AC Research
Anal. Chem. 1999, 71, 535-538
Articles
Rapid Amperometric Verification of PCR Amplification of DNA Thierry de Lumley-Woodyear,† Charles N. Campbell, Esti Freeman,‡ Amihay Freeman,‡ George Georgiou, and Adam Heller*
Department of Chemical Engineering, The University of Texas at Austin, Austin, Texas 78712-1062
Amplification of an 800-base template was verified in a 10-min test on a 2-µL sample of the PCR product solution. For verification, digoxigeninylated primers and biotinylated d-UTP-16-biotin were added to the amplification solution. The resulting amplified product was digoxigeninlabeled at its 3′-end and was also labeled with multiple biotin functions along its chain. The detecting electrode was coated with an electron-conducting redox hydrogel to which anti-digoxin monoclonal antibody was covalently bound. The amplified DNA was captured by the electrode through conjugation of its 3′-digoxigenin with the antibody. Exposure to a solution of horseradish peroxidase-labeled avidin led to capture of the enzyme and switched the redox hydrogel from a noncatalyst to a catalyst for H2O2 electroreduction. The switching resulted in an H2O2 electroreduction current density of 2.1 ( 0.9 µA cm-2 in 10-4 M H2O2 at Ag/AgCl potential and at 25 °C. Kopp et al.1 carried out recently a rapid continuous-flow polymerase chain reaction (PCR) on a chip. Using a micromachined chemical thermocycler, they carried out 20 cycles of amplification of a 176-base DNA segment in periods ranging from 1.5 to 18.7 min. Such rapid amplification is relevant to diagnostics and is potentially relevant also to prompt identification of biological warfare agents. Verification of the amplification, in a time period that is not longer than the amplification period itself, is of essence for operation of such a system. Gel electrophoresis and fluorescence detection of DNA using intercalating dyes is by far the most common technique for monitoring the formation of PCR products. However, even when gel electrophoresis is carried out in a MEMS † Present address: Cambridge Centre for Protein Engineering, University of Cambridge, Lensfield Rd, Cambridge, CB2 1EW, UK. ‡ Department of Molecular Microbiology and Biotechnology, Tel Aviv University, Ramat Aviv, 69978 Tel Aviv, Israel. (1) Kopp, M. U.; de Mello, A. J.; Manz, A. Science 1998, 280, 1046-7.
10.1021/ac980770h CCC: $18.00 Published on Web 01/05/1999
© 1999 American Chemical Society
device using gels less than 3 mm in length, the time required for DNA separation and detection has been reported to be of the order of 15 min.2 Alternatively, fluorescently labeled primers can be used to specifically label the DNA produced by amplification, but in this case, detection and quantitation of the amplified DNA first requires the removal of unincorporated primers. High-speed capillary electrophoresis can be employed for size separation of DNA fragments which must then be detected by spectroscopic or other methods.3 Here we describe a simple method for confirming, without high-speed separation, that all or part of the DNA was amplified. The method requires only a miniature lowpower potentiostat, which can be as small as a matchbox and can be made of inexpensive components.4 The verification is completed in 10 min, the required time being determined by mass transport considerations. The test utilizes 30 µL of the product solution/cm2 of electrode, amounting to 2 µL in case of the 3-mmdiameter electrodes actually used. It is based on a variant of the technique of doubly labeling the amplified DNA by biotin and digoxigenin5-11 and on the use of horseradish peroxidase-labeled avidin (Av-HRP) for enzyme amplification of the signal from the (2) Burns, A.; Johnson, B. N.; Brahmasandra, S. N.; Handique, K.; Webster, J. R.; Krishnan, M.; Sammarco, T. S.; Man, P.; Jones, D.; Heldsinger, D.; Mastrangelo, C. H.; Burke, D. T. Science 1998, 282, 484-7. (3) Effenhauser, C. S.; Paulus, A.; Manz, A.; Widmer, H. M. Anal. Chem. 1994, 66, 2949. (4) Quinn, C. P.; Wagner, J. G.; Heller, A.; Yarnitzky, C. N. Instrum. Sci., Technol. 1996, 24, 263. (5) Nickerson, D. A.; Kaiser, R.; Lappin, S.; Stewart, J.; Hood, L.; Landegren, U. Proc. Natl. Acad. Sci. U.S.A. 1990, 87, 6923. (6) Boeni, J.; Schuepbach, J. Mol. Cell. Probes 1993, 7, 361-71. (7) Ossewaarde, J. M.; Rieff, M.; van Doornum, G. J. J.; Henquet, C. J. M.; van Loon, A. M. Eur. J. Clin. Microbiol. Infect. Dis. 1994, 13, 732. (8) Xiao, L.; Yang, C.; Nelson, C. O.; Holloway, B. P.; Udhayakumar, V.; Lal, A. A. J. Immunol. Methods 1996, 199, 139. (9) Gutierrez, R.; Garcia, T.; Gonzalez, I.; Sanz, B.; Hernandez, P. E.; Martin, R. J. Appl. Microbiol. 1997, 83, 518. (10) Castillo, L.; Milano, G.; Santini, J.; Demard, F.; Pierrefite, V. Clin. Cancer Res. 1997, 3, 22137.
Analytical Chemistry, Vol. 71, No. 3, February 1, 1999 535
Figure 1. Scheme of the assay. The double-stranded 800-base DNA (top right) reacts with and binds to the anti-digoxin in the electronconducting redox hydrogel (top right). When the biotin residues of the DNA combine with avidin-labeled horseradish peroxidase, the enzyme is electrically “wired”. This “wiring” makes the electrode an electrocatalyst for H2O2 electroreduction to water through the cycles shown at the bottom.
PCR product. The scheme of the verification process is shown in Figure 1. EXPERIMENTAL SECTION Reagents. Sodium periodate (Catalog No. 31,144-8) and Tween 20 (Catalog No. 27,434-8) were purchased from Aldrich. Av-HRP (Catalog No. A-3151), biotin (Catalog No. B 4501), and monoclonal anti-digoxin (Catalog No. D8156; mouse IgG1 isotype) in ascites fluid were purchased from Sigma. Poly(ethylene glycol 400 diglycidyl ether) (PEGDGE), technical grade, was purchased from Polysciences (Catalog No. 08211). The redox polymer 7:1 copolymer of acrylamide and 1-vinylimidazole partially complexed with [Os(dmebpy)2Cl]+/2+ (PAA-PVI-Os) was synthesized as previously described.12 Its structure is shown in Figure 2. Electrodes and Electrochemical Equipment. Rotating disk electrodes were prepared by embedding a vitreous carbon rod (3-mm diameter, V10 Atomergic) in a Teflon shroud using a lowviscosity epoxy resin (Polyscience, Catalog No. 01916). Electrochemical measurements were performed with a Bioanalytical Systems CV-1B potentiostat. The signal was recorded on a Bioanalytical Systems XYT recorder. Rotating disk electrode experiments were performed with a Pine Instruments AFMSRX rotator with an MSRX speed controller. The electrochemical measurements were performed using a 14 mM phosphate pH 7.4 (11) Dorenbaum, A.; Vankateswaran, K. S.; Yang, G.; Comeau, A. M.; Wara, D.; Vyas, G. N. J. Acquired Immune Defic. Syndr. Hum. Retrovirol. 1997, 15, 35. (12) de Lumley-Woodyear, T.; Rocca, P.; Lindsay, J.; Dror, Y.; Freeman, A.; Heller, A. Anal. Chem. 1995, 67, 1332.
536 Analytical Chemistry, Vol. 71, No. 3, February 1, 1999
buffer in a three-electrode cell with a rotating (1000 rpm) glassy carbon working electrode, an Ag/AgCl Bioanalytical Systems reference electrode, and a platinum wire as counter electrode. Films. Films were formed on the 3-mm glassy carbon electrodes by cross-linking of hydrazide functions of the redox polymer with epoxide functions of PEGDGE. PAA-PVI-Os (5 mg mL-1) and PEGDGE (0.21 mg mL-1) were mixed in a 2:1 ratio. A 3-µL aliquot of the 2:1 solution was spread on the electrode and then allowed to dry and to cure at room temperature for 24 h. The films were uniform and shiny, swelled in water to electronconducting hydrogels, and adhered well to vitreous carbon. The anti-digoxin IgG solution (0.1 mL; concentration 6.7 mg mL-1) was purified using a HiTrap protein G affinity column from Pharmacia Biotech by elution with 0.1 M glycine and then by dialysis against pH 7.4 phosphate buffer solution (0.15 M NaCl, 0.020 M sodium phosphate), using Amicon Centriprep 30 tubes (4 × 30 min at 2700g). Next, carbohydrate groups on the antibody CH2 domain were oxidized to formyl groups by mixing 1 mL of the dialyzed antibody solution with 1 mL of sodium metaperiodate (10 mg mL-1). After oxidation of the antibody oligosaccharides, the excess sodium periodate was removed by dialysis using an Amicon Centriprep 30 tube. The concentration of the resulting anti-digoxin solution was 0.2 mg mL-1 as measured with a biuret kit, using BSA as the standard. An 8-µL drop of this mixture was deposited on the redox polymer-modified electrode. The electrode was kept for 16 h in a 25 °C chamber, in a saturated water atmosphere. The intact antibody molecules were immobilized by forming hydrazones between aldehydes of the monoclonal anti-
Figure 2. (A) Structure of the water-soluble copolymer of acrylamide and vinylimidazole, modified with hydrazine and [Os(dmebpy)2Cl]+/2+ (dmebpy ) 4,4′-dimethyl-2,2′-bipyridine functions) (PAA-PVI-Os). (B) Structure of the cross-linking agent, poly(ethylene glycol 600 diglycidyl ether) (PEGDGE).
digoxin and hydrazide functions of the polymer. The electrode was then rotated at 1000 rpm for 20 min in a 2% solution of PEGDGE in phosphate buffer. Next, the electrode was reacted for 20 min in pH 4.6 acetate buffer with 2% glutamate to eliminate all residual unreacted epoxide functions. The electrode was then incubated with 0.2% BSA for 10 min and stored in phosphate buffer solution at 4° C. PCR. An 800-base pair DNA fragment was amplified in a standard PCR amplification solution, to which the 3′-digoxigeninylated primer and dUTP-16-biotin were added. With these the PCR-amplified product was terminally labeled with digoxigenin and was also labeled with multiple biotin functions along the chain. The PCR was performed on 1 µL of the template solution, plasmid pTX152 encoding a scFvdig antibody gene. The template solution was mixed with 99 µL of amplification master mix (89 µL water, 10 µL of amplification buffer from Boehringer, 1 µM (each) BamH1 biotinylated primer (5′ biotin CTTCTTGATGGATCCGTCCTCGGGGTCTT) and #4 digoxigen conjugated primer from Genosys (5′ digoxigenin-TGGACCAACAACATCGGT), 200 µM dATP, dCTP, dGTP (each), 144 µM dTTP, 40 µM dUTP-16-biotin, and 5 units of Taq DNA polymerase from Boehringer (Catalog No. 1146165)). Amplification was performed in a thermocycler by using the following program: 1 cycle at 94 °C for 2 min, 53 °C for 2 min, and 72 °C for 3 min followed by 28 cycles at 94 °C for 1 min, 53 °C for 2 min, and 72 °C for 3 min, and 1 cycle at 72 °C for 7 min. At the end of the program, the tubes were held at 4 °C. The resulting double-stranded DNA consisted of 800 base pairs, some of which were modified with biotin and located at the end of a 16-carbon spacer arm. In addition, one end of the strand was labeled with digoxigenin.
Verification Procedure. The number and duration of the steps necessary for verifying the formation of the PCR product were minimized through multiparameter optimization. Both the DNA and the Av-HRP conjugation reactions with the film on the electrode were mass transport (reactant concentration and angular velocity) controlled. In the optimal procedure, 2 µL of the PCR sample was mixed with 700 µL of 0.1 M acetate buffer containing 0.1 M NaCl and 0.2% Tween 20. The amplified DNA was captured by the redox hydrogel on the electrodes through conjugating with their anti-digoxin antibodies,13 by rotating the electrodes at 1000 rpm in the diluted PCR solution for 5 min. Following capture of the amplified DNA, 700 µL of 1 µg mL-1 Av-HRP was added and the electrode was rotated for an additional 5 min, to label the captured DNA with peroxidase. RESULTS AND DISCUSSION When the completed electrode was immersed in 25 °C pH 7.4 phosphate buffer, the current density rose to 2.1 ( 0.9 µA cm-2 at 0.0 V versus Ag/AgCl upon adding H2O2 to a 10-4 M concentration (Figure 3). In an identical experiment, but with the binding sites of Av-HRP blocked with excess of biotin, to prevent specific binding of Av-HRP to the captured amplification product on the electrode, the current density rose only to 0.3 ( 0.2 µA cm-2. When the electrode was prepared with 2 µL of the amplification mixture (plasmid DNA, nucleotides, biotinylated oligonucleotides, biotinylated and digoxigeninylated primers), but (13) The equilibrium dissociation constant of the antibody’s FAB fragmentdigoxigenin conjugate is approximately 10-9: Chen, M. G.; Weiss, R.; Georgiou, G. unpublished.
Analytical Chemistry, Vol. 71, No. 3, February 1, 1999
537
Figure 3. Increase in the electroreduction current upon raising the H2O2 from zero to 10-4 M. The coated electrode was incubated with 700 µL of 0.1 M acetate buffer containing 0.1 M NaCl and 0.2% Tween 20 mixed with the following: (A) 0.7 µg mL-1 avidinylated HRP and 2 µL of the PCR sample; (B) 0.7 µg mL-1 avidinylated HRP and 2 µL of the plasmid DNA solution with nucleotides, biotinylated oligonucleotides, and biotinylated and digoxinylated primers, each at a concentration identical with its concentration in (A) before the PCR; (C) 0.7 µg mL-1 avidinylated HRP the binding sites of which were blocked with biotin and 2 µL of the PCR sample. Table 1. H2O2 Electroreduction Current Densities in the Presence and Absence of the Product of the PCR
PCR prod
Av-HRP
biotin
plasmid DNA, components of PCR mix
+ + -
+ + +
+ -
+
signal (µA cm-2) 2.1 0.5 0.3
without PCR amplification, the current density was 0.5 ( 0.2 µA cm-2 (Table 1). As is evident from Table 1 and from Figure 3, this simple and straightforward amperometric test rapidly detected the PCR product. Because the time required for completion of the verification was controlled by mass transport of the macromolecular
538 Analytical Chemistry, Vol. 71, No. 3, February 1, 1999
reactants DNA and Av-HRP to the surface of the electrode, it is expected that in a flow channel1 the greater than 100-fold dilution of the PCR product caused by the need for a large volume now required for a rotating electrode is obviated. Moreover, by using a PCR amplification mixture without dilution14 together with a more concentrated Av-HRP solution, the test times will be much shorter than the reported 10 min. Because 30 µL of the PCR solution was required per square centimeter, the volume of PCR solution to be used with a 10-µm-diameter microelectrode can be as low as 0.02 nL. Extrapolation of the observed current density of 2.1 µA/ cm2 and correction for the change from semi-infinite planar electron diffusion in the macroelectrode to radial electron diffusion in a similar H2O2 electroreducing 10-µm-diameter microelectrode15,16 suggests that the current will be as high as 4 nA and thus easily measurable with a simple and inexpensive lowpower potentiostat.4 Thus, amperometric detection with a microelectrode can be readily used for the detection of PCR amplification and be carried out in a nanofabricated solid-state device2 and is relevant to other applications where extremely small volumes of a DNA solution are used. While the method described in this report confirms the presence of amplified DNA, it provides no information on the size of the amplified product. Amplification products of incorrect size can be produced during PCR amplification, because of either the fortuitous annealing of the oligonucleotide primers to unrelated sequences or improper design of the amplification conditions. In diagnostic applications, such problems are commonly avoided by careful design of the oligonucleotide primers and of the PCR amplification conditions. ACKNOWLEDGMENT This work was supported by the Office of Naval Research, the National Science Foundation, and the Robert A. Welch Foundation. Received for review July 14, 1998. Accepted November 25, 1998. AC980770H (14) The PCR product solution was diluted 350-fold (2 µL was diluted with 700 µL of buffer). (15) Horrocks, B. R.; Schmidtke, D.; Heller, A.; Bard, A. J. Anal. Chem. 1993, 65, 3605. (16) Sakai, H.; Baba, R.; Hashimoto, K.; Fujishima, A.; and Heller, A. J. Phys. Chem. 1995, 99, 11896.
