Biomedical Microdevices

, Volume 11, Issue 2, pp 339–350

Magnetic-bead-based microfluidic system for ribonucleic acid extraction and reverse transcription processes


  • Chien-Ju Liu
    • Department of Engineering ScienceNational Cheng Kung University
  • Kang-Yi Lien
    • Institute of Nanotechnology and Microsystems EngineeringNational Cheng Kung University
  • Ching-Yi Weng
    • The Institute of Basic Medical SciencesNational Cheng Kung University
  • Jyh-Wei Shin
    • Department of ParasitologyNational Cheng Kung University
  • Tsuey-Yu Chang
    • Department of ParasitologyNational Cheng Kung University
    • Department of Engineering ScienceNational Cheng Kung University
    • Institute of Nanotechnology and Microsystems EngineeringNational Cheng Kung University
    • Medical Electronics and Device Technology CenterIndustrial Technology Research Institute

DOI: 10.1007/s10544-008-9240-1

Cite this article as:
Liu, C., Lien, K., Weng, C. et al. Biomed Microdevices (2009) 11: 339. doi:10.1007/s10544-008-9240-1


This paper presents a new integrated microfluidic chip that automatically performs ribonucleic acid (RNA) extraction and reverse transcription (RT) processes. The microfluidic system consists of a microfluidic control module and a magnetic bio-separator. The microfluidic control module can perform pumping and mixing of small amount of fluids and subsequent purification and concentration of RNA samples by incorporating with the magnetic bio-separator consisting of 2-dimension twisted microcoils. Notably, the magnetic bio-separators are developed either to generate the required magnetic field to perform the separation of magnetic beads or to work as a micro-heater to control the temperature field for the following RT process. Experimental results show that the total RNA can be successfully purified and extracted by using magnetic beads and the subsequent RT processing of the RNA can be performed automatically. Total RNA is successfully extracted and purified from T98 cells utilizing the microfluidic system, which is comparable with the conventional methods. The whole automatic procedure of RNA sample extraction only takes 35 min, which is much faster than the conventional method (more than 2 h). As a whole, the developed microfluidic system may provide a powerful platform for rapid RNA extraction and RT processes for further biomedical applications.


RNA extractionReverse transcriptionMagnetic beadMicrofluidicsMagnetic separatorμ-TASMEMS



application specific integrated circuit


complementary deoxyribonucleic acid


diethyl pyrocarbonate


deoxyribonucleic acid

DNase I

deoxyribonuclease I


deoxyribonucleotide triphosphate


electromagnetic valve


ethidium bromide


fetal bovine serum


glycerol-dehyde-3-phosphate dehydrogenase




messenger ribonucleic acid


phosphate buffered saline


polymerase chain reaction




ribonucleic acid



RNase H

ribonuclease H


reverse transcription


reverse transcription polymerase chain reaction


ribosomal ribonucleic acid




scanning electron microscope


transfer ribonucleic acid

1 Introduction

The process of purification and enrichment of biological samples such as DNA/RNA extraction are crucial for the detection and analysis in the biomedical and chemical applications. Extraction of DNA, RNA, and proteins from biological samples is a routine procedure in molecular biological laboratories for genome, transcriptome, and proteome analysis. It is necessary to pre-treat the clinical bio-samples which contain some substances that may interfere with the accuracy and compatibility of the biological assays. Recently, extraction of total RNA from the target cells have made a significant impact since these RNA samples can be significantly amplified to allow the production of probes for gene expression arrays (Madabusi et al. 2006; Shi and Bressan 2006). Briefly, it is the key process in the bio-sample pre-treatment process, which involves purification and enrichment of the RNA samples to perform the subsequent biomedical and biological analysis (such as Northern blotting, nuclease protection assays or reverse transcription) (Chomczynski and Sacchi 1987). However, it is also a labor-intensive and time-consuming process, involving various large-scale equipment (such as high-speed centrifuges, water baths and shakers) and making the RNA extraction procedure a lengthy and costly process. Furthermore, there is a risk of contamination during the RNA extraction process, which usually requires special caution. In addition, the operators may be under the risk of injury during the mixing of the organic solvents in the extraction process. Therefore, rapid and efficient extraction methods of total RNA from cells, suitable for further applications, are of great need.

Recently, various different methods for isolation and extraction of total RNA from eukaryotic tissues and cells have been reported (Bird 2005). For example, Chirgwin et al. reported the standard RNA extraction method from the target cells and tissues utilizing guanidinium thiocyanate, which was able to efficiently denature the endogenous ribonucleases (Chirgwin et al. 1979). Two standard methods for RNA preparation were investigated by using the guanidine, namely, (1) a single-step isolation method employing liquid-phase separation to selectively extract total RNA from tissues and cultured cells, and (2) the gradient precipitation of bio-samples utilizing a CsCl solution to isolate total RNA from the target cells (Chomczynski 1993; Millican and Bird 1998). There are also several surface immobilization techniques for DNA/RNA extraction by using surface modified probes or functional groups onto fixed spots or areas within the reaction reservoir. However, the tedious washing steps and labor-intensive processes are always required for the DNA/RNA purification. Besides, another complicate and time-consuming elution process is needed to collect and extract the immobilized RNA samples from the surface of the modified substrate. Therefore, using magnetic-bead-based extraction for RNA extraction has been extensively employed in recent years (Beaulieux et al. 1997; Haukanes and Kvam 1993; Jost et al. 2007; Melville et al. 1975; Roath et al. 1990). With this approach, surface-to-volume ratio of the magnetic beads can be significantly increased such that RNA samples immobilized on the beads can be greatly enhanced. Besides, the extraction of RNA samples can be easily carried out while applying an external magnetic field. Typically, large-scale electromagnets or permanent magnets are used to collect the magnetic beads so that the bound RNA can be purified and separated. For example, Muir et al. reported that enterovirus RNA can be efficiently and rapidly extracted from cerebrospinal fluid, stool, saliva, blood, pericardial fluid, urine, and cryopreserved or formalin-fixed solid tissue (Muir et al. 1993). With fewer lab-scale equipment and experimental procedures, the method of RNA extraction by using magnetic particles has the effect of extraction or purification on the target bio-sample, which can extremely improve sensitivity and specificity for biological analysis. The principle behind the magnetic-bead-based RNA sample extraction process is to use magnetic beads with switchable charges, or DNA probe-bound magnetic beads, to bind with the target RNA sample, followed by applying a magnetic field to collect the magnetic beads bound with the target RNA sample from the crude clinical samples (Lawson et al. 2006; Ogiue-Ikeda et al. 2003). There are many advantages in the magnetic-bead-based method for RNA extraction including the reduction of handling of hazardous reagents and reducing the centrifugation process time. RNA is a particularly labile molecule in organisms and would be easily digested by the nucleases during the RNA extraction process. This includes the use of diethyl pyrocarbonate (DEPC)-treated water, clean gloves, barrier pipette tips, working on ice and careful attention to minimizing the amount of time for each step (Chomczynski and Sacchi 2006; Pilcher et al. 2007). However, the whole protocol procedures are still time-consuming and labor-intensive, and the RNA sample could be consumed and digested by the surrounding ribonuclease (RNase) during the manual process.

In order to tackle these problems, various magnetic-bead-based sample purification and enrichment processes have been reported (Choi et al. 2000; Gijs 2004; Huang et al. 2002). With the incorporation of micro-electro-mechanical-systems (MEMS) technology and microfluidic techniques, these micro bio-devices and bio-systems provide powerful techniques to transport or mix the fluids and to perform sample pretreatment rapidly. In recent years, microfluidic systems utilizing MEMS techniques have demonstrated their potential for automatic biomedical diagnosis. A variety of functional bio-devices for sample preparation, reaction, separation, and manipulation have been successfully demonstrated (Huang et al. 2002; Vilkner et al. 2004). The reliability and functionality of these micro bio-devices can be greatly improved by integrating them with other functional microfluidic modules. For example, Hui et al. reported a microfluidic system built with a submicron filter for the extraction of viral RNA directly from viruses in the blood (Hui et al. 2007; Yobas et al. 2004). Lien et al. have also demonstrated rapid RNA extraction of dengue virus utilizing the magnetic-bead-based miniature microfluidic system (Lien et al. 2007a, b). By incorporating the function of bio-separation based on a bead-based approach in a microfluidic system, automatic sample pre-treatment and analytic tasks can be achieved in a shorter period of time. More importantly, the magnetic beads can be separated and manipulated by using meander-type inductors (microcoils) inside microchannels such that the localized magnetic field required for bio-separation can be generated. Therefore, using MEMS technology allows a compact system to perform magnetic bead manipulation and purification. The magnetic bio-separator to collect the magnetic beads with surface treatment to bind protein/DNA/RNA/bio-molecules suspended in solution has been realized (Choi et al. 2002; Ramadan et al. 2004). However, the temperature effect caused by Joule heating while a current passes through the microcoils can not be ignored. It should be taken into account when the planar microcoils were designed. Subsequently, after purification and extraction of the RNA samples, the RT process is a well-developed technique for transcribing single-stranded RNA into single-stranded DNA by using DNA polymerase enzymes such that the bio-samples can be stored stably for subsequent applications (Temin and Mizutani 1970). However, there still remain some crucial off-chip processes for sample preparation to be carried out manually before the RT process.

In order to improve this process, the study therefore proposes a new microfluidic system for total RNA extraction and reverse transcription process by integrating magnetic-bead-based RNA sample extraction devices into a single chip platform to automate the purification and RT-PCR process with less human intervention. By incorporating the microfluidic system, the micro bio-devices provide powerful techniques to transport or mix a small amount of fluids in the microchannels and to make the sample pretreatment analysis a quick and automatic process. Two major modules were integrated, namely a microfluidic control module and a magnetic bio-separator. Taking advantage of magnetic-bead-based technologies that provide a switchable surface charge dependent on the pH value of the surrounding fluid buffer, the total RNA (i.e. rRNA, tRNA and mRNA) of T98 cells (human Glioblastoma) were bound onto the surface of the beads and further purified by a magnetic field generated by the magnetic bio-separator. After the purification and enrichment processes, the purified RNA were then reverse transcripted into complementary deoxyribonucleic acid (cDNA) utilizing the same on-chip device, such that the resulting stable cDNA can be utilized in future biological applications. In addition, the following PCR process utilizing the primers specific for the human β-actin gene was performed by an external commercial PCR machine for comparison, and to demonstrate the successful extraction of RNA samples and reverse transcription of RNA to cDNA samples.

2 Materials and methods

2.1 Experimental procedure

The main purpose of automation for the developed microfluidic system is that the whole process including the RNA extraction and RT can be completed in an automatic fashion without any manual operation such that the extracted RNA can be stably extracted and converted to more stable cDNA for storage. Hence, the high-quality RNA can be extracted and enriched in a shorter period of time utilizing the automatic microfluidic system while compared with the manual operation that may cause the RNA samples to digest by the surrounding RNases. Instead of using traditional complex procedures in the extraction of RNA from the cell samples and excluding the unstable substances in the biological medium which can easily inhibit the purified RNA, the new microfluidic system tackles this problem by using charge switchable magnetic beads which carry positive charges in the solution buffer when the pH value lower is than 6.0. The experimental process using the integrated microfluidic system is schematically illustrated in Fig. 1. At first, the T98 cells were cultured and lysed by the lysis buffer manually in the Eppendorf tube, followed by loading the cells and magnetic beads (ChargeSwitch® Total RNA Cell Kits, Invitrogen™, USA) into a cell loading chamber (Fig. 1(a)). By adjusting the pH value of the buffer solution utilizing the binding buffer and mixing the cells with the magnetic beads, the DNA and total RNA of the T98 cells were trapped onto the surface of the magnetic beads in the solution buffer with a low pH value (Fig. 1(b)). Proteins and other contaminants were not bound onto the beads and can be simply washed away by using micropumps. Biological protocols (i.e. RT-PCR) require the elimination of even small traces of DNA, which can be easily accomplished by treatment with DNase. Therefore, the deoxyribonuclease I (DNase I) buffer was injected into the chamber to digest the trapped DNA and to trap the magnetic beads with purified total RNA (Fig. 1(c)). In order to wash out all the other interferent substances from the biological fluid, the magnetic bio-separators consisting of microcoils were supplied into a DC current to attract the magnetic beads onto the surface of the microchip while the washing buffer still flew through the reaction chamber (Fig. 1(d)). The washing process can be completed in 5 min and the RNA of the T98 cells was purified and concentrated by the on-chip magnetic bio-separators. Then, the low-salt elution buffer (10 mM Tris–HCl, pH 9.0, 1 mM EDTA in RNase-free water) was mixed with the RNA-bound magnetic beads to obtain the high quality total RNA (Fig. 1(e)). The yield of purified total RNA can be analyzed by checking the UV absorbance at 260 nm (NanoDrop® spectrophotometer ND-1000, USA). In order to obtain the stable samples, the RNA was then reverse-transcripted into cDNA by using on-chip micro-heaters to perform the subsequent RT process (Fig. 1(f)). Notably, the microheaters were basically the micro-coils supplied with a higher DC current. Housekeeping genes such as glycerol-dehyde-3-phosphate dehydrogenase (GADPH) and β-actin are widely used as internal controls. Because these two proteins are essential for the maintenance of cell function, it is generally assumed that they are constitutively expressed at similar levels in all cell types and tissues. Therefore, a PCR process was subsequently performed using primers specific for the human β-actin gene. Human β-actin gene can be used to assess successful RNA extraction (Ponte et al. 1984; Suzuki et al. 2000), to ascertain RNA integrity, and to normalize PCR results for the amount of RNA assayed.
Fig. 1

A schematic illustration of the operation process of the microfluidic chip capable of performing RNA extraction and reverse transcription processes: (a) loading the lysed cells and beads samples, (b) incubation process for total RNA binding with magnetic beads, (c) digestion of DNA by DNase I buffer, (d) washing process for purifying the RNA-bound beads, (e) elution of purified total RNA, (f) RT process for cDNA synthesis

2.2 Chip design

The main contribution of the current study is to develop a powerful platform for rapid RNA extraction and RT process by utilizing the magnetic particles technology in the microfluidic system. The whole process can be completed in an automatic fashion without any manual operation and can enhance the quality of the extracted RNA. Besides, stable cDNA can be synthesized using the same microfluidic system. Figure 2 shows a schematic illustration of the microfluidic system capable of performing RNA extraction and reverse transcription processes. The integrated microfluidic system comprises two major components, one is the microfluidic control module and the other is the magnetic bio-separator for magnetic field generation and temperature control. The microfluidic control module consists of a two-way pneumatic micropump with twisted connected air channels, three microvalves, multiple microchannels and reaction chambers. Figure 3 shows the design diagram of the new two-way pneumatic micropump and the experimental setup of the microfluidic system. The two-way pneumatic micropump was modified from the pneumatically driven peristaltic micropumps utilizing serpentine-shape (s-shape) channels (Wang and Lee 2005, 2006). The pumping rate might be influenced by the time interval between the deflections of adjacent membranes, and is therefore affected by the geometric characteristics of the serpentine microchannel. The connected air channels are designed for compressed air traveling along the air chambers formed from PDMS membranes that can generate the a peristaltic effect which drives the fluid along the microfluidic channel. Figure 3(a) shows the schematic illustration of the microfluidic control module. The proposed pneumatic micropump with four twisted connected air channels is developed to effectively minimize the design area of the micropump zone and also reduce the dead volume between the deflected PDMS membranes during the pumping process. It also can increase the pumping rate for fluids delivery in the microfluidic channel. Hence, the two-way pneumatic micropump capable of moving fluids forwards and backwards was developed for the mixing, washing, and transportation processes. Figure 3(b) shows the experimental setup of the microfluidic system. The microfluids in the microchannels could be driven forwards or backwards by utilizing a single electromagnetic valve (EMV, SMC Inc., S070M-5BG-32, Japan). The time-phased deflection of successive polydimethylsiloxane (PDMS, Sylgard 184A/B, Dow Corning Corp., USA) membranes along the microchannel length generates a peristaltic effect which drives the fluid along the microfluidic channel. The location of the injected compressed air determines the fluid direction such that the fluid can be transported forwards or backwards. Similarly, the PDMS membranes can be used as microvalves if the PDMS membrane is deflected completely to block the fluid flow in the microchannels. A microcontroller is used to control the microfluidic module by providing digital signals to regulate the EMV that causes the thin PDMS membranes to deflect downwards and upwards pneumatically under the applied compressed air pressure. As a result, the mixing and transporting processes can be completed automatically utilizing the microfluidic control module.
Fig. 2

A schematic illustration of the magnetic-bead-based microfluidic system for RNA extraction and reverse transcription processes
Fig. 3

(a) Design diagram of the two-way pneumatic micropump with twisted connected air channels for fluids delivery and mixing in the microfluidic channel. There are four twisted air channels for compressed air traveling between the air chambers. (b) Experimental setup of the microfluidic control module. A three-position EMV is used to control the direction of the microfluids in the microchannel by switching the injected location of the compressed air into the pneumatic micropump for

The magnetic bio-separator is made of a set of copper microcoils, which can be controlled by an application specific integrated circuit (ASIC) micro-controller. The magnetic bio-separator is used to trap the magnetic beads bound with RNA inside the cell/magnetic bead loading chambers, followed by washing and purifying the total RNA. After that, the purified RNA can be heated up utilizing the same microcoils that are used as the micro-heater to perform the RT process. A local magnetic field can attract the magnetic beads in the reaction chambers when a current is supplied to the microcoils. The ASIC micro-controller can be used to maintain a constant current in the microcoils of the magnetic bio-separator such that the required magnetic field or the temperature field inside the chamber can be controlled by adjusting the input power of the microcoils. A photograph of the assembled microfluidic chip is shown in Fig. 4(a). The dimensions of the chip are measured to be 43 mm × 51 mm. The scanning electron microscope (SEM) images of each device are shown in Fig. 4(b–c).
Fig. 4

(a) A photograph of the magnetic-bead-based microfluidic chip. The dimensions of the integrated chip are measured to be 51 mm × 43 mm. (b) SEM image of the SU-8 mold for the two-way, pneumatic micropump, and (c) SEM image of the magnetic bio-separator

2.3 Fabrication process

The microfluidic chip was fabricated using a MEMS fabrication process. The simplified fabrication processes are schematically illustrated in Fig. 5. There are two major steps involved in forming the integrated microfluidic chip, including the fabrication of magnetic bio-separators and the formation of the microfluidic control module. The fabrication of the microcoils used a standard lithography and electroplating processes (Chiou and Lee 2005). Copper (Cu) microcoils with a width of 50 μm and a thickness of 12 μm were electroplated. At first, a glass substrate was cleaned and then was deposited with a seed layer consisting of 30-nm chromium (Cr) and 100 nm Cu utilizing an electron-beam evaporation process (Fig. 5(a-1)). Then, a positive PR layer (AZ 4620, Clariant Corp., Switzerland) was spin-coated and was patterned to form the mold for the microcoils using a standard photolithography process (Fig. 5(a-2)). Subsequently, the Cu coils were electroplated and the seed layer was removed by using a Cr etchant solution (Cr-7T, CYANTEK Corp., USA) and 5 M diluted hydrochloric acid (Fig. 5(a-3)–(a-5)), followed by patterning of a 15 μm-thick SU-8 layer onto the surface of the microcoils as the dielectric layer to prevent electrical break-down (Fig. 5(a-6)). Figure 4(b) shows a scanning electron microscope (SEM) image of the microcoils for the magnetic bio-separators, which can be well defined using the developed fabrication process.
Fig. 5

Simplified fabrication process of the microfluidic system. There are two major steps including: (a) the fabrication process of the Cu magnetic bio-separators, followed by patterning a thin dielectric SU-8 layer and (b) the fabrication process for standard SU-8 lithography and PDMS casting processes for the microfluidic control module. Finally, these two modules are bonded together to complete the integrated microfluidic system

Figure 5(b) shows the fabrication process of the microfluidic control module. The microfluidic control module was fabricated using SU-8 photolithography and PDMS replication processes. Briefly, the negative thick PR SU-8_50 was first spin-coated and then patterned on a silicon substrate using a standard lithography process (Fig. 5(b-1)), followed by an exposure and developing process of the SU-8 mold (Fig. 5(b-2)–(b-3)). After the post-bake and hard bake process of the SU-8 mold, the PDMS was prepared by thoroughly mixing the PDMS prepolymer and curing agent (Sylgard 184A/B, Sil-More Industrial Ltd., USA) in a ratio of 10:1 by weight and spun onto the SU-8 mold (Fig. 5(b-4)). The polymer was then degassed under vacuum to remove any air bubble created during mixing. Then the mixture was poured onto the SU-8 masters and cured at 100°C for 4 h (Fig. 5(b-5)). The cured PDMS layers were then peeled off mechanically. Two PDMS layers were obtained from the replica molding and were then bonded by using an oxygen plasma treatment. Figure 4(c) shows a SEM image of the SU-8 mold for the two-way, pneumatic micropump. Finally, the integrated chip was assembled by bonding these modules together using an oxygen plasma treatment (Fig. 5(b-6)).

2.4 T98 cell preparation

Human glioma cell lines T98 were obtained from ATCC (accession no. CRL-1690; Manassas, VA, USA). Cells were grown in Dulbecco’s Modified Eagle’s Medium (DMEM, HyQ®, Hyclone™, US) supplemented with 1% antibiotics that included 10,000 U/ml penicillin G, 10,000 μg/ml streptomycin and 25 μg/ml fungizone (GIBCO®, Invitrogen™, USA), and 10% fetal bovine serum (FBS, Hyclone™, US) in 10-cm2 culture dishes (TPP®, Switzerland). Cells were then washed in 1× PBS and tested with an amount of 5 × 106 cells/ml in the developed microfluidic system. Three RNA extraction methods procedure were used for comparison and listed in Table 1, including a commercially available kit (TRIZOL® and ChargeSwitch®, Invitrogen™, USA), a manual magnetic-bead operation and an on-chip magnetic-bead operation. Note that extraction procedures using the commercially available kit were performed according to the manufacturer’s instructions. The cells preparation and lysis process were performed manually outside the system. At first, we flushed medium through the culture cells and washed them by 1X PBS. To remove the monolayer, we used rubber policeman and then transferred cells into an Eppendorf tube. Cells were washed again with 1× PBS and mixed with 500 μl of the prepared lysis buffer (ChargeSwitch®, Invitrogen™, USA), followed by pipetting the solution up and down thoroughly until the cells were lysed completely. Then the cell samples were loading into the microfluidic chip for RNA extraction.
Table 1

The protocol for the three different RNA extraction processes


Trizol reagent

Beads (manual)

Beads (chip)


1 ml Trizol reagent per 6 × 106 cells. Store homogenate for 5 min (room temperature)

1 ml lysis mix per 6 × 106 cells. Pipette up and down thoroughly

1 ml lysis mix per 6 × 106 cells. Pipette up and down thoroughly


Add 0.2 ml chloroform, mix vigorously. Store sample 15 min (room temperature)

Incubate at 60°C for 15 min. Mix the lysate briefly by vortexing

Incubate at 60°C for 15 min. Mix the lysate briefly by vortexing


Centrifuge at 12,000 g for 15 min (4°C)

Add magnetic beads, DNase I and binding buffer. Store sample 10 min (RT)

Add magnetic beads, DNase I and binding buffer for mixing utilizing the micropump. Store sample 10 min (room temperature)


Transfer aqueous phase into a new tube. Add 0.5 ml isopropanol, mix, store for 10 min (room temperature)

Place the sample on the magnet. Aspirate and discard the supernatant

Wash the bead sample while the magnetic field was turned on by the on-chip bio-separators


Centrifuge at 12,000 g for 8 min. (4–25°C)

Add wash buffer and pipette up and down

Incubate at room temperature for 5 min


Mix RNA pellet with 1 ml 75% ethanol

Place the sample on the magnet. Aspirate and discard the supernatant

Collect RNA


Centrifuge at 7,500 g for 5 min (4–25°C)

Add eluting buffer



Air dry the RNA pellet for 5–10 min

Incubate at room temperature for 5 min



Dissolve by pipetting in 50–200 μl water

Collect RNA



Incubate at 55–60°C for 10 min


2.5 Reverse-transcription (RT) process

The reverse transcription process was carried out by using extracted RNA samples obtained from three methods. A SuperScript™ II reverse transcription kit (Invitrogen™, USA) was used and the procedures were performed according to the manufacturer’s instructions. One microliter of random hexamer primers (3 μg/μl, Invitrogen™, USA), 1 μl dNTP mixture and 1 μl of Oligo (dT)12–18 primer (0.5 μg/μl, Promega, USA) were added to 1 μg purified total RNA in 4 μl DEPC-treated ddH2O. The mixture was incubated at 65°C for 5 min. Then, 4 μl of 5× First-Strand buffer, 2 μl of 0.1 M dithiothreitol (DTT), and 1 μl of DEPC-treated ddH2O were added, and the mixture was incubated at 25°C for 2 min and then 42°C for 2 min. 1 μl of reverse transcriptase (SuperScript™ II, Invitrogen™, USA) was then added into the reaction chamber at 25°C for 10 min and then incubated at 42°C for 2 h for cDNA synthesis. All the reverse transcription processes were performed using a 20-μl reaction volume. Then, the cDNA samples were transferred into a block heater at 70°C for 15 min to terminate the reverse transcription reaction. Ribonuclease H (RNase H, Invitrogen™, USA) is commonly used to destroy the RNA template after first-strand complementary DNA synthesis. Therefore, one microliter of RNase H was added and incubated at 37°C for 20 min to remove RNA complementary from the synthesized cDNA. Finally, the synthesized cDNA were ready to be used as a template for subsequent amplification in a PCR process or other biomedical assays.

2.6 Polymerase chain reaction (PCR) process

PCR amplification is a well recognized method for the amplification of nucleic acids and can be used to demonstrate the successful RT process of the cDNA samples. All the PCR processes were performed in the external PCR machine (MyCycler Thermal Cycler, Bio-Rad Laboratories, Hercules, CA, USA) with a 25-μl reaction volume, which consisted of 2 μl of cDNA obtained from three different RNA extraction methods after the reverse transcription process, 2.5 μl of 10× reaction buffer (200 mM Tris–HCl (pH 8.8), 100 mM KCl, 20 mM MgSO4·7H2O, 1% Triton X-100, 100 mM (NH4)2SO4, 1 mg/ml BSA; Yeastern Biotech), 1.5 units of Hi-Fi DNA polymerase (Yeastern Biotech), and 0.2 mM each deoxynucleotide triphosphate (Yeastern Biotech). Ten picomols of primers β-actin-forward (5′-TGGAATCCTGTGGCATCCATGAAAC-3′) and β-actin-reverse (5′-TAAAACGCAGCTCAGTAACAGTCCG-3′) were added for each PCR process for amplification in the human β-actin gene (348-bp). PCR conditions are described as follows: an initial denaturation step at 94°C for 2 min, denaturation at 94°C for 30 s, annealing at 51°C for 30 s, and extension at 72°C for 1 min for a total of 30 cycles, followed by a final extension cycle at 72°C for 5 min. The PCR products were then separated and detected using ethidium bromide (EtBr)-stained agarose gel (2%) (BioDoc-It 2UV System, UVP™, Canada).

3 Results and discussion

3.1 Characterizations of the microlfuidic system

In transporting microfluids using the microfluidic control system, the experimental results reveal that the pumping rate can be regulated by increasing the operational frequency of the EMV and the pressure of the supplied compressed air. The time-phased deflection of the successive PDMS membranes driven by the pneumatic force was able to generate a peristaltic effect to deliver liquid in the microfluidic channel. The new pneumatic micropump designed with twisted connected air channels can effectively enhance the pumping rate for fluids delivery in the microfluidic channel while compared with the micro pump utilizing the s-shape air channels (Lien et al. 2008). The relation between the pumping rate and the driving frequency at three different air pressures are shown in Fig. 6. It can be seen that the pumping rate increases along with the increase in air pressure. The maximum value of the fluid pumping rate is 1,233 μl/min at a driving frequency of 19 Hz under 15 psi of air pressure. In addition, experimental observation also show that the fluid flow can be shut off by microvalves whose membranes are completely deflected at a pressure of 25 psi. Hence, the microfluidic module is capable of transporting the bio-samples and reagents, or blocking the fluids inside the reaction chambers. Moreover, the controllable flow direction in the microchannel can be used for liquid delivery in the microfluidic system and also generate the effect of shaking-like fluid mixing by the swift alteration of flow direction. When the two-way pneumatic micropump was driven at an operating condition with a 13 Hz driving frequency and 20 psi air pressure, significant mixing can be observed due to the disturbance transferred into the fluids. The mixing efficiency of the unmixed and mixed condition located downstream of the exit of the two-way, pneumatic micro mixer is 20% and 89%, respectively. The results indicate that the microfluidic control module can successfully transport and mix the samples simultaneously.
Fig. 6

The relation between the pumping rate and the driving frequency of the EMV at three different air pressures. It can be seen that the pumping rate reaches a maximum value of about 1233 μl/min at a driving frequency of 19 Hz under 15 psi of air pressure

The built-in magnetic bio-separator consisting of twisted microcoils was used for magnetic bead collection in the extraction process of RNA samples and the subsequent RT reaction. Figure 7 shows the magnetic and temperature field generated by the magnetic bio-separators. At a current of 40 mA, the microcoils can produce a magnetic field with a magnitude up to 24.3 Gauss and a temperature field of 37.5°C inside the reaction chambers. From the experimental observations, the results demonstrated that the magnetic beads can be captured by the magnetic field generated by the magnetic bio-separator with a washing flow rate of 20 μl/min and revealed that enough magnetic force can be generated to capture the beads after the washing process. In addition, the temperature field of 42°C and 70°C in the RT process can also be generated by the same device when currents of 44.0 mA and 72.5 mA were applied to the microcoils, respectively. Therefore, the following RT process can be performed in an automatic fashion by the magnetic bio-separator in the same chamber without losing any pure RNA samples.
Fig. 7

The magnetic and temperature fields generated by the magnetic bio-separator at different applied currents

3.2 RNA extraction and RT processes

In terms of handling the nucleic acids, when compared with DNA, RNA samples are more unstable and could be digested by the surrounding RNase during the extraction process. In the traditional method of handling RNA samples, several time-consuming and labor-intensive centrifuge steps are required. Therefore, a new microfluidic system was developed in this study for RNA extraction and reverse transcription processing. In comparison with traditional RNA extraction utilizing the Trizol method (TRIZOL® Reagent, Invitrogen™, USA; Chomczynski and Sacchi 1987), the microfluidic system only takes 35 min for purifying and extracting the RNA samples while the traditional method takes more than 2 h. In addition, the quality and purity of the nucleic acid has been investigated using absorbance spectrophotometry, whose signals at 260 nm allowing calculation of the concentration of RNA in the sample and at 280 nm indicate the protein concentration. An optical density of 1 corresponds to approximately 40 μg/ml of single-stranded RNA. The measurements at 260 and 280 nm for the RNA samples have been used to deduce the amount of nucleic acid and the accompanying levels of protein carryover in a sample. Pure preparations of RNA have an A260/280 ratio of between 1.8 and 2.0. A ratio of A260/280 ≦ 1.8 is usually used as an indication for protein contamination. Table 2 list the A260/280 ratio, the concentrations of extracted RNA samples, the amount of total RNA, the duration of the extraction process and the cost per sample by using three different methods, including the traditional Trizol method, the manual magnetic beads extraction method, and the magnetic beads extraction method in the microfluidic system. The absorbance ratios of the RNA samples are at or above 2.0, indicating the high purity of the extracted RNA samples, which are ready for the subsequent biological applications. Besides, the reasonable extracted concentrations of RNA samples also demonstrate that the proposed microfluidic system can provide a rapid and high-quality extraction method of RNA samples. From the comparable results, 47.21 μg of RNA samples from 5 × 106 T98 cells is extracted by the magnetic beads in the microfluidic system while the extracted total RNA from Trizol method is 68.24 μg. However, the Trizol extraction method, which uses hazardous organic reagents such as chloroform/phenol, may create toxic wastes from the bio-samples during the operation process. Besides, the traditional method usually has a higher cost compared with the extraction method utilizing the magnetic beads in the microfluidic system and is a labor-intensive and time-consuming process during the extraction process.
Table 2

A comparison of the three different methods for RNA extraction


Traditional Trizol method

Magnetic beads (manual)

Magnetic beads in the microfluidic system

Consuming time

2 h (RNA purification)

50 min (RNA purification)

35 min (RNA purification)

2 h and 5 min (RT)

2 h and 5 min (RT)

80 min (RT)





RNA concentration (ng/μl)




Total RNA (μg)




Organic reagent




Another investigation of extracted RNA quality involves analysis of the ribosomal 18S (1.9-kb) and 28S (5-kb) species by denaturing agarose electrophoresis to determine the extent of degradation of the RNA sample. Figure 8 shows the slab-gel electropherograms of the extracted RNA samples by using the three methods, namely the traditional Trizol extraction method (lane 1), the manual magnetic beads extraction method (lane 2) and the automatic magnetic bead extraction method in the microfluidic system (lane 3). Note that the “L” represents the 1-kb marker (Fermentas, Canda). The 28S and 18S eukaryotic ribosomal RNA should reveal nearly a ratio of 2:1 indicating that a high extraction yield of total RNA has been achieved. DNA contamination was evaluated by the visual detection of a high molecular weight band in the gel. Experimental data reveals that each extraction method yields good quality total RNA, with 28S to 18S rRNA ratios ranging from 1.11 to 1.45 (Fig. 8). The gel image illustrates three extraction methods have no significant residual DNA in the gel and only one band was observed above the 28S band by using magnetic beads (both the manual and on-chip approaches). From these comparable results, a high-quality extraction of total RNA has been performed and purified by three different extraction methods. Still, the extraction procedure by the traditional Trizol and the manual magnetic beads method takes 2 h and 50 min, respectively. The developed system only takes 35 min for extracting the RNA samples (Table 2). In addition, the time saving might be increased up to 70% reduction if multiple extractions have to be carried out. For example, the operation time of the manual magnetic-bead-based RNA extraction may be increased to 150 min when performing extraction of ten samples, while the automatic RNA extraction only takes 40 min by using the microfluidic system if ten multiple RNA extraction modules were built on a single chip. As well, the quality of the RNA is an important issue especially when using microarray technology because it is the material used to prepare the detection probe. Therefore, poorly processed or partially degraded RNA may not faithfully represent transcription (Hegde et al. 2000). Hence, the proposed system may provide a powerful tool for automatic rapid RNA sample extraction and further biomedical analysis.
Fig. 8

Non-denaturing gel electrophoresis of RNA extracted with the three methods. The 28S (5 kb) and 18S (1.9 kb) rRNA bands are indicated. Lane 1, RNA was extracted using the traditional Trizol method. Lane 2, RNA was isolated from the manual magnetic beads method and RNA from the automatic magnetic beads extraction method in the microfluidic system (lane 3). Lane L represents the 1-kb DNA ladder

In this study, RNA samples from the human β-actin gene extracted from T98 cells was analyzed in the following PCR process to confirm the successful RT process of the RNA samples into cDNA samples. Figure 9 shows the PCR analysis of the RNA samples extracted using the three different methods. The cDNA samples were synthesized from 1 μg of RNA from each of the three different extraction methods. In Fig. 9, RNA extracted from the automated magnetic beads method from 5 × 106 T98 cells was used as a template for PCR of the RNA from the human β-actin gene by using a traditional Trizol method to be reverse transcripted to cDNA and amplified in PCR (lane 1), the manual magnetic beads method (lane 2) and the automatic magnetic beads extraction method in the microfluidic system (lane 3). Lane L is 100-bp ladder (Fermentas, Canda.). From the results, total RNA can all be extracted and purified successfully by all three different extraction methods including the magnetic-bead-based RNA extraction method in the developed microfluidic system. In addition, all the extraction methods can yield high-quality total RNA and the total RNA samples can be reverse transcripted into cDNA successfully. Experimental data showed that RNA extracted from all three different methods is suitable for downstream applications such as RT-PCR and micro-array gene expression analysis.
Fig. 9

PCR analysis of RNA extracted with the three methods. The PCR products were analyzed on a 2% agarose/EtBr gel; the amplified gene is β-actin. The PCR product has a length of 348 bp. Lane L, 100-bp ladder. RNA was extracted from three different methods, as described in Fig. 8, using a traditional Trizol method (lane 1), the manual magnetic beads method (lane 2), or the automatic magnetic beads extraction method in the microfluidic system (lane 3)

There are various methods available for isolating total RNA from eukaryotic tissues and cells. However, the time-consuming and labor-intensive process always makes the extraction a lengthy and costly procedure. For instance, the recommended approach in the TRIZOL product information begins with the homogenization of the tissue or cell sample, followed by a phase separation from the TRIZOL reagent after the addition of chloroform. The RNA is in the upper, aqueous phase, while the DNA and proteins are in the lower, organic phenol–chloroform phase. The RNA is then precipitated by the addition of isopropanol and incubation of the mixture for 10 min. After centrifugation, the RNA pellet needs another wash and air-dry step. These procedures are time-consuming and relatively complicated. Alternatively, magnetic bead technology makes use of magnetic field to accomplish RNA purification in the single tube. This technology could decrease the loss of potential sample in liquid phase. However, the manual magnetic beads method still could have artificial interference resulting from handling instrument, sample and materials to impair its performance. A key feature of the microfluidic process is the ability to extract total RNA rapidly and automatically. This approach could have far-reaching implications for gene expression studies in developmental and cancer biology or microarray.

4 Conclusion

The present study demonstrates a microfluidic device capable of rapid RNA extraction and RT processes. It can yield high-quality RNA for further applications such as microarrays and RT-PCR gene expression analysis. By using magnetic beads, the total RNA of cells can be released by a lysis buffer solution and then trapped onto the surface of the magnetic beads in a solution with a lower pH value in an incubation process performed by the built-in microfluidic control module. Then, the magnetic beads with total RNA were further purified and collected by the on-chip magnetic bio-separator consisting of twisted microcoils. Finally the purified RNA can be reverse transcripted into cDNA utilizing the built-in micro-heater so that the more stable cDNA can be utilized for any further biological and medical applications. The human β-actin gene was used to ascertain RNA integrity, and to normalize PCR results for the amount of RNA assay. It was successfully amplified using the developed microfluidic chip. The development of the microfluidic system can provide a powerful platform for further biomedical applications.


The authors would like to thank financial support from the National Science Council in Taiwan (NSC 96-2120-M-006-008). The authors also would like to thank Mr. Tsung-Min Hsieh for his assistance with the micro-controllers. The access provided to major fabrication equipment at the Center for Micro/Nano Technology Research, National Cheng Kung University is greatly appreciated.

Copyright information

© Springer Science+Business Media, LLC 2008