Anal. Chem. 1997, 69, 4828-4832
Ultrasensitive Voltammetric Detection of Underivatized Oligonucleotides and DNA Pankaj Singhal and Werner G. Kuhr*
Department of Chemistry, University of California, Riverside, California 92521
Electrochemical detection of nucleotides, ssDNA, and dsDNA was accomplished by using sinusoidal voltammetric detection at copper microelectrodes. Generally, detection of these molecules utilizes the electroactive nature of adenine and guanine residues at most electrode surfaces. The detection approach used in this study is based on the electrocatalytic oxidation of sugars and amines at copper surfaces. All nucleotides and DNA molecules comprise a ribose sugar backbone and primary amines present on the different nucleobases. Consequently, the detection approach is universal to all types of nucleotides. As the number of sugar moieties increases with the length of an oligonucleotide, the detection sensitivity is enhanced for bigger oligonucleotides. Irreversible adsorption of these oligonucleotides and other biomacromolecules like dsDNA on the electrode surface was avoided with sinusoidal voltammetry since it is a scanning electrochemical technique. The sensitivity of the detection strategy is, however, still preserved due to the effective decoupling of the faradaic signal from the capacitive background currents in the frequency domain. The ssDNA and dsDNA were detected in the picomolar concentration range. The electrochemical signal due to dsDNA is actually higher than that due to ssDNA due to the larger number of easily accessible sugars on the outer perimeter of a dsDNA double helix compared to those on a ssDNA of the same size. This is in contrast to the existing electrochemical detections techniques based on the electroactivity of the nucleobase. Detection of underivatized DNA is highly desirable in order to avoid any sample handling losses and contamination problems.1-3 Various analytical methodologies, including UV absorbance,4 fluorescence,5-9 electrochemical detection,3,10,11 and electrochemically generated chemiluminescence,12,13 have been used for the
sensitive and selective detection of DNA fragments. Generally, DNA analysis involves tedious derivatization and sample preparation protocols, which are undesirable, especially considering the limited sample availability and the possible ease of its contamination and degradation. Electrochemical methods3,10,14,15 seem particularly well suited for the generally sample-limited case of DNA analysis, since they can be miniaturized with ease (capable of working in nanoliter to picoliter volumes) without sacrificing sensitivity or selectivity. To date, most electrochemical detection protocols are based on the electroactivity of the nucleobases11,16-21 or the adsorption of ssDNA to complementary strands immobilized on an electrode surface (this also requires the use of an electroactive molecule that intercalates or otherwise associates with double-stranded DNA).10,12,13,22-25 Since only the purine bases, adenine and guanine, are significantly electroactive at most surfaces,11,20,21,26 it has been difficult to analyze pyrimidine-basecontaining nucleotides at an electrode surface. This, in turn, limits the applicability of these methods, since, for example, they cannot be used for sequencing DNA via the digestion of DNA using an exonuclease enzyme. All detection strategies which focus on the electroactivity of the nucleobase face a severe decrease in signal for dsDNA as compared to that for ssDNA, since the bases are on the inside of the double helix, and thereby their detection is sterically hindered due to the sugars surrounding them.3,20 Rather than focusing on the electrochemical processes of the nucleobases in DNA, we chose a more universal detection approach based on the electrocatalytic oxidation of sugars in nucleotides and DNA. Since all nucleotides and DNA molecules contain a ribose sugar backbone, electrochemical oxidation of this often-forgotten component of the oligonucleotide may lead to very sensitive measurements. Copper surfaces have been found to catalyze the oxidation of most sugars (at potentials greater than +0.4 V) due to an electrocatalytic mechanism involving the redox couple Cu(III) f Cu(II).27-35
(1) Conner, B. J.; Reyes, C. M.; Morin, C.; Itakura, K.; Teplitz, R. L.; Wallace, R. B. Proc. Natl. Acad. Sci. U.S.A. 1983, 80, 278. (2) Landegren, U.; Kaiser, R.; Caskey, C. T.; Hood, L. Science 1988, 242, 229. (3) Palecek, E. Electroanalysis 1996, 8, 7-14. (4) Holden, M.; Pirie, N. W. Biochim. Biophys. Acta 1955, 16, 317-321. (5) Goodwin, P. M.; Johnson, M. E.; Martin, J. C.; Ambrose, W. P. a. o. Nucleic Acid Res. 1993, 21, 803-806. (6) Huang, Z.; Petty, J. T.; O’Quinn, B.; Longmire, J. L.; Brown, N. C.; Jett, J. H.; Keller, R. A. Nucleic Acids Res. 1996, 24, 4202-4209. (7) Karsten, U.; Wollenburger, A. Anal. Biochem. 1977, 77, 464-470. (8) Le Pecq, J. B.; Paoletti, C. Anal. Biochem. 1966, 17, 100-107. (9) Markovits, J.; Roques, B. P.; Le Pecq, J. B. Anal. Biochem. 1979, 94, 259264. (10) Mikkelson, S. R. Electroanalysis 1996, 8, 15-19. (11) Kafil, J. B.; Cheng, H. Y.; Last, T. Anal. Chem. 1986, 58, 285-289. (12) Rodriguez, M.; Bard, A. J. Anal. Chem. 1990, 62, 2658-2662. (13) Xu, X. H.; Bard, A. J. J. Am. Chem. Soc. 1995, 117, 2627-2631.
(14) Palecek, E. Nature 1960, 188, 656-657. (15) Palecek, E.; Postbieglova, I. J. Electroanal. Chem. 1986, 214, 359-371. (16) Lin, H.; Xu, D. K.; Chen, H. Y. J. Chromatogr. A 1997, 760, 227-233. (17) Nakahara, T.; Okuzawa, M.; Maeda, H.; Hirano, M.; Matsumoto, T.; Uchimura, H. J. Chromatogr. 1992, 15, 1785-1796. (18) Pang, D.-W.; Qi, Y.-P.; Wang, Z.-L.; Cheng, J.-K.; Wang, J.-W. Electroanalysis 1995, 7, 774-777. (19) Singhal, P.; Kuhr, W. G. Anal. Chem. 1997, 69, 3552-3557. (20) Wang, J.; Chen, L.; Chicharro, M. Anal. Chim. Acta 1996, 319, 347-352. (21) Xu, D. K.; Hua, L.; Chen, H. Y. Anal. Chim. Acta 1996, 335, 95-101. (22) Millan, K. M.; Mikkelsen, S. R. Anal. Chem. 1993, 65, 2317-2323. (23) Millan, K. M.; Saraullo, A.; Mikkelsen, S. R. Anal. Chem. 1994, 66, 29432948. (24) Wang, J.; Cai, X.; Rivas, G.; Shiraishi, H.; Farias, P. A. M.; Dontha, N. Anal. Chem. 1996, 68, 2629-2634. (25) Wang, J.; Rivas, G.; Luo, D.; Cai, X.; Valera, F. S.; Dontha, N. Anal. Chem. 1996, 68, 4365-4369. (26) Palecek, E.; Boublikova, P.; Jelen, F. Anal. Chim. Acta 1986, 187, 99-107.
4828 Analytical Chemistry, Vol. 69, No. 23, December 1, 1997
S0003-2700(97)00731-2 CCC: $14.00
© 1997 American Chemical Society
Therefore, one would expect that the sensitivity for detection would also improve as the number of ribose sugar moieties increases with the chain length of the nucleotide, thereby enabling ultrasensitive detection of native, underivatized oligonucleotides and DNA strands. Previously, we have shown that sinusoidal voltammetry is able to detect sugars at a copper surface with very high sensitivity and selectivity compared to existing electrochemical methods.33 Sinusoidal voltammetry uses a sine wave excitation to elicit a current response at the electrode surface.19,33 The response obtained is converted into the frequency domain, and all the harmonics of the fundamental excitation are monitored, since these contain almost all of the current response obtained. The more nonlinear nature of the faradaic current compared to the background charging current is capitalized upon to discriminate the analyte signal sensitively over the background. The measurement at the higher harmonics is, consequently, much more sensitive than all time domain-based electroanalytical methods, leading to detection limits almost 2 orders of magnitude lower than those obtained with dc methods.33 Another issue which limits the sensitivity of direct electrochemical measurements of most biopolymers is the irreversible adsorption of these materials onto most electrode surfaces.18,36 In order to minimize fouling, the potential applied to the electrode surface is often pulsed through the region at which the electrode material is oxidized to regenerate a fresh electrode surface (pulsed ameperometric detection, or PAD).37 While this leads to sensitivity comparable to that of dc amperometric measurements, no qualitative information is generated during this process. To obtain qualitative information, a scanning waveform must be used. Unfortunately, most electrochemical methods which scan the applied potential do not have the same sensitivity as dc detection schemes.38 Thus, sinusoidal voltammetry, which utilizes a scanning potential, provides qualitative information (i.e., frequency spectra) and has the additional advantage of minimizing fouling by the adsorption of large molecules (i.e., DNA strands). EXPERIMENTAL SECTION Reagents. The water used was deionized and then passed through a Milli-Q water purification system (Millipore Corp., Bedford, MA). The oligonucleotides containing repeated units of deoxycytosine (dC) were custom synthesized by Genemed Synthesis, Inc. (San Francisco, CA). Three different lengths containing 10 bases, 120 bases, and 400+ bases (poly dC), respectively, were obtained and used as received. Cytosine 5′monophosphate (98-99%, Sigma Chemical Co., St. Louis, MO) and sodium hydroxide (ACS grade, Fisher Scientific, Fair Lawn, NJ) were used as received. All experiments were done with 0.10 (27) Colon, L. A.; Dadoo, R.; Zare, R. N. Anal. Chem. 1993, 65, 476-481. (28) Kano, K.; Ueda, T. J. Electroanal. Chem. 1994, 372, 137-143. (29) Luo, P.; Zhang, F.; Baldwin, R. P. Anal. Chim. Acta 1991, 244, 169-178. (30) Marioli, J. M.; Kuwana, T. Electrochim. Acta 1992, 37, 1187-1197. (31) Prabhu, S. V.; Baldwin, R. P. Anal. Chem. 1989, 61, 2258-2263. (32) Prabhu, S. V.; Baldwin, R. P. Anal. Chem. 1989, 61, 2258-2263. (33) Singhal, P.; Kawagoe, K. T.; Christian, C. N.; Kuhr, W. G. Anal. Chem. 1997, 69, 1662-1668. (34) Vassilyev, Y. B.; Khazova, O. A.; Nikolaeva, N. N. J. Electroanal. Chem. 1985, 196, 127-144. (35) Xie, Y.; Huber, C. O. Anal. Chem. 1991, 63, 1714-1719. (36) Johnson, D. C.; LaCourse, W. R. Anal. Chem. 1990, 62, 589A-597A. (37) LaCourse, W. R.; Johnson, D. C. Anal. Chem. 1993, 65, 50-55. (38) Wiedemann, D. J.; Kawagoe, K. T.; Kennedy, R. T.; Ciolkowski, E. L.; Wightman, R. M. Anal. Chem. 1991, 63, 2965-2970.
M sodium hydroxide as the running electrolyte. Several 0.01 M stock solutions of the cytosine 5′-monophosphate (CMP) were prepared in deionized water. Subsequent dilutions were made whenever required using the running electrolyte. The oligonucleotide samples were prepared in the running electrolyte and used immediately to avoid any degradation by sodium hydroxide. Preparation of dsDNA and ssDNA. A size-selected (>2 kb) cDNA population was synthesized from mRNA isolated from 3 day embryo of the tobacco hornworm, Manduca sexta.39 cDNAs were cloned into SalI/NotI-digested pSPORT (Gibco-BRL) vector (∼4 kb) and amplified in Escherichia coli strain DH5 R. A 5.5 kb cDNA clone was obtained by screening this library. The cDNA codes for a 745 amino acid hydrophilic protein. Plasmid preparation was done using Promega Wizard miniprep protocols. The total size of the double-stranded DNA ()9.5 kb) is from 5.5 kb of cDNA insert and 4 kb of the vector in to which it is cloned. The ssDNA sample was prepared from the above dsDNA. The 9.5 kb dsDNA was denatured by standard steps using alkali denaturation and precipitation in ethanol.40 Copper Microelectrodes. The fabrication of 20 µm Cu microelectrodes has been described elsewhere.33 Prior to use, the electrodes were polished with a 1 µm diamond polish, followed by sonication in water. No electrochemical activation was performed in an effort to minimize the occurrence of background faradaic processes at the electrode surface. The potential at the electrode was cycled under the experimental conditions for about 1 h prior to the collection of data to achieve a stable background response. Instrumentation. The flow injection analysis (FIA) apparatus has been described previously33,41 and includes a pneumatic actuator (Rheodyne, Model 5701) controlled via a solenoid valve (Rheodyne kit, Model 7163). This cell was designed to match the internal diameter of the FIA tubing (0.75 mm) in order to minimize convective dispersion of the analyte as it was transported to the microelectrode. Finally, the flow rate (0.5 mL/min) was controlled by gravity flow by maintaining a height difference of 25 cm between the running electrolyte container and the flow cell. The volume of sample injected into the flow stream was determined by the flow rate and the length of the injection. Typically, an injection time of 60 s was used, producing an injection volume of 500 µL. This injection protocol allowed the electrode to see the full concentration of the injected sample without dispersion or dilution and to obtain a flat-top response. Sinusoidal voltammetry33 was performed as described previously. The specific conditions used for these experiments are described here. Sinusoidal voltammetry was performed by digitally generating a 2 Hz sine wave (exactly 1.95 Hz, 0.05-0.55 V vs Ag/AgCl) with software provided by Axon Instruments (SineVolt, Axon Instruments Inc., Foster City, CA). This excitation waveform was filtered with a four-pole low-pass filter having a 3 dB point at a frequency three times the fundamental frequency using a Cyberamp (Model 380, Axon Instruments Inc.). The filtered excitation waveform was supplied to the Cu electrode through a three-electrode potentiostat (Geneclamp, Axon Instru(39) Mbungu, D.; Ross, L.; Gill, S. Arch. Biochem. Biophys. 1995, 318, 489497. (40) Sambrook, J. J.; Fritsch, E. F.; Maniatis, T. Molecular Cloning. A Laboratory Manual.; Cold Spring Harbor Laboratory Press: Cold Spring Harbor, NY, 1989. (41) Kristensen, E. W.; Kuhr, W. G.; Wightman, R. M. Anal. Chem. 1987, 59, 1752.
Analytical Chemistry, Vol. 69, No. 23, December 1, 1997
4829
ments Inc.). The current output of the potentiostat was filtered with a four-pole low-pass filter having a 3 dB point at 200 Hz to avoid aliasing and then by a second four-pole filter with a Cyberamp set at 80 Hz to further minimize noise (at 4 times the maximum frequency component analyzed (20 Hz)). The current was sampled digitally with a 12 bit A/D converter (1200A, Axon Instruments Inc.) at a rate of 500 Hz using an 80486 IBMcompatible personal computer. Leakage was avoided by sampling a wide bandwidth (over 10 000 points). Unless otherwise specified, two sinusoidal cycles were obtained in a single scan, and 240 such scans were collected for one FIA measurement. Acquiring a larger number of scans increased the resolution at the lower frequencies, and a longer sampling time minimized artifacts due to convolution with the window function of the data. Background subtraction was performed continuously as follows. A background signal was acquired digitally prior to each FIA experiment and then converted back into an analog signal, which was subtracted from all subsequent current measurements prior to digitization of the instantaneous signal. This was done to minimize the low-frequency components associated with the background at a copper electrode, so as to increase the dynamic range for the measurement of the signal due to nucleotides. The time domain data acquired in this manner were converted into the frequency domain with Fourier transform methods using commercial software (MATLAB 4.2.c.1, The Mathworks, Inc., Englewood Cliffs, NJ). The protocol for analyzing frequency spectra has been described previously.33 Briefly, frequency spectra of the signal (only) were obtained simply by digital subtraction of a residual background vector from the instantaneous current vector. Time course data were obtained through the digital equivalent of lock-in amplification. Successive 512 point segments of the data were Fourier transformed sequentially into the frequency domain, generating the magnitude and phase information of each frequency element. Since all of the Faradaic information is contained within the harmonics of the excitation waveform,33,42-44 only these frequency elements were examined. Phase selectivity was obtained by taking the projection of the instantaneous current on the background-subtracted signal. Finally, the phase-resolved projections of each segment were low-pass filtered as a function of time. The time course information was generated after averaging 10 such projections together by using a moving average smooth (essentially moving boxcar integration) as a low-pass filter. RESULTS AND DISCUSSION Sinusoidal voltammetry at copper electrodes is capable of directly detecting and characterizing the response due to the direct oxidation of carbohydrates and nucleotides.19 Both purine- and pyrimidine-based nucleotides can be detected at copper surfaces as the sugar backbone is present universally on all the nucleotides. Previously, model nucleotides (AMP and CMP) were analyzed with sinusoidal voltammetry.19 The detection of nucleotides does not foul the electrode surface, and the detector can be used for continuous analysis of nucleotides in flow streams. The limits of detection for the adenine nucleotides are on the order of 100 nM, and it is possible to differentiate the adenine (purine base)(42) Cullison, J. K.; Kuhr, W. G. Electroanalysis 1996, 8, 314-319. (43) Smith, D. E. In Electroanalytical Chemistry; Bard, A. J., Ed.; Marcel Dekker: New York, 1966; Vol. 1, pp 1-155. (44) Pettit, F. R. Fourier Transforms in Action; Chartwell-Bratt Ltd.: Bromley, Kent, UK, 1985.
4830 Analytical Chemistry, Vol. 69, No. 23, December 1, 1997
Figure 1. Background-subtracted frequency spectra for different length oligonucleotides: (A) 1.1 µM CMP, (B) 109 nM 10-mer dC, (C) 9.1 nM 120-mer dC, and (D) 2.7 nM 400-mer poly-dC. Spectra shown correspond to signal obtained at time (t) ) 95 s in the FIA injections. All responses were optimized independently with respect to harmonic magnitude and phase angle.33 Sinusoidal excitation with 2 Hz sine wave, 0.05-0.55 Vpeak-peak vs Ag/AgCl; 60 s FIA injection; flow rate, 0.5 mL/min; 0.1 M NaOH running electrolyte.
containing nucleotide from cytosine (pyrimidine base)-containing nucleotide on the basis of the difference in the frequency domain response due to the two bases. As the number of sugar moieties present on the nucleotide increases with the length of the nucleotide, it should be possible to enhance the sensitivity of detection with increasing nucleotide length, while preserving the qualitative nature of the signal. Figure 1 shows the frequency spectra for four different size synthetic oligonucleotides, containing repeated units of the base cytosine. The spectra demonstrate that the qualitative nature of the frequency domain response is retained for nucleotides ranging from one individual base to nearly 400 bases in length. The similar distribution of signal at different harmonics indicates that the detection mechanism is not significantly affected by the increasing number of monomer units present in the sequential samples analyzed. The sample containing the longest oligomer (a 400mer of poly-dC) starts exhibiting sluggish kinetic effects, which leads to a decrease in the high-frequency components present in its spectrum with respect to those observed in the shorter oligonucleotides. However, it is important to note that the signal magnitude increases linearly as the length of the oligonucleotide increases. Figure 2 A demonstrates the dependence of signal in terms of its magnitude at various harmonics. The response at the first six harmonics is shown since most of the signal was found in these harmonics. The magnitude of the signal increases linearly at almost all the harmonics, although some saturation effect is seen for the poly-dC sample at the higher harmonics. For example, a regression line for the second harmonic is shown in Figure 2A, with the equation for the line, i(pA) ) 1.00x + 0.752, r ) 0.991 26, n ) 4, where x is the number of bases in the oligonucleotide. Since this is a log-log plot with a slope of 1.00
A
A
B
B
Figure 2. Dependence of magnitude (A) and phase angle (B) response on the length of the oligonucleotide. Response is shown only for the first six harmonic components, as they contain most of the signal measured for these analytes. The magnitude values plotted correspond to the maximum response observed at each harmonic for a nucleotide in an FIA experiment. Phase angles of these maximum responses for each nucleotide are plotted in the lower figure. All other FIA experimental conditions are as described in Figure 1.
over a range of 4 orders of magnitude, this indicates that the signal increases extremely linearly with the number of bases present on the oligonucleotide. Figure 2B highlights another trend observed in the qualitative aspect of the response for oligonucleotides of different lengths. The phase angles at the first six harmonics are plotted for the four nucleotides analyzed. Previously, Long and Weber have found a linear relationship (where E1/2 decreases as phase angle increases at a given harmonic for a given potential window) between the phase angle and the E1/2 of the redox couple.45 The phase angles observed for the oxidation of the DNA oligomers increase (more positive) with increasing chain length, indicating that the E1/2 of the oxidation reaction involving the longer DNA oligomers occurs at more negative potentials (i.e., less overpotential). This, in turn, would mean that the oxidation of longer DNA strands is occurring more easily, confirming the linear increase in current magnitude with increasing oligomer strand length observed in Figure 2A. This provides further evidence that the qualitative nature of the signal is preserved for different length oligonucleotides and that this information can be used to further qualitatively identify the oligonucleotide in terms of its length and, consequently, its molecular weight. This is an important aspect of this detection strategy, as not only is the sensitivity of the measurement improved as the length of the oligonucleotide increases, but also the oligonucleotide can possibly be character(45) Long, J. T.; Weber, S. G. Electroanalysis 1992, 4, 429.
Figure 3. (A) Background-subtracted frequency spectra for 1 nM, 9.5 kb ssDNA (0) and dsDNA (9). ssDNA was obtained by denaturing a fraction of the dsDNA sample. Frequency spectra shown correspond to signal obtained at time (t) ) 95 s. Again, all conditions were optimized independently with respect to their specific harmonic magnitude and phase angle.33 The magnitude of the electrochemical response is shown for each frequency component, and the phase angle of the signal is reported above each bar (top, ssDNA; bottom, dsDNA). (B) Time course for the FIA injection for 1 nM dsDNA monitored at the sixth harmonic, which was found to be the most sensitive for the detection of dsDNA. All other experimental conditions are the same as those given in Figure 1.
ized by the magnitude and phase angle of its response at the copper surface using sinusoidal voltammetry. Since DNA is an oligonucleotide which contains both purine and pyrimidine bases, it is also possible to detect underivatized DNA with this approach. An important feature of this scheme is that the sinusoidal voltammetric signal is proportional to the strand length (even with dsDNA), because the electroactive sugars lie on the outer perimeter of DNA and thereby will be available for detection at the copper surface. This facilitates the detection of large DNA strands, and the signal from a dsDNA is proportionately higher than that arising from the oxidation of ssDNA at a copper surface, as it capitalizes on the electroactivity of these easily available sugar moieties. This is in contrast to other electrochemical detection schemes at carbon and other surfaces, which are dependent only on the electroactivity of the purine bases, where the bases are shielded by the sugar and phosphate backbone present in a double helix for dsDNA. Figure 3A shows the sinusoidal voltammetry frequency spectra due to the oxidation of a 9.5 kb segment of ssDNA and its complemetary dsDNA at a copper electrode. The signal observed for the oxidation of the dsDNA is roughly twice the magnitude of that due to ssDNA derived by denaturation of the dsDNA, since the dsDNA has exactly twice the number of ribose sugar moieties (and bases) available for oxidation at the copper surface. The normalized current response for oligonucleotides (e.g., a response Analytical Chemistry, Vol. 69, No. 23, December 1, 1997
4831
of 81 fA/µM nucleotide was found for the 9.5 kb dsDNA fragment at the fifth harmonic) always exceeded the response observed for the individual nucleotide (e.g., 7 fA/µM nucleotide for CMP), indicating that the oxidation of the nucleotides in the oligomer is more efficient than that of the individual unit. Figure 3B shows the time course for the signal due to the oxidation of 1.0 nM dsDNA at the sixth harmonic, which was found to be optimal for detection of this analyte. The flow profile of the injection shows that there is no significant adsorption of the dsDNA at the copper surface. No loss in signal was observed in the performance of the electrode after repeated injections of dsDNA (rsd ) 5.6%, n ) 3), indicating that the DNA does not foul the electrode surface. This illustrates the reproducibility of the sensor for the detection of DNA. The concentration detection limit (S/N ) 3) for this size of dsDNA at the sixth harmonic is 3.2 pM. The mass detection limit for this experiment can be determined by approximating the sample volume. This can be obtained by taking the total analysis time for one time point (two cycles at 2 Hz, or 1 s) and the diffusion coefficent for a 9.5 kB DNA (1 × 10-6 cm2 s-1) and estimating the sampling distance using linear diffusion (l ) (Dt)1/2, or l ) 1.41 × 10-3 cm). Thus, the sample volume is the volume of the cylinder dictated by the area of the copper microelectrode and this length (V ) 2πr2l, where r ) 10 × 10-4cm; thus V ) 4.4 nL). Thus, the mass sensitivity of this technique is the concentration detection limit (3.2 pM) multiplied by the volume (4.4 nL), or 1.4 × 10-20 mol (or 8400 molecules) of DNA. Obviously, these mass detection limits can only be achieved if this corresponds to the injection volume of 4.4 nL. Since this is commonly available with capillary electrophoresis, it seems likely that these detection limits can be achieved when sinusoidal voltammetry is combined with capillary electrophoresis or other capillary separation techniques.
metry. The detection approach at copper is based on the electrocatalytic oxidation of amine-containing nucleobases and the ribose sugar-containing backbone of the nucleotides. The electrochemical response increases with the number of bases present on the oligonucleotide in a linear manner. The magnitude and the phase angle of the response can be used to qualitatively identify the oligonucleotide in terms of the number of bases present. The sensitivity for ssDNA and dsDNA is even better due to the larger number of sugars present in these macromolecules compared to a single nucleotide or smaller oligonucleotides. Ten kilobase segments of ssDNA and dsDNA are detected in the picomolar range. Sinusoidal voltammetry makes it possible to detect these big molecules with high sensitivity by preventing any fouling of the electrode surface and by effectively decoupling the faradaic signal from the large charging current background in the frequency domain. When coupled with a low-volume separation technique like capillary electrophoresis (with nanoliter sample volumes), this should allow detection of zeptomole (10-21 mol) quantities of DNA.
CONCLUSIONS The detection of oligonucleotides, ssDNA, and dsDNA is achieved at a copper electrode surface using sinusoidal voltam-
AC970731Q
4832
Analytical Chemistry, Vol. 69, No. 23, December 1, 1997
ACKNOWLEDGMENT The authors would like to thank Dr. Ashok Pullikuth in the Department of Entomology at UC Riverside for donating the ssDNA and the dsDNA samples. Dr. N. Dontha’s help with sample preparation and discussions is also appreciated. This work was supported by the National Institutes of Health (GM44112-01A1) and National Science Foundation (CHE-9414410).
Received for review July 8, 1997. Accepted September 15, 1997.X
X
Abstract published in Advance ACS Abstracts, October 15, 1997.
