A PROGRAMMABLE MICROFLUIDIC SYSTEM FOR SELECTIVE RNA OR DNA EXTRACTION FROM VARIOUS RAW BIOLOGICAL SAMPLES Michael Johnson 1* , Jungkyu Kim 1, Angela Williams 2, and Bruce Gale 1 1University of Utah, UNITED STATES OF AMERICA and 2Integrated Explorations, CANADA

ABSTRACT A system is presented that promises to be a step towards universal sample preparation in its ability to selectively extract and purify nucleic acids from various sources. A three-layer polydimethylsiloxane (PDMS) microfluidic chip is fabricated to perform the main fluid handling tasks. Nucleic acid is purified on the chip using standard solid phase extraction on a disposable glass fiber filter. The ability of the system to handle different sample inputs is demonstrated by extracting DNA from both a stock DNA solution and whole blood, and by extracting RNA form a both a stock RNA solution and living E. Coli cells.

KEYWORDS: Nucleic Acid, Sample Preparation, Microfluidics

INTRODUCTION Microfluidic systems are progressing towards the goal of reducing complex and time consuming laboratory analysis processes onto a single lab-on-a-chip device. These devices often rely on nucleic acid as the main analysis target because of its ubiquity, stability, and specificity. While genetic on-chip analysis techniques show great promise in expanding the role of microfluidic systems towards reducing human intervention in performing bio-assays, there is still a problem in that these sensing platforms demand purified nucleic acids, requiring typically manual genetic extraction and purification methods to remove unwanted biological material from the sample. This hurdle has limited the use of lab-on-a-chip devices to speciallized laboratories with the capability to perform the sample preparation steps, so widespread development of benchtop microfliudic analysis systems with true sample-in to answer-out capabilities is highly contingent on the development of nucleic acid purification sample preparation devices. Thus, while the number and type of microfluidic analysis techniques has grown in recent years, the practical use of lab-on-a-chip systems to perform wide- ranging laboratory-type tests has been limited and would be greatly benefited by the development of a sample preparation device that purifies genetic material from raw biological samples 1. The varied conditions of the sources of such material and the stringent requirements of analysis techniques force system-specific sample preparation operations, which often require time consuming and expensive manual handling 2. It would be of great benefit to have a self-contained lab-on-a- chip system capable of preparing genetic material out of raw biological samples from a variety of sources for microfluidic analysis. We present a system that promises to be a step towards universal sample preparation in its ability to take samples from various sources and selectively extract and purify nucleic acids (DNA or RNA).

THEORY The main purpose of the sample preparation device is to extract and purify nucleic acids from biological samples. The nucleic acid is purified on the chip using standard solid phase extraction on a disposable glass fiber filter 3. The extraction process operates by pumping the sample and various reagents through integrated pipetting/mixing reservoirs and through an extraction filter where the nucleic acid is selectively bound and released depending on buffer conditions. Altering the amounts, types, and sequences of reagents used allows selective adjustment of the protocol to handle different sample types and control the output as DNA, RNA or both and is easily accomplished by adjusting the control program. The vision is to produce an entire extraction system, such as that seen in Figure 1, that provides a compact, portable control platform with minimal inputs. (electrical power, a USB connector, and an optional pressure supply). The polydimethylsiloxane (PDMS) microfluidic chip seen in Figure 2(a) that performs the main fluid handling tasks of the system is fabricated in three layers: the fluidic channel layer, Figure 1: Programmable Microfluidic Nucleic Aci d the functional flexible membrane layer, and the pneumatic Extraction System. This compact device uses control layer. The PDMS chip contains on-chip valves, electric vacuum and pressure pumps along with a channels, reservoirs, and pumps that are all pneumatically bank of solenoid valves to control a PDMS controlled via electrically actuated solenoid valves (Lee microfluidic chip. A computer connection allows Company). A LabView computer program (National easy programming to adjust the protocol to handle Instruments) operates the solenoid valves in sequence to a variety of sample inputs.

978-0-9798064-3-8/µTAS 2010/$20©2010 CBMS 387 14th International Conference on Miniaturized Systems for Chemistry and Life Sciences 3 - 7 October 2010, Groningen, The Netherlands Figure 2: PDMS Chip Overview. (a) Photograph of the fabricated microfluidic chip measuring 125 mm x 100 mm. (b) Diagram of the chambers and channels of the microfluidic chip with inputs labeled for extraction of RNA from E. coli cells. Input reagents can be easily adjusted for other samples. perform the required range of tasks such as sample input, lysis, incubation, mixing, extraction, cleaning, etc. The PDMS chip also has input ports for the reagents required to extract nucleic acid from a variety of samples. Figure 2(b) is a schematic of the main chip functional layers with labels indicating which reagents are connected to the input ports. Following extraction, the purified genetic material can be mixed with other reagents required for analysis techniques— electrochemical detection in this case.

EXPERIMENTAL The automated system (Figure 1) contains a PDMS microfluidic platform complete with on-chip valves, reservoir pumps, and a disposable extraction filter. Fabrication of the PDMS chip involves a modified xurographic 4 technique for multi-level structures. Molds were created using a scotchcal vinyl tape (3M) cut on a sign plotter (Graphtec), with higher structures being cut from acrylic on a CO 2 laser and positioned over the tape with an adhesive. The control and operational fluidic layers are separated by a flexible silicone membrane (BISCO), and the layers are bonded using a novel masked corona discharge method. The disposable glass fiber filter used for solid phase extraction was fabricated from thin acrylic sheets and glass filter paper (Whatman). The acrylic sheets were cut on a laser and were bonded to enclose the glass filter with a double-sided adhesive. Known concentrations of RNA from an RNA standard stock (Invitrogen) were used as the sample. An RNA- containing solution was placed at the sample inlet port of the PDMS device. A protocol for on-chip mixing of the sample with reagents was used to bind the RNA to the glass filter, wash off any contaminants, and subsequently release the RNA in an elution of nuclease-free water. RNA extraction results were determined using a fluorescent Ribogreen RNA quantification kit obtained from Invitrogen. Fluorescence data was obtained using a 96- well plate reader (Bio-TEK). Binding of the nucleic acid to the PDMS and silicone surfaces was minimized using various coatings and solution stabilizers. RNA recovery yields from a sodium poly-phosphate (NaPP) coated system approached those from the commercial spin kit. However, these high-yield products displayed some NaPP contamination that interfered with reverse-transcript Figure 1: Manifold Mounting System. Photograph of polymerase chain reaction (RT-PCR) analysis methods. A the manifold mounting system that allows for rapid polyvinylperilidone (PVP) coating was used to produce replacement of the PDMS microfluidic chip. RT-PCR compatible outputs at the cost of a slightly reduced extraction yield. Nucleic acid extraction from raw biological samples was demonstrated by performing lysis and RNA extraction from E. Coli bacteria cells (New England Biolabs). The cells were incubated in cell culture medium and growth was monitored using optical density measurements to ensure viability. The E. Coli cells were then loaded into the automated system where they were enzymatically lysed, and RNA was extracted and quantified. Similarly, pre-purified human genomic DNA was used as the sample to be extracted in the system. The sample was drawn into the device and mixed with the appropriate reagents. The DNA extraction was quantified and verified by performing PCR on the eluate. Following verification of the ability to recover DNA, whole blood from a volunteer was used as the extraction sample and DNA was again extracted in the system and quantified using fluorescence and PCR.

388 A cleaning step was included in the program to allow for multiple tests on a single PDMS chip. A solution of RBS Neutral (Sigma-Adrich) is flushed twice over the system followed by three flushes of nuclease-free water (Qiagen). The programmable nature of the device allows for other disinfectants and cleaning agents to be introduced into the PDMS device, and the entire PDMS chip is also manifold-mounted [Figure 3] for easy replacement for sensitive tests where disposable units are desired.

RESULTS AND DISCUSSION The device was used to demonstrate successful automated RNA recovery at yields approaching results attained by a commercial spin kit. The ability of the system to handle different sample inputs is demonstrated by extracting DNA from both a stock DNA solution and whole blood, and by extracting RNA form a both a stock RNA solution and living E. Coli cells as shown in Table 1.

Table 1. Extraction results from operation of microfluidic system to recover nuceleic acid (NA) from various samples. These preliminary results show promise, especially since most optimization work done to date has been focused on the RNA extraction protocols with DNA extraction results providing a proof of concept.

Sample Total nanograms NA Retrieved E.coli 0157:H7 429 ng RNA Stock RNA 35.7 ng RNA Whole Blood 22.0 ng DNA Stock DNA 4.1 ng DNA

The total mass of retrieved nucleic acid shown in the table is to demonstrate the ability of the system to recover nucleic acid from a variety of samples and does not provide a comparison between their relative efficiencies as the starting concentrations were widely different. The percent yield of extracted nucleic acid was unavailable in some cases—due to uncertainty in starting concentration in raw samples—but all of the extracted samples were of sufficient quantity for subsequent detection and analysis. PCR and RT-PCR showed that RNA and DNA could be extracted from a variety of starting samples on a single programmable device. While some optimization of the extraction protocols has been performed, the DNA extraction protocols would benefit from additional work to increase the yield.

CONCLUSION This system has proven capable of handling a variety of sample preparation tasks in an effort to make automated nucleic acid sample preparation more practical and available. Such work is a step towards a universal sample preparation unit that can handle many kinds of samples from viruses to tissue samples and provide rapid, meaningful results from a lab-on-a-chip platform. The RNA detection abilities were demonstrated for two main applications. The first is the detection of E. Coli RNA to monitor drinking water supplies for harmful bacteria where RNA is a more appropriate genetic material for detection. RNA will also be extracted from viral samples to demonstrate another form of pathogen detection. These applications will open the door for future pathogen screening and detection tests. This project satisfies global demand for fast, simple, accurate, and adaptable genetic testing systems. There are numerous applications and industries that would widely benefit from such a device. For example, one of the specific applications kept in mind during the creation of this device was the detection of mutant DNA from circulating tumor cells (CTC’s). The DNA output from the sample preparation microfluidics will be delivered to a digital PCR device capable of detecting small concentrations of genetic mutations found in cancers. Having a rapid, automated, bench-top device with CTC detection capabilities would open the door to more research on the impact of CTC’s on cancer incidence and positively impact clinical testing for cancer metastasis and treatment efficacy.

ACKNOWLEDGEMENTS Funding for this project was generously provided by Early Warning LLC and Indian Immunologicals Ltd.

REFERENCES 1. “Nucleic acid extraction techniques and application to the microchip,” C.W. Price, D.C. Leslie and J.P. Landers, Lab on a Chip , 9, 2484-2494 (2009). 2. “Dealing with ‘real’ samples: sample pre-treatment in microfluidic systems,” A.J. de Mello and N. Beard, Lab on a Chip , 3, 11N-19N (2003) 3. “Microfluidic sample preparation: cell lysis and nucleic acid purification” Jungkyu Kim, Michael Johnson, Parker Hill and Bruce K. Gale, Integr. Biol ., 1, 574–586 (2009). 4. “Xurography: Rapid Prototyping of Micro-Structures Using a Cutting Plotter,” Daniel A. Bartholomeusz, Ronald W. Boutté, and Joseph D. Andrade, J. Microelectromech. Syst ., 14, 1364–74 (2005).

CONTACT *Michael Johnson, tel: +1-801-581-6549; [email protected]

389