, to realize the advantages of a portable bioanalytical microsys∗
Corresponding author. Tel.: +852 23587131; fax: +852 23580054. E-mail address: [email protected] (I.-M. Hsing).
0956-5663/$ – see front matter © 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.bios.2005.03.008
tem, a DNA-based micro-total-analysis-system (TAS), also called lab on a chip, should have the ability to perform sample preparation, DNA amplification, and sequence-specific detection (either based on electrochemical or fluorescent reporters) on chip. Chip-based techniques for cell lysis (Han et al., 2003; Taylor et al., 2001; Carlo et al., 2003), PCR (Northrup et al., 1993; Poser et al., 1997; Lee et al., 2000; Schneegaß et al., 2001) and capillary electrophoresis (Lagally et al., 2000; Nicole and Brazill, 2003; Huang et al., 2004) were largely demonstrated. Among them, the on-chip sample preparation technique (e.g. separating genetic materials from crude cell samples) is known to be the most challenging task for an integrated lab-on-a-chip system. Although novel DNA extraction methods utilizing adsorptive silica beads (Breadmore et al., 2003) and dielectrophoretic forces (Prinz et al., 2002; Gao et al., 2004; Gunasekera et al., 2004) were reported, on-chip integration issues were largely not addressed. One of chip-amenable approaches would be the use of magnetic particles (MPs), which
990
S.W. Yeung, I.-M. Hsing / Biosensors and Bioelectronics 21 (2006) 989–997
were demonstrated for the protein separation (Bucak et al., 2003; Hubbuch and Owen, 2002), the bacterial cell accumulation (Gu et al., 2003), electrical detection of DNA (Wang et al., 2003; Weizmann et al., 2003) and DNA extraction from crude blood samples by polyvinyl–alcohol-based (Oster et al., 2001) and hyperbranched polyamidoamine dendrimer modified (Yoza et al., 2003) MPs. Most of the published methods relied on the electrostatic interaction of the genomic materials and functionalized surfaces of magnetic particles for the genome extraction. Their approaches did not provide “species-specific” capturing of genomic DNAs and interferences of extraction by other cellular components (e.g. RNA and protein) are hardly avoidable. The DNA hybridization process can be used for the sequence-specific recognition and in recent years this method has been applied for the extraction and purification of DNAs. A few studies were previously reported for the capturing of mRNA via polyA–polyT interaction (Carlo and Lee, 2002) and for the extraction of plasmid DNA (Matteo et al., 2003). Moreover, the collection of trace amounts of DNA/mRNA using MPs functionalized by molecular beacons was recently demonstrated (Zhao et al., 2003). Magnetic particles modified by species-specific probes opens a possibility to capture/extract specific genome/DNA on chip. The genomic DNAs anchored on the MPs can be easily separated from the rest of cellular components by a simple buffer washing when an external magnetic force is applied. In addition, the release of the captured DNA from MPs can be effectively controlled by the on-chip heater, instead of using running buffers at different ionic strengths. In our work, the direct extraction/capturing of genomic DNA from cultured E. coli samples is demonstrated in a silicon-based micro-reactor with functionalized MP capturers. Factors, such as probe concentration and density, the sequences of probes and the states of hybridization that may impact the genome capturing efficiency are investigated. Ag- and Au-based sequence-specific electrochemical detection of PCR amplicons (Li et al., 2004; Cai et al., 2004; Lee et al., 2004) using the genome template extracted by MPs is studied. Finally, the cell lysis and genome capturing using MPs are demonstrated on integrated silicon-based microchip where the on-chip heater and sensors are used to control the temperatures for the cell lysing and genome probe hybridization.
operon (acc. no.: X07850) with the following primers: ECfwd (5 -GAC-AAG-AAA-ATC-TCC-AAC-ATC-C-3 ) and, EC-rev (5 -ACA -ACA-CGT-TTA-GCC-TGA-CC-3 ). The sequences of DNA probes used to demonstrate the capturing of E. coli genome via hybridization were 5 end biotinylated forward primers (biotin-EC-fwd) and a biotinylated random nucleotide sequences (biotin-R12, 5 -biotin-NNNNNN-NNN-NNN-3 ). LB agar and LB broth were purchased from Amersham Bioscience. DNA purification kit was ordered from Promega (catalog no. A1120). Oligonucleotides bearing a pyrrole group at 5 end (pyrrole-reverse primer, pEC-rev) for electrochemical detection were ordered from APIBIO (Cedex, France). Streptavidin/colloidal gold (5 nm) and sodium nitrate were obtained from Sigma (St. Louis, MO). Silver nitrate was ordered from Fisher Chemicals. Lithium hyperchlorate was obtained from Fluka. Sodium chloride, trisodium citrate dehydrate, disodium hydrogen phosphate, pyrrole monomer and HEPES (N-[2-Hydroxyethyl] piperazine-N -[2-ethanesulfonic acid] were purchased from Sigma-Aldrich (Germany). All solutions were prepared with ultrapure water from a Millipore Milli-Q system. 2.2. Instrumentation PCR was carried out in an Eppendorf® Mastercycler. Ultraviolet (UV) transillumination of agarose gel electrophoresis was performed by a Gel Documentation System (Alpha Innotech Corp.). DNA concentrations were measured by an Eppendorf® BioPhotometer. The control system for chip-based genome extraction consisted of signal conditioning board (SC-2042-RTD, National Instruments), a data acquisition DAQ card (PCI-MIO-16-1, National Instruments), a power supply unit (HP6629A), and a GPIB card (HP82350A). They were connected together for the computational control by a LabVIEW program (National Instruments, Texas, USA). The electrochemical detection methods: cyclic voltammetry (CV), chronoamperometry (CA) and potentiometric stripping analysis (PSA) were performed with a 16-channel VMP Multichannel Potentiostats/Galvanostats (PerkinElmer Instruments, USA) controlled by the EC-Lab (V6.70) software (Bio-Logic-Science Instruments). 2.3. Chip fabrication
2. Experimental 2.1. Reagents The avidin-coated MPs used in the experiment were purchased from Spherotech Inc. (catalog no. VMS-30-20). They are paramagnetic magnetic particles with average diameter of 3.0 m. The binding capacity of these MPs is 0.103 nmol of biotin-FITC to 1 mg of particles. PCR primers and biotinylated DNA probes were purchased from Invitrogen. During PCR, a 300 bps product is amplified from the E. coli groE
The silicon chip consists of one reaction chamber (∼8 L in volume) having the specific dimensions of 5 mm × 5 mm × 300 m (Fig. 1, inset I) and a patterned platinum heater and temperature sensors. The dimensions of heater line and sensor line are L: 3.6 cm × W: 600 m × H: 0.1 m and L: 0.49 cm × W: 20 m × H: 0.1 m, respectively (Fig. 1, inset II). After the chip was cleaned with acetone and deionized water in a ultrasonic bath (5 min for each washing step), a closed chamber was formed by bonding the chip with an indium tin oxide (ITO) glass via UV glue (Type UV-69,
S.W. Yeung, I.-M. Hsing / Biosensors and Bioelectronics 21 (2006) 989–997
991
Fig. 1. Chip temperature during the DNA extraction process (cell lysis and genome denaturation at 85 ◦ C, hybridization between denatured genome and the MP linked probe at 50 ◦ C). Insets: (I) Front side of silicon chip showing a square reaction chamber. (II) Back side of silicon chip showing the fabricated heater and temperature sensors. The silicon chip and glass were bonded together with UV glue.
Summers Optical, UV illumination: 90 min, vacuum oven: 2 h at 60 ◦ C). The ITO glass was drilled with holes for sample input and output and cleaned with deionized water before binding with the silicon chip. 2.4. Procedure 2.4.1. Functionalized MPs preparation To prepare the avidin-coated MPs for genome extraction via DNA–DNA pairing, 10 L of avidin-coated MPs were first extracted from the stock solution into Eppendorf tube and washed with 0.5 × SSC buffer. After removing the supernatant via centrifugation and pipetting, 10 L of biotinylated probes (biotin-EC-fwd) at different concentrations were added to the residual MPs and the well mixed mixture was then overnight incubated to achieve various probe densities on the particle surface. Finally, the DNA probes functionalized MPs were washed with 0.5 × SSC to remove the excess DNA-oligos and stored in 10 L 0.5 × SSC buffer at room temperature. 2.4.2. Tube-based genome extraction Two MPs-based genome extraction strategies were studied: solid phase hybridization approach and liquid phase hybridization approach. To start with, E. coli genome purified with DNA purification kit was adopted as the target DNA to illustrate the two proposed extraction strategies. In solid phase hybridization approach, base-pairing takes place at the MPs surface, as shown in Fig. 2a. In brief, each Eppendorf tube containing 15 L of 0.5 × SSC buffer, 1.0 L of probe-modified MPs, and 4.0 L of purified E. coli genome (1.0 ng/L) was first heated at 94 ◦ C for 5 min to denature the genome, followed by the incubation at 60 ◦ C for 10 min
to allow the hydrogen bonds formation between the immobilized probes and its complementary sequence on the denatured genomes to achieve DNA capturing. MPs-based genome capturing via liquid phase hybridization, as shown in Fig. 2b, was achieved by first mixing the genome samples (4 ng) with 1 L biotinylated probes (biotinEC-fwd, 1 nM) and 15 L of 0.5 × SSC buffer. After thermal denaturation (94 ◦ C for 5 min) and hybridization (60 ◦ C for 10 min), MPs were introduced to the samples and incubated for another 10 min to link the biotinylated probe-carrying genomic DNA to the particle surface via biotin–avidin interaction. In each approach, the DNA-attached particles were then washed with 0.5 × SSC buffer and separated from the washing buffer by an external magnet at the bottom of Eppendorf tube. Finally, PCR reagents without DNA template were added to the MPs to amplify a selected region of the extracted genome after they were released from the MPs during the initial denaturation step in PCR. 2.4.3. PCR protocols PCR reagents and Taq polymerase were purchased from Bioline Ltd. (catalog no. BIO-21040). For E. coli PCR experiment, reaction mixture contained 5 L of 10× PCR buffer, 1.0 L of 10 mM dNTP, 1.0 L of 50 mM MgCl2 , 2.0 L of forward and reverse primers (10 M), 0.5 L of 5 units/L Taq polymerase and 34.5 L of autoclaved double-deionized water. In the positive control, 4 ng of E. coli genome purified from commercial kit was added. In the tested samples, in which genomic DNA-captured MPs presented, ddH2 O was used instead to make up 50 L for PCR experiment. The master mix was subjected to the following thermal cycling profile: initial denaturation at 94 ◦ C for 5 min, 30 cycles at
992
S.W. Yeung, I.-M. Hsing / Biosensors and Bioelectronics 21 (2006) 989–997
Fig. 2. (a) Schematic diagram showing genome extraction by solid phase hybridization approach. Step 1: probe immobilization on MPs; Step 2: thermal denaturation of genome and its hybridization on the probes modified-MPs. (b) Schematic diagram showing genome extraction by liquid phase hybridization approach. Step 1: thermal denaturation of genome and genome-biotinylated probes hybridization; Step 2: addition of MPs for genome capturing via biotin–avidin interaction.
94 ◦ C for 30 s, at 60 ◦ C for 30 s, at 72 ◦ C for 30 s, and a final extension at 72 ◦ C for 5 min. PCR products were loaded in a 2% agarose gel (USB Corp. catalog no. 75817) prepared with 1 × TAE buffer for gel electrophoresis at 100 V for 1 h. The gel was immersed in ethidium bromide solution and visualized by UV transillumination. Asymmetric PCR was performed to produce singlestranded amplicons for the hybridization step in electrochemical detection. During amplification, concentration of EC-rev was adjusted to 0.1 M and the biotin-EC-fwd used is 10 M. It was cycled 40 times with the same temperature profile mentioned above. 2.4.4. On-chip genome extraction To demonstrate the compatibility of MPs-based genomic DNA capturing in the micro-environment, liquid phase hybridization approach was adopted and the 10 L sample was prepared with 0.5 L biotin-fwd probes (1 nM), 2.0 L of purified E. coli genomic DNA (1 ng/L) or living E. coli cells in LB (cultured overnight at 37 ◦ C), 7.5 L 0.5 × SSC buffer, 0.5 L BSA (20 mg/mL), 0.5 L dimethylformamide (DMF). Then, 8 L of well mixed sample was injected (bubble free) into the reaction chamber via the drilled hole. The injection holes are then sealed with a plastic tape (3M 5419,
3M Corporation) to prevent evaporation during heating and an insulation tape. Here, DMF was added to lower the denaturation temperature to 85 ◦ C for power consumption reduction. Once the chip was electrically connected, the chamber was heated to 85 ◦ C for 5 min and then 50 ◦ C for 10 min under computational control. MPs (1.0 L) were then injected into the chip via the injection hole. The MPs can be well dispersed by the internal mixing induced by the random movement of an external magnet in the chip area. The magnet was then removed for the incubation process. After 10 min incubation, genome-anchored MPs were washed three times by pipetting 0.5 × SSC buffer in and out of the chamber to wash out the unlinked genome and other cell components in the chamber. During washing, an external magnet was held under the chip to keep the MPs retained in the chamber. In the final wash, magnet was removed so that MPs were transported out of the chip for PCR amplification in a thermocycler. 2.4.5. Electrochemical detection 2.4.5.1. Probe immobilization and DNA hybridization. Electrochemical reaction was carried out on a glass chip electrode patterned with four Indium Tin Oxide (ITO) working electrodes, one platinum counter electrode and reference electrode. After the electrode-patterned glass
S.W. Yeung, I.-M. Hsing / Biosensors and Bioelectronics 21 (2006) 989–997
993
was cleaned with water and electrically connected to the VMP system, 10 L of freshly prepared reaction mixture (60 mM pyrrole, 0.1 M LiClO4 , and 20 M pEC-rev) was added to the electrodes surface for cyclic voltammetry to co-electropolymerize the oligonucleotide-linked pyrrole (pEC-rev) and pyrrole monomer (potential sweeping between −0.5 and 0.65 V at a scan rate of 50 mV/s for three cycles). The chip was then cleaned with water and dried by compressed nitrogen gas before the loading of asymmetric PCR products (0.5 L) on any two of four working electrodes for 2 h-incubation. Finally, chip is washed with 2 × SSC buffer to remove excess, nonspecifically bound PCR products. 2.4.5.2. Gold nanoparticle labeling. Gold nanoparticles were bound to the PCR amplicons by exposing the electrode to 0.5 L streptavidin-gold solution (the stock was 10 times diluted with HEPES buffer: 0.05 M HEPES and 0.2 M NaCl) at room temperature for 35 min. Any unbound gold nanoparticle labels were removed by rinsing with phosphate-buffered nitrate solution (PBN, 0.3 M NaNO3 /10 mM sodium phosphate, pH 7.0). 2.4.5.3. Silver electro-deposition and electrochemical stripping. Two cyclic voltammograms were scanned simultaneously from +0.2 to −1.0 V (scan rate: 100 mV/s) in 1.0 M KNO3 solution containing 1.0 mM AgNO3 on the probetarget electrode and probe only electrode. Because of the catalytic effect of gold nanoparticle (Lee et al., 2004), an appropriate silver electrodeposition potential was chosen where metallic silver only deposited on the probe-target electrode surface. After electrodeposition of silver onto the gold nanoparticle surface under the chosen potential for 20 s, the amount of silver deposited was determined by measuring the oxidative silver dissolution response in a potentiometric stripping analysis (PSA) scan. All the measurements were conducted with an applied anodic current of 10 A in the same silver nitrate solution at room temperature. 3. Result and discussion To miniaturize the sample preparation step of bio-analysis on a micro-fabricated chip, most of the current fluidic control schemes involved in washing will increase the complexity of the chip design and fabrication. On the other hand, the compatibility of MPs and the proposed genomic DNA extraction strategy to PCR simplify the DNA retrieval procedure such that the captured DNA can be released directly to the PCR mixture by breaking the hydrogen bonds between the probe and the genome during the initial denaturation step of PCR. In this paper, the amount of resulting PCR amplicons produced using the released genomic DNA from the MPs as the initial template in PCR is used to measure the extraction/capturing efficiency of the genome using MPs.
Fig. 3. (a) Gel electrophoresis photograph showing the influence of probe densities on the MPs on the genome capturing performance based on the solid phase hybridization approach. Lane 1: negative control (sample without DNA template); Lane 2: probe density (nmol probe/mg MPs) = 0.1; Lane 3: probe density, 0.01; Lane 4: probe density, 0.001; Lane 5: probe density, 1e-4; Lane 6: probe density, 1e-5; Lane 7: probe density, 0; Lane 8, PCR positive control. (b) Bar chart showing the relative band intensities in gel electrophoresis with the lane numbers corresponding to (a) to compare the influence of probe densities on MPs on the genome capturing performance based on the solid phase hybridization approach. The band intensity of positive control equals to 100%.
3.1. Characterization of probe modified magnetic particle for genomic DNA extraction To capture genome at a solid surface utilizing the hybridization process between a specific region of the genome and the complementary immobilized probe, the coverage of the DNA probes (i.e. probe density) on MPs is a deciding factor because of the anticipated hindrance effect of the large size genome and the strong electrostatic repulsion between negatively charged genome molecules. As it can be seen from Fig. 3a, the relative band intensity data of the gel electrophoresis results show that the genome extraction efficiency increases with the decrease of the probe density at the
994
S.W. Yeung, I.-M. Hsing / Biosensors and Bioelectronics 21 (2006) 989–997
range of 1.0 nmol/mg MP ∼ 1.0 × 10−3 nmol/mg MP. High probe coverage on the MPs could inhibit the probes to access the specific complementary binding site of the genome. As the probe density/coverage decreases, the large genome and the immobilized probes are given opportunities to rearrange among themselves to establish the hydrogen binding. (Fig. 3a, Lanes 2–4). With a further decrease of probe density on MPs, non-specific adsorption of genome could occur (Fig. 3a, Lanes 5–7) without improving the capturing process. This finding suggests that a sufficient amount of DNA probes should be immobilized to build up a resistant layer against non-specific adsorption. The genome capturing process based on the hybridization interaction between the genome and immobilized probe on MPs (the so-called solid phase hybridization approach) clearly requires a further improvement, as revealed by Fig. 3b. The optimal capturing efficiency relative to the positive control stays at around 30%. The gigantic circular bacterial genome with a complicated entanglement structure after thermal denaturation could explain the difficulty of the capturing process relying on the hybridization between a huge genome and tiny MP linked probes. 3.2. Capture genome by liquid phase approach To address the steric effect caused by the large genome structure and to improve the performance of MPs-based genome extraction, the liquid phase hybridization approach is adopted where the biotinylated probes are first introduced and hybridized with genomic DNA before being immobilized on MPs. MPs were pretreated to build up an oligo blocking layer (0.01 nmol/mg MPs) on the surface. This oligo layer can serve as a blocking layer to reduce the non-specific DNA adsorption as evidenced by a control experiment where the fluorescent signal can be seen from the complementary probe modified MPs, but not on the non-complementary probe modified MPs (Figure not shown). The fluorescein-labeled target DNA contributes to the fluorescent image of the complementary probe MPs after the hybridization. The absence of a fluorescent signal from the non-complementary probe MPs indicates the elimination of non-specific adsorption of DNAs by the oligo blocking layer. The relative band intensities of PCR products after the gel electrophoresis (Fig. 4) using the solid and liquid phase hybridization approaches suggest that when the liquid phase hybridization approach is employed, a much higher genome capturing efficiency can be achieved (∼75% of the relative to the positive PCR control) as compared to the solid phase one. However, in the liquid phase hybridization step the excessive use of capturing probes might also lower the DNA extraction performance. As displayed in Fig. 5, the amount of the PCR products amplified from the extracted genome increases with the continuous decrease of the capturing probe concentration from 10 M to 1.0 nM. Although, a high probe
Fig. 4. Gel electrophoresis photograph showing the effect of solid and liquid phase hybridization methods and the use of different DNA probes. Lanes 1 and 2: sequence-specific probe; Lanes 3 and 4: degenerated probe; Lanes 1 and 3: solid phase hybridization; Lanes 2 and 4 liquid phase hybridization; Lane 5: positive control.
concentration should increase the chance for these probes to anneal on the specific site of genomic DNA, nonetheless these biotinylated probes will also compete with the biotin reporters bound with hybridized genomes for the avidin sites of the MPs and thus lead to the reduction of the capturing efficiency. To characterize the hybridization process in the MP-based DNA extraction approach, biotin-R12 are allowed to capture E. coli genome according to the base pairing reaction between the degenerated sequences and the corresponding complements at the genome so that the one genome may carry several biotins for the extraction. Based on the respective PCR results, the relative band intensity of the gel electrophoresis as shown in Fig. 4, the difference in genome capturing
Fig. 5. Gel electrophoresis photograph showing the effect of the concentrations of liquid phase probes on the genome capture efficiency during the liquid phase hybridization process. Lane 1: probe concentration, 10 M; Lane 2: probe concentration, 1 M; Lane 3: probe concentration, 1 nM.
S.W. Yeung, I.-M. Hsing / Biosensors and Bioelectronics 21 (2006) 989–997
995
performances between degenerated probes and sequencespecific probes is not as obvious as expected. Therefore, it is not conclusive to state that the randomized hybridization process using degenerated probes provides any noticeable improvement in the DNA extraction process. One possible explanation is that the hybridized genome with multiple biotin labels resulted from the random hybridization might take away more than one avidin binding sites, thus decreasing the available binding sites of the MPs. It is possible that large surface areas when smaller MPs are used could contribute more biotin binding sites and thus improve the extraction process. 3.3. On-chip genome extraction from biological sample The realization of a fully integrated DNA-based bioanalytical microsystem requires the various bioprocessing modules for the sample preparation, target amplification and product detection to be integrated on chip. In this study, on-chip genomic DNA extraction using MPs is to be demonstrated in a silicon/glass-based microdevice previously developed in our group for the simultaneous DNA amplification (PCR) and electrochemistry-based (EC) sequence-specific detection (Lee et al., 2003). This PCR-EC microdevice has an integrated platinum-based thin film heaters and sensors patterned on the silicon surface and the working and reference electrodes are coated on the glass side of the device for the electrochemical detection. Unlike the previous experiments that utilized on-chip Pt heaters and sensors to control the real time temperatures of the PCR thermal cycles, the present investigation uses the chip heater and sensors for the cell lysis and genomic DNA denaturation (∼85 ◦ C) and the hybridization reaction between a specific region of the genomic DNA and its complementary probe linked to the MPs (∼50 ◦ C). As shown in Fig. 1, the chip temperature closely resembles to the desired temperature set in the LabVIEW. The successful collection of E. coli genome from the diluted samples of cell culture medium (O.D.600nm = 0.678) by probe modified MPs under the thermal manipulation of silicon chip is shown in Fig. 6. The image of gel electrophoresis reveals a clear band at 300 bps of PCR products amplified from the extracted genome. From this result, the presence of other cellular components such as enzymes, proteins and RNA did not inhibit the PCR process, which can be possibly explained by the deactivation or degradation of those biomolecules during the high temperature cell lysis step and the flush out of those inhibiting components during the washing step. The capturing efficiency of on-chip genome extraction is examined by the amplified PCR products after the extraction process. The chip-based extraction (∼75% of the positive control) efficiency is comparable to the extraction done in the Eppendorf® tube (figure not shown). The ability of MPs to collect genomic DNA via liquid phase hybridization and to be retained by an external magnetic field for washing presented suggests its potential applications in mi-
Fig. 6. Gel electrophoresis image of the PCR products showing on-chip genomic DNA extraction from crude E. coli cells by MPs in the microfabricated silicon chip and in Eppendorf tube. Lane 1: Eppendorf tube-based extraction, Lane 2: silicon chip-based extraction.
crochip environment as an active medium in sample preparation. 3.4. Electrochemical detection of PCR amplicons from extracted genome In addition to the gel electrophoresis analysis, a silver and gold-based electrochemical detection approach (Li et al., 2004; Cai et al., 2004; Lee et al., in press) was used to exanimate the compatibility of MPs to the EC-detection system. The chronopotentiometric stripping analysis as shown in Fig. 7a reveals a much higher electrochemical signal for the PCR amplicons than that of the probe only surface, indirectly confirming the success of capturing E. coli genome using MPs and the presence of probe-modified MPs does not contribute significantly to the background noise during ECdetection. The electrochemistry-based detection assays were repeated on a number of E. coli culture mediums at different dilution conditions. As shown by the PSA data of Fig. 7b, silver dissolution signal increases as a function of increasing number of the E. coli cells in the sample solution. Reproducible detection signals were obtained from the samples in the range of 2700 ∼ 2.7 × 106 cell/mL (stepwise diluted from E. coli culture medium: O.D.600nm = 0.678; 1.0 O.D.600nm = 1.0 × 108 E. coli cell/mL), suggesting that genomic DNAs of E. coli can be successfully extracted, amplified and electrochemically detected when the number of the bacteria cell is in the range of 102 ∼ 103 .
996
S.W. Yeung, I.-M. Hsing / Biosensors and Bioelectronics 21 (2006) 989–997
compatible and simple platform for the miniaturization of sample preparation step in micro-system. In the future work, specificity of the MPs-based genomic DNA extraction, together with the on-chip PCR reaction and sequence-specific electrochemistry-based detection will be demonstrated on the same device for the on-spot monitoring of food and water quality.
Acknowledgements The work described in this study was substantially supported by a grant from the Research Grants Council of the Hong Kong Special Administrative Region, China (project no. 601504). The laboratory facilities provided by the Bioengineering Laboratory and Microelectronics Fabrication Facility (MFF) of HKUST are also acknowledged. The authors thank Dr. Cai Hong for conducting the electrochemical detection experiments.
References
Fig. 7. (a) Chronopotentiometric stripping responses of (A) the probe-only and (B) the probe-target surfaces, in 1.0 mM AgNO3 /1.0 M KNO3 solution at a constant current of 10 A (1 h for hybridization with PCR product amplified from extracted genome, 35 min for gold nanoparticles labeling, 20 s for silver electrodeposition at −0.18 V). (b) The calibration curve of the silver dissolution signal with different concentrations of E. coli cells for genome extraction using functionalized MPs (PCR amplification and electrochemical detection conditions are the same as (a)).
4. Conclusions A novel strategy to capture genomic DNA directly from bacterial cell culture by functionalized MPs through DNA–DNA hybridization in a silicon-based microchip has been reported. The silicon chip is fabricated with a heater and temperature sensors for thermal cell lysis and genomic DNA denaturation to provide an accessible DNA template for MPs to capture through the biotinylated DNA probes. After washing under an external magnetic field, the genomic DNA anchored on MPs are ready for PCR and electrochemical detection by gold labeling and silver deposition/stripping. This on-chip genomic DNA extraction technique offers a
Breadmore, M.C., Wolfe, K.A., Arcibal, I.G., Leung, W.K., Dickson, D., Giordano, B.C., Power, M.E., Ferrance, J.P., Feldman, S.H., Norris, P.M., Landers, J.P., 2003. Microchip-based purification of DNA from biological samples. Anal. Chem. 75, 1880–1886. Bucak, S., Jones, D.A., Laibinis, P.E., Hatton, T.A., 2003. Protein separations using colloidal magnetic nanoparticles. Biotechnol. Prog. 19, 477–484. Cai, H., Shang, C., Hsing, I.M., 2004. Sequence-specific electrochemical recognition of multiple species using nanoparticle labels. Anal. Chim. Acta 523, 61–68. Carlo, D.D., Lee, L.P., 2002. Mechanical cell lysis results of a sample preparation module for functional genomics. In: Second Annual International IEEE-EMBS Special Topic Conference on Microtechnologies in Medicine and Biology, pp. 527–530. Carlo, D.D., Jeong, K.H., Lee, L.P., 2003. Reagentless mechanical cell lysis by nanoscale barbs in microchannels for sample preparation. Lab on a Chip 3, 287–291. Gao, J., Yin, X.F., Fang, Z.L., 2004. Integration of single cell injection, cell lysis, separation and detection of intracellular constituents on a microfluidic chip. Lab on a Chip 4, 47–52. Gu, H.W., Ho, P.L., Tsang, K.W.T., Wang, L., Xu, B., 2003. Using biofunctional magnetic nanoparticles to capture vancomycii-resistant enterococci and other gram-positive bacteria at ultralow concentration. J. Am. Chem. Soc. 125, 15702–15703. Gunasekera, N., Olson, K.J., Musier-Forsyth, K., Arriaga, E.A., 2004. Capillary electrophoretic separation of nuclei released from single cells. Anal. Chem. 76, 655–662. Han, F., Wang, Y., Sims, C.E., Bachman, M., Chang, R., Li, G.P., Allbritton, N.L., 2003. Fast electrical lysis of cells for capillary electrophoresis. Anal. Chem. 75, 3688–3696. Huang, M.F., Kuo, Y.C., Huang, C.C., Chang, H.T., 2004. Separation of long double-stranded DNA by nanoparticle-filled capillary electrophoresis. Anal. Chem. 76, 192–196. Hubbuch, J.J., Owen, R.T.T., 2002. High-gradient magnetic affinity separation of trypsin from porcine pancreatin. Biotechnol. Bioeng. 79, 301–313. Lagally, E.T., Simpson, P.C., Mathies, R.A., 2000. Monolithic integrated DNA amplification and capillary electrophoresis analysis system. Sens. Actuators B 63, 138–146.
S.W. Yeung, I.-M. Hsing / Biosensors and Bioelectronics 21 (2006) 989–997 Lee, T.M.H., Hsing, I.M., Lao, A.I.K., Carles, M.C., 2000. A miniaturized DNA amplifier: its application in traditional Chinese medicine. Anal. Chem. 72, 4242–4274. Lee, T.M.H., Carles, M.C., Hsing, I.M., 2003. Microfabricated PCRelectrochemical device for simultaneous DNA amplification and detection. Lab on a Chip 3, 100–105. Lee, T.M.H., Cai, H., Hsing, I.M., 2004. Gold nanoparticle-catalyzed silver electrodeposition on an indium tin oxide electrode and its application in DNA hybridization transduction. Electroanalysis 16, 1628–1631. Li, L.L., Cai, H., Lee, T.M.H., Barford, J., Hsing, I.M., 2004. Electrochemical detection of PCR amplicons using electroconductive polymer modified electrode and multiple nanoparticle labels. Electroanalysis 16, 81–87. Matteo, D.C., Igor, F., Fr´ed´eric, G.F., Frank, H., Ruth, F., 2003. DNA purification by triple-helix affinity precipitation. Biotechnol. Bioeng. 81, 535–545. Nicole, E.H., Brazill, S.A., 2003. Microchip capillary gel electrophoresis with electrochemical detection for the analysis of known SNPs. Lab on a Chip 4, 241–247. Northrup, M.A., Ching, M.T., White, R.M., Watson, R.T., 1993. DNA amplification with a microfabricated reaction chamber. In: Transducers, Seventh International Conference of Solid State Sensors and Actuators, vol. 34, pp. 924–926. Oster, J., Parker, J., Brassard, L., 2001. Polyvinyl–alcohol-based magnetic beads for rapid and efficient separation of specific or un-
997
specific nucleic acid sequences. J. Magn. Magn. Mater. 225, 145– 150. Poser, S., Schulz, T., Dillner, U., Baier, V., Kohler, J.M., Schimkat, D., Mayer, G., Siebert, A., 1997. Chip elements for fast thermocycling. Sens. Actuators A 62, 672–675. Prinz, C., Tegenfeldt, J.O., Austin, R.H., Cox, E.C., Sturm, J.C., 2002. Bacterial chromosome extraction and isolation. Lab on a Chip 2, 207–212. Schneegaß, I., Brautig¨am, R., K¨ohler, J.M., 2001. Miniaturized flowthrough PCR with different template types in a silicon chip thermocycler. Lab on a Chip 1, 42–49. Taylor, M.T., Belgrader, P., Furman, B.J., Pourahmadi, F., Kovacs, G.T.A., Northrup, M.A., 2001. Lysing bacterial spores by sonication through a flexible interface in a microfluidic system. Anal. Chem. 73, 492– 496. Wang, J., Polsky, R., Merkoci, A., Turner, K.L., 2003. Electroactive beads for ultrasensitive DNA detection. Langmuir 19, 989–991. Weizmann, Y., Patolsky, F., Katz, E., Willner, I., 2003. Amplified DNA sensing and immunosensing by the rotation of functional magnetic particles. J. Am. Chem. Soc. 125, 3452–3454. Yoza, B., Matsumoto, M., Matsunaga, T., 2003. DNA extraction using bacterial magnetic particles modified with hyperbranched polyamidoamine dendrimer. J. Biotechnol. 101, 219–228. Zhao, X., Tapec-Dytioco, R., Wang, K., Tan, W., 2003. Collection of trace amounts of DNA/mRNA molecules using genomagnetic nanocapturers. Anal. Chem. 75, 3476–3483.