An integrated temporary negative pressure assisted microfluidic chip for DNA isolation and digital PCR detection

Abstract

An integrated DNA purification and digital PCR (dPCR) detection microfluidic chip was developed in this study. This temporary negative pressure assisted microfluidic chip constructed of polydimethylsiloxane contained two distinct functional zones, one for DNA purification and the other for digital PCR (dPCR) detection. Sample lysate and reagent segments (washing buffer) for DNA purification were pre-loaded in a Teflon tube and pulled into the nucleic acids (NA) capture zone successively by negative pressure from the pipettor. The magnetic particles in the sample lysate were captured by the magnet and washed with buffer to get the purified DNA. DNA on magnetic particles were eluted by the PCR mix and then taken into the dPCR layer for dPCR reaction by the temporary negative pressure provided by the suction layer. Lastly, the suction layer was filled with water to avoid evaporation and ensure the efficiency of PCR in each chamber. This microdevice carried advantages such as ease of manufacture, ease of operation, and low cost.

---

Introduction

Food adulteration has occurred in some areas such as China and the European Union. In China, pork or even fox meat has been adulterated into beef or mutton to entrap the consumer. In 2013, the same stunning thing happened in the European Union with some processed meat sold in British and Irish supermarkets as beef actually containing horsemeat. Although governments have tightened control over food safety, food adulteration is an important consideration. Some chromatographic methods have been employed to detect several species of meat proteins.^1,2 However, these methods suffered from several practical problems, including poor column life and high costs. Spectroscopic determination^3–5 of adulteration is needed.

Food adulteration has been detected by means of PCR such as duplex PCR assay,⁶ single primer multiplex PCR (CSP-M-PCR) method,⁷ common primer multiplex PCR (CP-M-PCR) method,⁸ and real time PCR.^9–12 DNA testing is a useful way to confirm whether a product is authentic. When compared to other nucleic acid test methods, digital PCR (dPCR) is another way to detect food products.

Digital PCR is of higher sensitivity, which is achieved by diluting the starting sample to, at most, only one molecule in each well. After amplification, the number of wells containing the template is counted to determine the number of molecules in the starting sample using binomial Poisson statistics.^13,14 Some microfluidic devices were developed to perform dPCR such as integrated fluidic circuit (IFC) chip,¹⁵ megapixel digital PCR,¹⁶ femtoliter array,¹⁷ emulsion PCR or droplets,¹⁸ integrated self-priming compartmentalization chip,¹⁹ and a localized temporary negative pressure assisted microfluidic device.²⁰ In recent years, dPCR methods have been used to test fetal nucleic acids in maternal plasma for non-invasive prenatal diagnosis,²¹ to quantify the amount of the transgenic event in genetically modified organisms^22,23 and have been applied in performing the quantitative analysis of RNA transcripts from single cells.²⁴

The objective of lab-on-a-chip systems is to integrate all laboratory processes into a single device in which sample pre-treatment, biochemical reactions, separation, detection, and data analysis operations are combined.²⁵ The integration of multiple processes on-chip can limit sample loss, reduce analysis time, and enable new detection methods for microfluidic analyses.²⁵ Commercialized kits for the purification of nucleic acids (NA) were based on the method by Boom et al.,²⁶ which was based on the lysing and nuclease-inactivating properties of a chaotropic agent together with the nucleic acid-binding properties of silica magnetic particles in the presence of this agent. The microfluidics have shown prominent advantages such as low reagent and sample consumption, enhanced sensitivity, and increased speed.^27–29

There also have been some studies that integrate NA extraction to other bio-analyses on a microdevice. Easley et al.³⁰ reported an integrated microfluidic genetic analysis system that could extract and purify DNA from crude whole blood samples, and carry out PCR-based amplification. Wu et al.³¹ fabricated a glass microdevice that integrated the extraction of NAs and loop-mediated isothermal amplification (LAMP) on a single glass chip. Zhang et al.³² developed a droplet microfluidic device that exploited superparamagnetic silica particles for the solid binding of DNA, washing, elution and amplified DNA for genetic analysis within discrete droplets.

In the present study, dPCR and DNA isolation were integrated on a chip. The nucleic acids bound on magnetic particles were captured by a magnet, washed using three washing steps with buffer and eluted by the PCR mix. Then, the PCR mix took DNA into the dPCR module using the negative pressure of the chip for dPCR. This integrated microfluidic genetic analysis system is simple to fabricate and highly functional to detect bovine DNA in ovine meat for the detection of food adulteration.

Experimental

Microdevice fabrication

The temporary negative pressure assisted (TNPA) microfluidic chip was fabricated using soft lithography techniques with photoresist (SU-8, MicroChem) and polydimethylsiloxane (PDMS, Dow Corning). The TNPA chip was composed of three layers of PDMS, a thin blank PDMS layer, a digital PCR layer with NA isolation zone and digital PCR zone, and a suction layer. The patterns of the TNPA chip were designed using Corel DRAW X4 and printed on transparent film. The molds of the TNPA chip were prepared via spin-coating photoresist onto silicon wafers to create male molds.

The digital PCR layer mold was created using three layers of photoresist to create branch channels, main channels and chambers. First, a 4-inch silicon wafer was spin-coated with photoresist (SU-8 3025) at 3000 rpm for 30 s to create 25 μm high branch channels. After the wafer was soft baked at 95 °C for 15 min, the branch channels pattern were exposed using ultraviolet light. Then, the wafer was spin-coated with photoresist (SU-8 3025) at 1000 rpm for 30 s again to create 95 μm high main channels. After the wafer was soft baked at 95 °C for 30 min, the main channels pattern were exposed using ultraviolet light with a mask aligner. Lastly, the wafer was spin-coated with photoresist (SU-8 3050) at 1000 rpm for 30 s and exposed using ultraviolet light with the chambers pattern to create 210 μm high chambers with a mask aligner. The features of branch channels, main channels and chambers on the masks should be aligned well by the mask aligner. The 115 μm high suction layer mold was prepared by spin-coating with photoresist (SU-8 3050) onto silicon wafer (1000 rpm, 30 s), soft baked for 40 s and exposed using ultraviolet light with the suction layer pattern. After exposure and development, the molds were baked on a hotplate at 240 °C for 30 min to harden the features and then treated with a vapor of trimethylchlorosilane to prevent adhesion of the PDMS.

The TNPA chip was made from PDMS, which is a two component (A and B) elastomer. First, an approximately 2 mm thin suction layer was created by pouring 15 g of a 10A : 1B mixture of PDMS on the suction layer mold and baking at 80 °C for 30 min. Second, the suction layer was peeled off the mold to punch holes. The 500 μm thin digital PCR layer was created by spin-coating a 5A : 1B (excess Si–H groups) mixture of PDMS (1000 rpm, 30 s) on the digital PCR layer mold and baking at 80 °C for 5 min. Then, the suction layer covered the digital PCR layer, a soft rubber (10 mm × 8 mm) was placed above the NA capture zone and a 5A : 1B (excess Si–H group) mixture of PDMS was poured on them to bury the suction layer inside. The thin blank PDMS layer was created by spin-coating the 30A : 1B (excess vinyl groups) mixture of PDMS onto a clear silicon wafer at 1000 rpm for 30 s and baking at 80 °C for 40 min. Furthermore, the PDMS block on the digital PCR layer mold was peeled off to punch holes and aligned to the thin blank PDMS layer. After being baked at 100 °C for 1 h, the multilayer TNPA chip was peeled off.

Sample preparation

Commercial bovine meat and ovine meat were purchased from the local market. According to the specification of E.Z.N.A.™ Mag-Bind Tissue DNA Kit (Omega Bio-tek, M6223-01), 10 mg meat tissue was lysed to 750 μL sample lysate with magnetic particles to capture the DNA. To eliminate the weight error, bovine tissue lysate was mixed with ovine tissue lysate to make up a different weight ratio (WR) (w/w). The bovine tissue lysate was serially diluted with ovine tissue lysate at 10-fold dilutions from 1 : 10 to 1 : 10¹¹. Then, 20 μL of the diluted sample was introduced into the NA capture zone to purify the DNA.

For glyceraldehyde phosphate dehydrogenase (GAPDH) gene detection, about 1000 A549 lung carcinoma cells were collected to extract the total RNA using a AxyPrep™ Multisource Total RNA Miniprep Kit (Corning Life Sciences), and the total RNA was reverse transcribed into cDNA. The template cDNA was diluted at 5-fold dilutions from 0.2 to 1.6 × 10⁻³.

Microchip operation

In the commercial E.Z.N.A.™ Mag-Bind Tissue DNA Kit, multiple additions of reagents and washing ensure the high quality and high purity of the DNA. We applied a bubble-based method of reagent delivery to deliver the 5 separate reagent segments (sample lysate, washing buffer 1, two steps of washing buffer 2 and PCR mix) in a Teflon tube.

The capture zone in the digital PCR layer was a teardrop-shaped three-way tube (inlet, outlet 1 and outlet 2 to the digital PCR zone) with a magnet above. The Teflon tube with regent segments was connected with the inlet of the capture zone. When the pipettor that was connected to outlet 1 was turned up, the reagent segments in the Teflon tube entered into the NA capture zone in single file and the DNA on magnetic particles was purified. Purified DNA was eluted by the PCR mix and introduced into the digital PCR zone by the negative pressure from the suction layer.

Digital PCR

The performance of the TNPA chip was tested using serial dilutions of the cDNA solution of GAPDH at 4 orders of magnitude from 0.2 to 1.6 × 10⁻³. The PCR mix (40 μL) was prepared off-chip, comprised of 2 × TaqMan Gene Expression Master Mix 20 μL (Applied Biosystems, PN 4369016), forward primer 2 μL (10 nM), reverse primer 2 μL (10 nM), probe 2 μL (5 nM), RNase-free water 12 μL and the serially diluted template 2 μL. The PCR mix was loaded into the TNPA chip and dPCR performed on a bench-top PCR machine (MGL96G, Long Gene) using a flat-bed heating block. The thermocycling protocol included a 2 min heating step at 50 °C, 10 min hot start at 95 °C and then 40 cycles of: 30 s at 95 °C and 1 min at 60 °C. The experiments were repeated three times to ensure the robustness and reproducibility of the TNPA chip.

For the bovine meat detection, 20 μL bovine tissue lysate at 5 orders of magnitude from 1 : 10⁸ to 1 : 10¹¹ were introduced into the NA capture zone to purify the DNA. The PCR mix (40 μL) was prepared off-chip, comprised of 2 × TaqMan Gene Expression Master Mix 20 μL (Applied Biosystems, PN 4369016), forward primer 2 μL (10 nM), reverse primer 2 μL (10 nM), probe 2 μL (5 nM), RNase-free water 14 μL and eluted the DNA on magnetic particles into digital PCR zone. The thermocycling protocol of PCR included a 2 min heating step at 50 °C, 10 min hot start at 95 °C and then 40 cycles of: 30 s at 95 °C and 1 min at 60 °C.

Data acquisition and analysis

The chip after amplification was detected using a Maestro Ex INVIVO Imaging System (CRI Maestro). Fluorescence images were acquired using a large area CCD system. The fluorescence was excited at 455 nm and the emitted light was accepted by the CCD through a 495 nm long-pass filter. The image analysis and data processing were performed using Image-Pro Plus V 6.0 software in the Maestro EX IN-VIVO Image System. This software could detect the fluorescent intensity of each chamber in the images.

Results and discussions

Fabrication of the integrated chip and the principle of the negative pressure assisted pumping

The microfluidic chip was designed with a NA isolation zone and a digital PCR zone (Fig. 1a). The NA isolation zone was a 36.9 μL teardrop-shaped three-way tube (inlet, outlet 1 and outlet 2 to the digital PCR zone) with a groove for magnet (Fig. 1b). The digital PCR zone was a series of 2560 rectangular microwells (200 μm × 200 μm × 210 μm), which were located on interlaced bifurcation channels (25 μm in height) comprised of 16 main channels (2.5 cm in length and 95 μm in height). The ends of the 8 parallel main channels converged to a silicone oil injection port. Above the digital PCR zone was a rectangular suction layer (24 mm × 24 mm) with pillars (500 μm in diameter) to prevent collapsing under negative pressure (Fig. 1b). The diameter of the ports in the NA isolation zone, digital PCR zone and suction layer was 0.5 mm. The bottom of the chip was sealed with a lamina of PDMS and supported with a piece of cover slip (0.17 mm high).

Fig. 1 A schematic drawing shows the design of the microfluidic chip for nucleic acid purification and digital PCR. (a) A schematic of the microfluidic chip device, with the insets showing the array and chamber geometries. (b) A schematic of the layered device structure, which is composed of two layers of PDMS (a suction layer for negative pressure and hydration, a digital PCR layer with NA isolation zone and digital PCR zone containing hundreds of micro-wells and a thin blank PDMS layer).

When outlet 2 (Fig. 1a) in the three-way shaped NA isolation zone was shut down, the sample reagent could be sucked into outlet 1 from inlet using the negative pressure from outlet 1. According to this principle, when the silicone oil injection ports of the digital PCR zone were shut down by adhesive tape, the reagent could pass through the NA isolation zone from the inlet to outlet 1 and could not enter the digital PCR zone under negative pressure from outlet 2 (Fig. 1a and S3†). For DNA isolation, 20 μL sample lysate, three segments of 20 μL washing buffer and 30 μL PCR mix were pre-loaded by the pipettor in the Teflon tube one by one spaced with 20 μL air (Fig. S1†). When the pipettor connected to outlet 1 was pulled, the reagents in the Teflon tube entered the NA capture zone in single file. The leader magnetic particles with DNA were collected by a magnet (Fig. 2a), and the three washing steps using buffer were used to wash the particles to get pure DNA on the particles (Fig. 2b–d). The waste liquor was sucked into the pipettor from outlet 1. The flow rate of the reagents was controlled by adjusting the knob of the pipettor. Then, purified DNA on the particles was eluted using a PCR mix (Fig. 2e). The magnet was slid above the NA capture zone to drive the magnetic particles movement around and to improve the efficiency of the washing and elution steps.

Fig. 2 The principle and operational procedure of the TNPA microfluidic chip. (a) After the silicone oil injection ports were sealed using adhesive tape, when the pipettor that connected to outlet 1 was pulled, the sample lysate in the Teflon tube was sucked into the NA capture zone. The magnetic particles with NAs in the lysate were captured using the magnet. (b–d) When the pipettor continued to be pulled, the magnetic particles were washed using three washing steps with buffer. The magnet was slid in the groove to improve washing. (e) The PCR mix eluted the NAs off the magnetic particles. The magnet was also slid in the groove to improve elution. (f) When the syringe connected to the suction layer was pulled, air in the chambers permeated the PDMS into the suction layer with the two air pressure differences and brought the PCR mix into the chambers. (g) All the chambers were full of PCR mix. (h) Silicone oil was injected into the main channels from the silicone oil injection ports and rushed away the sample in the main channels. (i) The suction layer was filled with water to moisturize the digital PCR layer and avoid evaporation. (j) All the ports in chip were sealed with adhesive tape.

Because of the air permeability of PDMS, when the syringe connected to the suction layer was pulled to provide a negative air pressure, air in the digital PCR layer permeated the PDMS into the suction layer using the two air pressure differences and brought the PCR mix into each chamber (Fig. 2f and g). Then, the mixture of 4 g of silicone oil and 1 g of uncured PDMS (10A : 1B) was injected into the main channels from the silicone oil injection ports and rushed away the sample in the main channels (Fig. 2h) as described in previous study.¹⁹ The uncured PDMS polymers in the silicone oil could be cured during thermal cycling. Lastly, the negative air pressure was cancelled and the suction layer was filled with water for moisturizing (Fig. 2i). The sample was held in the chambers and digitized by the oil phase. A temporary negative pressure in the suction layer provided by a syringe implemented sample loading, which could reduce the complexity of the operation and the assistance of equipment such as a vacuum pump.

The breathability of PDMS facilitates sample loading when bringing evaporation during thermocycling. The evaporation in the chip not only leads to dryness in some chambers but also reduces the efficiency of PCR, and finally, brings an unfaithful dPCR result.²⁰ It is uneconomic and complicated to introduce a parylene C membrane¹⁶ and low-permeability fluorosilane polymer¹⁹ to restrict evaporation. To resolve this trouble, the suction layer serves a double-purpose in this microdevice not only for providing the negative pressure but also for moisturizing. The area of the suction layer was larger than the digital PCR layer, which moisturized the digital PCR layer to avoid evaporation and ensure the efficiency of PCR in each chamber (Fig. S3†).

Quantization of cDNA molecules by dPCR

To test the digital PCR response on the TNPA chip, the cDNA solution of GAPDH was prepared at 4 orders of magnitude from 0.2 to 1.6 × 10⁻³. The pre-mixed PCR mix (40 μL) contained 2 μL of diluted DNA template solution and was loaded into the TNPA chip using negative pressure. After thermocycling of the PCR, green fluorescence colour images (positive chambers) of the digital PCR on the microchip with concentrations of GAPDH cDNA were obtained using the Maestro Ex IN-VIVO Imaging System (Fig. 3a). The experiment for the same concentration was repeated three times. DNA molecules were randomly partitioned into each chamber; the probability of having at least one copy template per chamber was fitted with the Poisson distribution (Fig. 3b). According to the Poisson distribution, the stock concentration of DNA templates could be determined by the observed fraction of positive chambers under successful reaction in each chamber. The calculation was derived from the total number of chambers in the chip and the number of positive chambers. When compared the test using conventional real-time quantitative PCR (Fig. 3c), the results demonstrated the robustness of the dPCR chip (R² = 0.999). The raw statistical data are included in the electronic ESI (Table S1 and S2†). The stock concentration of the cDNA was determined from the linear regression fit (Fig. 3b), yielding a value of a stock concentration of c₀ = 7354 ± 647 copies per μL.

Fig. 3 Digital PCR results on the TNPA microchip. (a) Digital PCR fluorescent images (part) with a serial dilution of the target GAPDH DNA template ranging from 0.0016 to 0.2 dilutions. Complete images are shown in Fig. S4 found in the ESI.† (b) A regression curve was acquired by plotting −ln(1 − f₀) against the dilution factor X_dil, which is the linear form of the Poisson distribution equation. According to the equation, the copy number of the stock solution of DNA can be calculated. (c) A regression curve was acquired by plotting the C_t value in qPCR against the dilution factor X_dil.

NAs extraction and dPCR detection

After a section of 20 μL lysed sample with magnetic particles was pulled into the capture zone using negative pressure from the other side, the magnetic particles with NAs were collected using the magnet. Then, three washing steps using buffer in the Teflon tube were pulled into the capture zone and washed away proteins and/or other contaminants on the magnetic particles to get pure DNA. Finally, the DNA was eluted using a PCR mix and introduced into the dPCR region for subsequent dPCR analysis using negative pressure from the suction layer.

The reagents for the NA extraction and PCR in the amplification were incompatible. A two-step process was applied to keep fluidic isolation using negative pressure. With the pulling force from the other side of the capture zone using negative pressure, the reagents for NA extraction could not seep into the outlet 2 channel that connected the dPCR region, in case the reagents disturbed the subsequent dPCR reaction. As shown in Fig. 2, the reagents solution were introduced into the NA isolation area from the Teflon tube by the pipettor and did not seep into the dPCR region (also see Fig. S3†). Loading reagents with the pulling force using negative pressure is much simpler than the method using a pushing force. When the magnetic particles were collected together using the magnet, the magnetic particles were clumped together against washing and elution. A groove in the chip for the magnet to slide above the NA capture zone was introduced to drive the magnetic particles movement around and improve the efficiency of the washing and elution steps.

The DNA extraction efficiency of the microdevice was compared with the benchtop kit protocol, which was determined using real-time PCR. Purified nucleic acids obtained from the benchtop kit and microdevice were quantified using real-time PCR with the Real Time PCR Bovine and Ovine DNA Detection Kit (TAKARA, RR913). The experiment was tested in a gradient (WR = 10⁻³) and the DNA was quantified in triplicate using real-time PCR (Fig. S2†). Yield differences between the benchtop kit and microdevice were evaluated by the Mann–Whitney U statistical test (SPSS 19.0) (ESI Table S3†). On average, when the WR was 10⁻³, 0.628 ng μL⁻¹ bovine DNA were obtained from the microdevice in comparison to 0.747 ng μL⁻¹ using the benchtop kit (P = 0.127 > 0.05) in the tubes. There were non-significant differences in the DNA yields using the on-chip isolation method when compared to the benchtop procedure (Fig. 4). However, the error bars of the results from the microdevice were slightly high. The deviation of the reference standard in qPCR would enlarge the error bars of the samples. The loss would also affect the error bars when some PCR mix residue appeared from microdevice to the 0.2 mL tube. However, this situation did not influence the dPCR in the microdevice because the PCR mix for dPCR was redundant and there was no need for a reference standard in the dPCR.

Fig. 4 DNA obtained from the microdevice. (a) Average DNA concentration obtained in the microdevice method compared to the benchtop method. (b) Genomic DNA on magnetic particles was eluted using a PCR mix. Lane A was the PCR result of the genomic DNA from the microdevice (the PCR result was 89 bp).

The bovine meat in different mixtures with the ovine meat was detected using dPCR on the TNPA chip. Bovine lysate was mixed with ovine lysate in 10-fold serially ratios from 1 : 10 to 1 : 10¹⁰. Next, 20 μL mixed lysates, the DNA isolated and eluted with the PCR reagent mix in the microdevice. Degraded TaqMan probes in the positive chambers after PCR were excited at 455 nm and the emitted 518 nm light was detected by CCD through a 495 nm long filter. The images were analysed using the Image-Pro Plus V 6.0 software to count the number of positive chambers. The images of the dPCR on the chips with concentrations of bovine meat are shown in Fig. 5. As the ratio of bovine meat increased, the fraction of positive chambers increased, and when the WR was 10⁻¹¹, there was no positive chamber appearance (Fig. S6†). The bovine meat in different mixtures with the ovine meat was also detected using qPCR. The quantitative analysis of bovine DNA in qPCR is shown in Fig. S5.† Real-time PCR (qPCR) is a relative quantitative method with a reference standard and usually when the C_t > 35, the result of detection is negative by qPCR. However, the detection of dPCR could give a specific result whether it is negative or positive. It was positive by dPCR when WR = 10⁻¹⁰, but it was hard to say whether there was bovine DNA without a reference standard by qPCR.

Fig. 5 Digital PCR fluorescent images (part) on the microdevice with different concentrations of bovine lysate. Complete images are shown in Fig. S6 found in the ESI.†

Conclusions

Improvements in efficiency can be obtained by integrating DNA isolation with other bio-analytical technology on a microdevice for detecting impurities in food samples, pathogenic bacteria germs or cancer. In this study, a microdevice was developed by applying a commercial DNA isolation kit to purify DNA with magnetic beads and capture of the magnetic beads with a magnet in the microdevice. This microdevice could also suit other commercial kits with magnetic beads for RNA isolation and cell capture.

Step by step treatment is a good idea to integrate different modules in a microdevice. In this study, the DNA of the samples was isolated initially using negative pressure from the pipettor without disturbing the next dPCR step. In the next step, the purified DNA was introduced into the dPCR zone using negative pressure from a syringe. This step by step idea could extend to introduce more functional modules in a microdevice. In addition, negative pressure is a better approach to move fluid in a microdevice. In this paper, we applied negative pressure to move regents for DNA isolation from the inlet to outlet 1. If we applied a positive pressure to move the reagents, the reagents would also go into the outlet 2 and disturb the next dPCR step. Based on this idea, we could integrate more different functional modules in a microdevice using negative pressure and step by step treatment.

In some published study,³¹ DNA was eluted using water from the DNA isolation zone and mixed with the PCR mix in the PCR zone. However, this method would bring a problem in controlling the volume ratio of the PCR mix and eluted DNA in the microdevice to constitute functional PCR buffer. Under the chaotropic reagent, DNA was absorbed on the magnetic beads, purified by 70% ethanol and could be eluted using water, TE buffer or the PCR mix. In this study, PCR mix was applied to elute the DNA from a small amount of magnetic beads to reduce the complexity of the microdevice.

RNA isolation and specific cell capture were achieved using a commercial kit with cooperation between the magnetic beads and magnet. Besides DNA isolation, this microdevice could also be applied in RNA isolation and specific cell capture, and the acquired RNA or cells could also be placed downstream for dPCR analysis in the future. Therefore, this microdevice could also integrate RNA isolation and dPCR with some commercial RNA isolation and RT-real time PCR kits.

Acknowledgements

This study was supported by the National Natural Science Foundation of China (No. 31270907), the National key foundation for exploring scientific instrument (No. 2013YQ470781) and the Autonomous Research Project of the State Key Laboratory of Industrial Control Technology, China (No. 1501).

References

1. R. Wissiack, B. de la Calle, G. Bordin and A. R. Rodriguez, Meat Sci., 2003, 64, 427–432.
2. I. M. Oellingrath, A. Iversen and G. Skrede, Meat Sci., 1990, 28, 313–320.
3. H. B. Ding and R. J. Xu, J. Agric. Food Chem., 2000, 48, 2193–2198.
4. A. Rohman, Sismindari, Y. Erwanto and Y. B. Che Man, Meat Sci., 2011, 88, 91–95.
5. M. Zhao, G. Downey and C. P. O'Donnell, Meat Sci., 2014, 96, 1003–1011.
6. S. Soares, J. S. Amaral, I. Mafra and M. B. Oliveira, Meat Sci., 2010, 85, 531–536.
7. W. Xu, W. Bai, Y. Luo, Y. Yuan, W. Zhang, X. Guo and K. Huang, J. Sci. Food Agric., 2008, 88, 2631–2637.
8. W. Bai, W. Xu, K. Huang, Y. Yuan, S. Cao and Y. Luo, Food Control, 2009, 20, 366–370.
9. E. Delibato, D. Rodriguez-Lazaro, M. Gianfranceschi, A. de Cesare, D. Comin, A. Gattuso, M. Hernandez, M. Sonnessa, F. Pasquali, Z. Sreter-Lancz, M. J. Saiz-Abajo, J. Perez-de-Juan, J. Butron, E. Prukner-Radovcic, D. Horvatek Tomic, G. S. Johannessen, D. Jakociune, J. E. Olsen, M. Chemaly, F. le Gall, P. Gonzalez-Garcia, A. A. Lettini, M. Lukac, S. Quesne, C. Zampieron, P. de Santis, S. Lovari, B. Bertasi, E. Pavoni, Y. T. Proroga, F. Capuano, G. Manfreda and D. de Medici, Int. J. Food Microbiol., 2014, 184, 134–138.
10. E. J. Heo, B. R. Song, H. J. Park, Y. J. Kim, J. S. Moon, S. H. Wee, J. S. Kim and Y. Yoon, J. Food Prot., 2014, 77, 453–458.
11. A. Iwobi, D. Sebah, I. Kraemer, C. Losher, G. Fischer, U. Busch and I. Huber, Food Chem., 2015, 169, 305–313.
12. S. Soares, J. S. Amaral, M. B. Oliveira and I. Mafra, Meat Sci., 2014, 98, 41–46.
13. N. Blow, Nat. Methods, 2007, 4, 869–875.
14. L. B. Pinheiro, V. A. Coleman, C. M. Hindson, J. Herrmann, B. J. Hindson, S. Bhat and K. R. Emslie, Anal. Chem., 2012, 84, 1003–1011.
15. T. Thorsen, S. J. Maerkl and S. R. Quake, Science, 2002, 298, 580–584.
16. K. A. Heyries, C. Tropini, M. VanInsberghe, C. Doolin, O. I. Petriv, A. Singhal, K. Leung, C. B. Hughesman and C. L. Hansen, Nat. Methods, 2011, 8, 649–U664.
17. Y. F. Men, Y. S. Fu, Z. T. Chen, P. A. Sims, W. J. Greenleaf and Y. Y. Huang, Anal. Chem., 2012, 84, 4262–4266.
18. B. J. Hindson, K. D. Ness, D. A. Masquelier, P. Belgrader, N. J. Heredia, A. J. Makarewicz, I. J. Bright, M. Y. Lucero, A. L. Hiddessen, T. C. Legler, T. K. Kitano, M. R. Hodel, J. F. Petersen, P. W. Wyatt, E. R. Steenblock, P. H. Shah, L. J. Bousse, C. B. Troup, J. C. Mellen, D. K. Wittmann, N. G. Erndt, T. H. Cauley, R. T. Koehler, A. P. So, S. Dube, K. A. Rose, L. Montesclaros, S. L. Wang, D. P. Stumbo, S. P. Hodges, S. Romine, F. P. Milanovich, H. E. White, J. F. Regan, G. A. Karlin-Neumann, C. M. Hindson, S. Saxonov and B. W. Colston, Anal. Chem., 2011, 83, 8604–8610.
19. Q. Y. Zhu, L. Qiu, B. W. Yu, Y. N. Xu, Y. B. Gao, T. T. Pan, Q. C. Tian, Q. Song, W. Jin, Q. H. Jin and Y. Mu, Lab Chip, 2014, 14, 1176–1185.
20. Q. Tian, Q. Song, Y. Xu, Q. Zhu, B. Yu, W. Jin, Q. Jin and Y. Mu, Anal. Methods, 2015, 7, 2006–2011.
21. Y. M. D. Lo, F. M. F. Lun, K. C. A. Chan, N. B. Y. Tsui, K. C. Chong, T. K. Lau, T. Y. Leung, B. C. Y. Zee, C. R. Cantor and R. W. K. Chiu, Proc. Natl. Acad. Sci. U. S. A., 2007, 104, 13116–13121.
22. P. Corbisier, S. Bhat, L. Partis, V. R. D. Xie and K. R. Emslie, Anal. Bioanal. Chem., 2010, 396, 2143–2150.
23. D. Morisset, D. Stebih, M. Milavec, K. Gruden and J. Zel, PLoS One, 2013, 8(5), e62583.
24. A. K. White, K. A. Heyries, C. Doolin, M. VanInsberghe and C. L. Hansen, Anal. Chem., 2013, 85, 7182–7190.
25. P. N. Nge, C. I. Rogers and A. T. Woolley, Chem. Rev., 2013, 113, 2550–2583.
26. R. Boom, C. J. Sol, M. M. Salimans, C. L. Jansen, P. M. Wertheim-van Dillen and J. van der Noordaa, J. Clin. Microbiol., 1990, 28, 495–503.
27. D. R. Reyes, D. Iossifidis, P. A. Auroux and A. Manz, Anal. Chem., 2002, 74, 2623–2636.
28. P. A. Auroux, D. Iossifidis, D. R. Reyes and A. Manz, Anal. Chem., 2002, 74, 2637–2652.
29. P. S. Dittrich, K. Tachikawa and A. Manz, Anal. Chem., 2006, 78, 3887–3908.
30. C. J. Easley, J. M. Karlinsey, J. M. Bienvenue, L. A. Legendre, M. G. Roper, S. H. Feldman, M. A. Hughes, E. L. Hewlett, T. J. Merkel, J. P. Ferrance and J. P. Landers, Proc. Natl. Acad. Sci. U. S. A., 2006, 103, 19272–19277.
31. Q. Wu, W. Jin, C. Zhou, S. Han, W. Yang, Q. Zhu, Q. Jin and Y. Mu, Anal. Chem., 2011, 83, 3336–3342.
32. Y. Zhang, S. Park, K. Liu, J. Tsuan, S. Yang and T. H. Wang, Lab Chip, 2011, 11, 398–406.

---

Footnotes

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c5ra18166f

‡ These authors contributed equally to this article.

---