3691
Electrophoresis 2009, 30, 3691–3695
Erik Pettersson Patrik L. Sta˚hl Hovsep Mahdessian Max Ka¨ller Joakim Lundeberg Afshin Ahmadian Division of Gene Technology, AlbaNova University Center, Stockholm, Sweden
Received April 28, 2009 Revised July 15, 2009 Accepted July 20, 2009
Research Article
Visual DNA as a diagnostic tool We report on the incorporation of the Visual DNA concept in a genotyping assay as a simple and straightforward detection tool. The principle of trapping streptavidin-coated superparamagnetic beads of micrometer size for visualization of genetic variances is used for PrASE-based detection of a panel of mutations in the severe and common genetic disorder of cystic fibrosis. The method allows a final investigation of genotypes by the naked eye and the output is easily documented using a regular hand-held device with an integrated digital camera. A number of samples were run through the assay, showing rapid and accurate detection using superparamagnetic beads and an off-the-shelf neodymium magnet. The assay emphasizes the power of Visual DNA and demonstrates the potential value of the method in future point-of-care tests. Keywords: Cystic fibrosis / Diagnostics / Mutation / PrAse / Visual DNA DOI 10.1002/elps.200900273
1 Introduction The medical and forensic sciences are heavily dependent on various types of diagnostic testing tools. Different objectives, such as characterizing a pathogen, measuring the level of glucose in the blood, or obtaining an early indication of pregnancy, are all of great importance to the affected individuals as well as their physician. Although many parameters may be of importance and taken into careful consideration when selecting a diagnostic method, regardless of assay, the common requests are cost-effectiveness, speed, and accuracy. Improvements in these areas may enable testing of larger fractions of the population – in prenatal screening programs or in the case of a pandemic outbreak – and allow for quicker and more reliable predictions leading to improved health for the individual and for the population in general. However, detecting the presence of a hormone, such as the human chorionic gonadotropin in urine when using a pregnancy test [1] or measuring the levels of metabolites, such as glucose in the blood [2], is not as complex as elucidating the sequence composition of a DNA strand in the genome [3]. As a consequence, genetic testing still requires access to diagnostic labs with trained technicians, advanced
Correspondence: Dr. Afshin Ahmadian, Division of Gene Technology, Royal Institute of Technology (KTH), AlbaNova University Center, SE-106 91 Stockholm, Sweden E-mail:
[email protected] Fax: 146-8-5537-8481
Abbreviations: CF, cystic fibrosis; CFTR, cystic fibrosis transmembrane conductance regulator
& 2009 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
equipments, and extensive protocols, resulting in expensive and tedious tests. Although interrogating millions of genetic markers at a time [4], current state-of-the-art SNP scoring arrays from Illumina [5](www.illumina.com) and Affymetrix (www.affymetrix.com) are not always the optimal choice. There are situations when more accurate and efficient analysis of just a few markers is preferred, such as when testing for a specific genetic disorder or a drug response in a patient. An improvement in cost-effectiveness and rapidness of these systems would allow for distribution to more laboratories and test sites as well as a reduction in the time of analysis. The concept of Visual DNA [6] provides efficient and accurate genotype visualization of genetic markers due to the highly improved readout. Printing and immobilization of allele-specific primers in an array format onto a glass slide provides the means for running an in situ allele-discriminating assay with biotinylated nucleotides. Using a magnet underneath the slide, streptavidin-coated superparamagnetic beads of micrometer size can be attracted toward the glass surface and form covalent bonds with the DNA molecules. Thereby a 1000-fold magnification can be achieved, making nanometerscale DNA molecules visible to the naked eye. Genotyping is carried out in real time by simple ocular inspection and is documented using an offthe-shelf digital camera reducing the time of analysis and the need for expensive downstream scanning instrumentation and software tools. The method can, therefore, be employed for diagnostic purposes when screening hundreds of markers on a standardized platform. These authors contributed equally to this work.
www.electrophoresis-journal.com
3692
Electrophoresis 2009, 30, 3691–3695
E. Pettersson et al.
To illustrate the benefits of a Visual DNA-aided diagnostic test, an assay was developed for detecting mutations in cystic fibrosis (CF), one of the most common autosomal recessive genetic diseases [7]. Defective alleles of the cystic fibrosis transmembrane conductance regulator (CFTR) gene [8–10], an ion channel responsible for chloride transport across epithelial cell membranes, affect most organs in the body including the lungs, intestines, pancreas, and sweat glands. The symptoms can vary in type and severity, but the malfunction in epithelial cells causes thick mucus production resulting in recurrent bacterial infections in the lungs and hampered nutrient uptake in the intestines. In the Caucasian population approximately 1:3000–1:4000 [11] are born with the life-shortening disease, but with the proper care life beyond the third decade is now possible [12]. For this reason newborn screening is routinely recommended and procedures applied by hospitals encompass immunoreactive trypsinogen test [11], sweat tests [13, 14],] as well as genetic tests [15]. The latter, which often involves hybridization, ligation, or polymerization to interrogate DNA variations [16], is the most informative while also requiring advanced detection instrumentation, hence it is also the most expensive. Therefore, as a proof of concept, a panel of mutations was selected to demonstrate the concept of making DNA sequence variation visible to the naked eye as a part of an efficient diagnostic assay using only standard laboratory instrumentation.
2 Materials and methods 2.1 Selected mutations and genomic DNA samples Ten common mutations in the CFTR gene were selected for this study (www.genet.sickkids.on.ca/cftr/) (Supporting Information Table 1). The analysis was conducted on extracted genomic DNA from the blood of one voluntary Swedish individual and eight genomic DNA samples (Coriell, Camden, NJ, USA), each harboring one or more mutations. All samples were diluted to a concentration of 30 ng of genomic DNA/mL. Synthetic oligonucleotide templates were designed to evaluate the probes targeting 394delTT since none of the available samples contained the mutation. The template sequences were GATCCTTACCCCTAAATATAAAAAGATTCCATAGAACATA (wild type) and GAGATCCTTACCCCTAAATATAAAGATTCCATAGAACATA (mutation).
2.2 Amplification Six primer pairs were designed to target different exons of the receptor DNA sequence, covering the ten selected mutations. For each targeted template, one of the primers was biotinylated (Supporting Information Table 2). The biotinylation strategy was aimed to facilitate PCR cleanup and strand separation. The six amplicons (with lengths & 2009 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
ranging between 140 and 238 bp), harboring the ten variable positions, were generated using 1 U of Platinums Taq DNA Polymerase (Invitrogen, Carlsbad, CA, USA) in a final concentration of 20 mM Tris-HCl (pH 8.4), 50 mM KCl, 2 mM MgCl2, 0.2 mM dNTPs, and 0.1 mM of each of the two primers in a reaction volume of 50 mL. The reactions were cycled 42 times (951C 30 s, 541C 45 s, and 721C 45 s) using a 96-Well GeneAmps PCR System 9700 (Applied Biosystems, Foster City, CA, USA). Before the cyclic amplification, the reaction mixture was incubated at 951C for 5 min. In addition, after the cyclic amplification reactions, a final elongation step of 721C for 10 min was performed.
2.3 Microarray preparation Ten pairs of allele-specific oligonucleotides were designed to function as target capture probes on the glass slide as well as for genotyping the target template in the following in situ PrASE reaction (Supporting Information Table 3). The oligonucleotides were designed with a 50 -poly (T) spacer of 15 thymine residues as well as a 50 -terminus amino link with a C6 spacer. All oligonucleotides were synthesized by MWG-Biotech AG (Ebersberg, Germany). The allele-specific oligonucleotides were suspended at a concentration of 20 mM in 150 mM sodium phosphate, pH 8.5 and 0.06% sarkosyl solution (sarkosyl for improved spot uniformity) and were spotted using a Q-array (Genetix, New Milton, Hampshire, UK) onto CodeLinkTM Activated Slides (Surmodics, Eden Prairie, MN, USA). After printing, surface blocking was performed according to the manufacturer’s instructions. The oligonucleotides were printed in 12 identical arrays on the slide, and each array contained a predefined printing pattern (Supporting Information Fig. 1). The 12 subarrays were separated during hybridization by a 16-pad mask on a firm holder (ChipClipTM Schleicher & Schuell BioScience, Keene, NH, USA).
2.4 Hybridization The biotinylated PCR products were pooled (15 mL from each of the six simplex reactions) and immobilized to streptavidin-coated superparamagnetic beads (M-270 Dynal, Invitrogen). Using the Magnetic Particle Concentrator Dynal MPCTM-6 (Invitrogen), single-stranded products were generated by treating the complex with 8 mL of 0.1 M NaOH. The eluted non-biotinylated strand was then kept and neutralized with 6.67 mL 0.1 M HCl and 1.33 mL 10 AB (annealing buffer, 200 mM Tris-Ac/20 mM MgAc2, pH 7.6). The ssDNA samples were mixed with a pre-warmed (501C) 2 hybridization buffer (10 SSC 0.4% SDS) and deionised and filtered water (Milli-Q) to a total volume of 60 mL (1 hybridization buffer). The mixture was then warmed to 951C for 3 min before hybridization to array. Hybridization was performed at 501C for 40 min while shaking at a speed of 75 rpm. The 16 subarrays were separated during hybridization www.electrophoresis-journal.com
Electrophoresis 2009, 30, 3691–3695
by a 16-pad ChipClipTM mask (Schleicher & Schuell BioScience) on a firm holder. The mask was covered with standard white tape to prevent mixture evaporation and drying. The array was then washed with 2 SSC, 0.1% SDS buffer at 501C in a beaker while shaking at a speed of 45 rpm, followed by a 1 min 0.2 SSC buffer wash and a 1 min 0.1 SSC buffer wash, both at room temperature. The glass slides were dried by 10–20 s centrifugation in a standard table microcentrifuge (VWR International AB, Stockholm, Sweden).
2.5 In situ PrASE Adding two pre-warmed reaction mixtures (371C) denoted P1 and P2 consecutively; an in situ PrASE reaction [17, 18] utilizing biotinylated nucleotides was performed. A volume of 40 mL of the first solution (P1) consisting of 16 U of exonucleasedeficient (exo–) Klenow DNA polymerase (Fermentas, St. LeonRot, Germany), 1 EB (extension buffer, 50 mM Tris-HCl, pH 8, 5 mM MgCl2, 1 mM DTT), and 0.3% BSA was applied to the array and incubated for 30 s. The second solution (P2), containing 32 mg Proteinase K (Invitrogen) with 1 EB, 0.3% BSA, and nucleotides at a concentration of 1.25 mM (where dATP and dCTP were labeled with biotin (Invitrogen)) were then added to the array and incubated for 2 min. The mixture was then discarded and the slide was washed by the washing protocol used during the hybridization.
2.6 DNA visualization by bead trapping and digital imaging Detection was performed by applying a mixture of 3 mL (30 mg) of MyOneTM streptavidin-coated superparamagnetic beads (Invitrogen) and 27 mL 2 PBS, 0.2% Tween 20 onto the array. The beads were attracted toward the array surface by a 2-cm-diameter neodymium magnet held underneath the array for 20 s, allowing biotin–streptavidin bonds to form. Following the bead trapping onto the DNA and at the slide surface, the magnet was placed above the array for 5 s to attract and remove non-trapped beads. The array was then left to rest on top of the magnet for 1 min, allowing bonds to fully form while the excess solution was discarded by aspiration with a pipette. The array was then washed with Milli-Qs water (Millipore, Billerica, MA, USA) at 701C for 2 min in a glass chamber on a shaker (45 rpm) followed by a 10 s of centrifugation in a standard tabletop microcentrifuge (VWR International AB). Following naked eye investigation of the array, to document the results, the slide was held to a non-reflecting clear object and a digital image was captured using a Canon Digital IXUS 900 Ti (Canon Svenska AB, Solna, Sweden).
3 Results and discussion Simplified and more efficient genetic testing is highly desired in a clinical setting, reducing the time of analysis & 2009 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Nucleic acids
3693
and costs while promising a more widespread use of diagnostics in terms of targeted variations and number of genetic tests carried out. Outlined in this work is a diagnostic assay that illustrates a novel and user-friendly mutation screening assay for CF. The described proof of principle employs the Visual DNA concept [6] that is based on a magnetic bead detection platform. In short, it is constituted of an integrated system with a DNA microarray, an on-chip multiplex proteasemediated allele-specific primer extension [17, 18], and a visual detection of the mutations by trapping of micrometersized superparamagnetic beads allowing final investigation and genotype characterization by the naked eye. A printing pattern was designed and created to facilitate recognition and genotype identification without requiring any additional instrumentation (see Supporting Information Fig. 1). Large, the so-called megaspots were constructed by building a subpattern of spots with each allele-specific oligonucleotides rendering a 3 5-pattern with an area of 450 750 mm. The larger spots eliminate the need for scanning and magnification and facilitate naked eye analysis. In addition, a column of megaspots made up of mixtures of all oligonucleotides was added to the array to facilitate orientation and navigation on the array. Following hybridization of the prepared single-stranded PCR products, biotinylated nucleotides were incorporated onto the 30 termini of the immobilized oligonucleotides in the competitive enzymatic PrASE assay [17, 18], allowing extension discrimination between the match and mismatch strands (Fig. 1A). The addition of streptavidin-coated superparamagnetic beads (Fig. 1B) combined with the positioning of a magnet underneath the glass slide (Fig. 1C) allowed for an attraction of beads toward the surface and formation of biotin– streptavidin bonds, coupling the 1-mm-sized beads to the attached and extended DNA complex. By placing the magnet on top of the array, the excess of beads was attracted toward the magnet and removed from the array followed by quick washing. Finally, visual detection of the polymorphic positions and determination of the genotype (Fig. 1D and E) could be performed by simple inspection. The magnetic drag force on the beads and the required rupture force for breaking a biotin–streptavidin bond are both similar in size with a few pN each, as previously estimated [19, 20]. For this reason a single biotin–streptavidin bond should find it hard to resist rupture from the magnetic drag force, whereas a number of biotin–streptavidin bonds working together would be able to firmly attach a bead to the DNA-probe-coated surface. The assay was evaluated by genotyping nine individuals with known genetic profiles for ten selected mutations (394delTT, R117H, I507del, F508del, 1717-1, G542X, R553X, N1303K, G551D, and dele2.3) (see Supporting Information Table 1). All mutations were correctly genotyped. The first mutation (394delTT at position 1, see Supporting Information Fig. 1 and Table 1) was evaluated using synthetic oligonucleotide templates since no available www.electrophoresis-journal.com
3694
Electrophoresis 2009, 30, 3691–3695
E. Pettersson et al.
A
E
D
B
C Figure 1. Detection by Visual DNA. Allele-specific primers covalently attached to the surface of the array, in combination with an amplified target, provide the means for an in situ PrASE reaction. The allele-specific extension technique includes a protease in the reaction, thereby degrading the polymerase before unspecific mismatch extensions take place. After a successful PrASE reaction, the 450 750 mm megaspots on the array surface (A) are covered with biotinylated nucleotides if the targeted allele was present in the sample. By adding streptavidin-coated superparamagnetic beads, MyOneTM, with a diameter of 1 mm (B), and placing a magnet (C) directly underneath the glass slide for 20 s, the beads are attracted toward the surface and form bonds to the biotinylated strands present. After removal of excess beads and a brief wash the resulting pattern, visible to the naked eye, indicates the genotype of the sample. Theoretical images are shown where no detectable mutations are present (D) and where a CF patient homozygous for the mutation G542X in exon 11 (E) is diagnosed. The actual genotyping results of this mutation and another example can be seen in Figs. 2 and 3.
sample contained the mutation. The results, as seen in Fig. 2, show a very high S/N ratio and high specificity and accuracy with inspection by the naked eye. Since 2001, when the American College of Obstetricians and Gynecologists and the American College of Medical Genetics established a recommended screening panel of 25 CFTR mutations [15, 21], a number of commercial kits have been introduced to the market [16]. The methods are in most cases based on DNA hybridization and/or enzymatic interrogation but differ in their detection setup. Although they have wide commercial use, the methods could benefit from less tedious protocols and no additional instrumentation requirements. In this study, the array-based detection methodology simply relies on biotin–streptavidin interactions while providing an extremely rapid and simple visual detection utilizing neither expensive detection equipment nor software programs for data analysis. Hence, testing may also be performed in remote laboratories where such equipment is absent. Therefore, it may be preferable over existing screening technologies with respect to criteria such & 2009 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Figure 2. Genotyping results. (Left) No mutations detected (healthy individual). (Right) Homozygous for the mutation G542X in exon 11 (CF patient).
as mobility, hands-on time, and analytical simplicity making genetic analysis available to a vastly broader field of users. In addition, the method enables improvements in signal and background clarity that also makes it interesting for larger scale research and diagnostic groups. As mentioned earlier, the method herein uses printed slides for efficient DNA typing without the need for using dedicated detection instruments. Thus, the end user saves time and cost while the upstream array fabrication may be performed in an industrial setting. In addition, allele-discriminating tags could easily be designed for targeting different mutations and SNPs enabling other diagnostic application areas such as characterizing viral infections, drug-tolerance/resistance, bacterial infections, or, for instance, scoring SNPs in CYP450 genes for identifying fast and slow metabolizers [22]. Obviously, an important parameter in diagnostic assays is sensitivity and the PrASE reaction used to detect DNA variations is preceded by PCR amplification and has been demonstrated to be very sensitive as less than 1 ng genomic DNA, even in highly multiplex assays [18], has generated reliable results. Although most testing still requires a laboratory, the Internet has played a role in introducing genotyping to the public. Companies such as Mountain View-based 23andMe (www.23andme.com) and Iceland-based deCODE genetics (www.decodeme.com) are offering kits where upstream sampling and downstream data mining can be performed by the customer at home while the analysis, requiring heavy instrumentation for amplification, hybridization, allele discrimination, and detection are performed at the companies’ sites. Another example is testing for a venereal disease such as Chlamydia that can be diagnosed by ordering a free test via the web (www.klamydia.se) and sampling as well as the final communication of the test result are taking place within the privacy of one’s own home. It is also likely that the more frequent use of hand-held devices, integrating communication tools as well as cameras, will soon enter the clinics. As an example, the Apple App Store (www.apple.com/iphone/appstore/) is currently housing hundreds of medical software tools for the iPhone with many more to come and the platform has recently been demonstrated interacting with portable glucose monitors [2](www.lifescan.com) further indicating www.electrophoresis-journal.com
Nucleic acids
Electrophoresis 2009, 30, 3691–3695
3695
5 References [1] Vaitukaitis, J. L., Ann. NY Acad. Sci. 2004, 1038, 220–222. [2] Montagnana, M., Caputo, M., Giavarina, D., Lippi, G., Clin. Chim. Acta 2009, 402, 7–13. [3] Pettersson, E., Lundeberg, J., Ahmadian, A., Genomics 2009, 93, 105–111. [4] Kaller, M., Lundeberg, J., Ahmadian, A., Expert Rev. Mol. Diagn. 2007, 7, 65–76.
Figure 3. Screening patterns for an individual with no detectable mutations (top), an individual heterozygous for mutation R117H (middle), and an individual homozygous for mutation G542X (bottom). The actual size of the readout is visualized by using the reverse side of a 5-cent coin.
future use of such devices in medical sciences. In a clinical setting, the iPhone would be ideal for documenting Visual DNA images and for communicating genotype information from the technician, via the doctor, to the patient. The recent progress in diagnostics mentioned earlier, in combination with true naked eye detection of mutations provided by Visual DNA (Fig. 3), has the potential of further contributing in the domestication of genetic testing.
[5] Steemers, F. J., Gunderson, K. L., Biotechnol. J. 2007, 2, 41–49. [6] Stahl, P. L., Gantelius, J., Natanaelsson, C., Ahmadian, A., Andersson-Svahn, H., Lundeberg, J., Genomics 2007, 90, 741–745. [7] Rowe, S. M., Miller, S., Sorscher, E. J., N. Engl. J. Med. 2005, 352, 1992–2001. [8] Kerem, B., Rommens, J. M., Buchanan, J. A., Markiewicz, D., Cox, T. K., Chakravarti, A., Buchwald, M., Tsui, L. C., Science 1989, 245, 1073–1080. [9] Riordan, J. R., Rommens, J. M., Kerem, B., Alon, N., Rozmahel, R., Grzelczak, Z., Zielenski, J. et al., Science 1989, 245, 1066–1073. [10] Rommens, J. M., Iannuzzi, M. C., Kerem, B., Drumm, M. L., Melmer, G., Dean, M., Rozmahel, R. et al., Science 1989, 245, 1059–1065. [11] Sharp, J. K., Rock, M. J., Clin. Rev. Allergy Immunol. 2008, 35, 107–115.
4 Concluding remarks This effort illustrates an efficient and rapid principle of using superparamagnetic beads for naked eye visualization of genetic variances in a diagnostic assay. In its current format, a regular PCR apparatus available in most laboratories is sufficient and the overall time required for the CF screening protocol adds up to 3 h. A possible future improvement of the assay includes proper adaption of an isothermal amplification protocol, eliminating the need of PCR machinery while rendering a single-stranded product. In addition, although the present panel represents the most common mutations, increasing the number of targeted mutations to match current standards [15] would be required for clinical use, a task possible since the PrASE assay has been demonstrated genotyping more than 140 different alleles in a single reaction [23]. We wish to thank the Swedish Research Council and VINNOVA for financial aid. Further we would like to thank Anders Hedrum and Ulf Klangby at Devyser AB for valuable advice and Anna Westring and Annelie Walde´n for their assistance in the production of glass slides. The authors have declared no conflict of interest.
& 2009 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
[12] Proesmans, M., Vermeulen, F., De Boeck, K., Eur. J. Pediatr. 2008, 167, 839–849. [13] Gibson, L. E., Cooke, R. E., Pediatrics 1959, 23, 545–549. [14] Voter, K. Z., Ren, C. L., Clin. Rev. Allergy Immunol. 2008, 35, 100–106. [15] Grody, W. W., Cutting, G. R., Klinger, K. W., Richards, C. S., Watson, M. S., Desnick, R. J., Genet. Med. 2001, 3, 149–154. [16] Johnson, M. A., Yoshitomi, M. J., Richards, C. S., J. Mol. Diagn. 2007, 9, 401–407. [17] Hultin, E., Kaller, M., Ahmadian, A., Lundeberg, J., Nucleic Acids Res. 2005, 33, e48. [18] Pettersson, E., Lindskog, M., Lundeberg, J., Ahmadian, A., Nucleic Acids Res. 2006, 34, e49. [19] Ota, T., Sugiura, T., Kawata, S., Appl. Phys. Lett. 2005, 87, 43901–43903. [20] Pamme, N., Manz, A., Anal. Chem. 2004, 76, 7250–7256. [21] Watson, M. S., Cutting, G. R., Desnick, R. J., Driscoll, D. A., Klinger, K., Mennuti, M., Palomaki, G. E. et al., Genet. Med. 2004, 6, 387–391. [22] Ingelman-Sundberg, M., Trends Pharmacol. Sci. 2004, 25, 193–200. [23] Pettersson, E., Zajac, P., Stahl, P. L., Jacobsson, J. A. Fredriksson, R., Marcus, C., Schioth, H. B. et al., Hum. Mutat. 2008, 29, 323–329.
www.electrophoresis-journal.com
