Anal. Chem. 2010, 82, 8272–8277
Sensitive DNA-Based Electrochemical Strategy for Trace Bleomycin Detection Bin-Cheng Yin, Di Wu, and Bang-Ce Ye* Lab of Biosystems and Microanalysis, State Key Laboratory of Bioreactor Engineering, East China University of Science and Technology, Meilong Road 130, Shanghai, 200237, China Bleomycins (BLMs) are widely used in combination with chemotherapy for the treatment of a variety of cancers. The clinical application of BLMs is featured by the occurrence of sometimes fatal side effects, such as renal and lung toxicity, and the potential dose-limiting side effect of pulmonary fibrosis. Therefore, it is highly desirable to develop a sensitive method to quantitatively determine the BLM content in both pharmaceutical analysis and clinical samples, to make full use of therapeutic efficacy and to weaken its toxicity. Here, we proposed a simple, rapid, and convenient electrochemical assay for trace BLM detection. A reported DNA motif, as substrate for BLMs, is prepared to self-assemble onto the gold electrode to fabricate an electrochemical DNA (EDNA) sensor, with a terminus tethered on the electrode surface and the other terminus labeled with ferrocenyl moiety as a signal reporter to form a stem-loop structure, giving an arise of remarkable faradaic current. In the presence of Fe(II) · BLM, the E-DNA sensor undergoes the irreversible cleavage event, which can be transduced into a significant decrease in current peak. This proposed sensor reveals an impressive sensitivity as low as 100 pM BLMs and exhibits a good performance as well as in serum sample. Considering the high sensitivity and specificity of this proposed sensor, as well as the costeffective and simple-to-implement features of the electrochemical technique, we believe that this method shows distinct advantages over conventional methods and it is a promising alternative for the determination of trace amounts of BLMs in clinical samples. Bleomycins (BLMs) are a family of glycopeptide antibiotics produced by Streptomyces verticillus.1,2 The antitumor activity of BLMs is generally believed to be due to its chemotherapeutic effect through the selective cleavage of single-stranded (ss) or double-stranded (ds) DNAs3-6 and possibly also shape-selective * To whom correspondence should be addressed. Fax: (+)00862164252094. E-mail:
[email protected]. (1) Umezawa, H.; Maeda, K.; Takeuchi, T.; Okami, Y. J. Antibiot. (Tokyo) 1966, 19, 200–209. (2) Ishizuka, M.; Takayama, H.; Takeuchi, T.; Umezawa, H. J. Antibiot. (Tokyo) 1967, 20, 15–24. (3) Claussen, C. A.; Long, E. C. Chem. Rev. 1999, 99, 2797–2816. (4) Steighner, R. J.; Povirk, L. F. Proc. Natl. Acad. Sci. U.S.A. 1990, 87, 8350– 8354. (5) Povirk, L. F.; Han, Y. H.; Steighner, R. J. Biochemistry 1989, 28, 5808– 5814.
8272
Analytical Chemistry, Vol. 82, No. 19, October 1, 2010
cleavage of some RNAs.7-11 BLM-mediated DNA cleavage is via oxidation of the deoxyribose moiety in the presence of oxygen and a redox-active metal ion, such as Fe (II). In the presence of Fe (II) and oxygen, BLMs form a binary complex Fe(II) · BLM, subsequently oxidize Fe(II) to Fe (III), and reduce the oxygen to free radicals, exerting its antitumor effect on the cell death ultimately via DNA breaks induced by the free radicals. Mixtures of BLMs have been widely used in combination with chemotherapy for the clinical treatment of a variety of cancers, notably squamous cell carcinomas, germ cell tumors, Kaposi’s sarcoma, cervical cancers, and malignant lymphomas,3,13 with the advantages of low myelosuppression and low immunosuppression. However, they can cause bad renal and lung toxicity. In addition, the most feared and dose-limiting side effect of bleomycin is the potential for pulmonary fibrosis.14 In order to make full use of the antitumor drugs and to weaken their toxicity, it is a compelling need for sensitive detection of BLMs in both pharmaceutical analysis and clinical sample. In order to meet these objectives, a number of methods have been developed for BLM determination over the past three decade, such as high-performance liquid chromatography (HPLC),15-19 radioimmunoassay (RIA),20-22 enzyme immunoassay (EIA),23,24 microbiological assay,25,26 and fluorescence reso(6) Burger, R. M. Chem. Rev. 1998, 98, 1153–1170. (7) Carter, B. J.; de Vroom, E.; Long, E. C.; van der Marel, G. A.; van Boom, J. H.; Hecht, S. M. Proc. Natl. Acad. Sci. U.S.A. 1990, 87, 9373–9377. (8) Holmes, C. E.; Carter, B. J.; Hecht, S. M. Biochemistry 1993, 32, 4293– 4307. (9) Hecht, S. M. Bioconjug. Chem. 1994, 5, 513–526. (10) Morgan, M. A.; Hecht, S. M. Biochemistry 1994, 33, 10286–10293. (11) Holmes, C. E.; Duff, R. J.; van der Marel, G. A.; van Boom, J.; Hecht, S. M. Bioorg. Med. Chem. 1997, 5, 1235–1248. (12) Ma, Q.; Akiyama, Y.; Xu, Z.; Konishi, K.; Hecht, S. M. J. Am. Chem. Soc. 2009, 131, 2013–2022. (13) Bayer, R. A.; Gaynor, E. R.; Fisher, R. I. Semin. Oncol. 1992, 19, 46–52; discussion 52-53. (14) Liu, J.; Liu, Z.; Hu, X.; Kong, L.; Liu, S. Luminescence 2008, 23, 1–6. (15) Klett, R. P.; Chovan, J. P. J. Chromatogr. 1985, 337, 182–186. (16) Klett, R. P.; Chovan, J. P.; Danse, I. H. J. Chromatogr. 1984, 310, 361–371. (17) Aszalos, A.; Crawford, J.; Vollmer, P.; Kantor, N.; Alexander, T. J. Pharm. Sci. 1981, 70, 878–880. (18) Shiu, G. K.; Goehl, T. J. J. Chromatogr. 1980, 181, 127–131. (19) Shiu, G. K.; Goehl, T. J.; Pitlick, W. H. J. Pharm. Sci. 1979, 68, 232–234. (20) Broughton, A.; Strong, J. E. Cancer Res. 1976, 36, 1418–1421. (21) Crooke, S. T.; Luft, F.; Broughton, A.; Strong, J.; Casson, K.; Einhorn, L. Cancer 1977, 39, 1430–1434. (22) Teale, J. D.; Clough, J. M.; Marks, V. Br. J. Cancer 1977, 35, 822–827. (23) Fujiwara, K.; Isobe, M.; Saikusa, H.; Nakamura, H.; Kitagawa, T.; Takahashi, S. Cancer Treat Rep. 1983, 67, 363–369. (24) Fujiwara, K.; Yasuno, M.; Kitagawa, T. Cancer Res. 1981, 41, 4121–4126. (25) Onuma, T.; Holland, J. F.; Masuda, H.; Waligunda, J. A.; Goldberg, G. A. Cancer 1974, 33, 1230–1238. 10.1021/ac101761q 2010 American Chemical Society Published on Web 09/03/2010
nance energy transfer (FRET).12 However, these methods are generally time-consuming and/or labor-intensive and/or costexpensive and/or harmful to health. For example, RIA is a highly sensitive and specific method, but the radioactive isotope is harmful to the health and its half-life limits the effective life of the instruments. As for the HPLC assay, although it is a widely used method, it is expensive for the equipment, is labor intensive for the complex pretreatment, and has a time cost for the separation procedures. Thus, it is still highly desirable to develop a sensitive and selective method that can provide a simple, fast, and practical routine to determine trace amounts of BLMs in clinical analysis. In the recent years, the electrochemical DNA (E-DNA) sensor has attracted increasing attention owing to its remarkable features of high sensitivity, simple and inexpensive instrumentation, low fabrication cost, fast response, and portability.27-32 It is well-known that a typical E-DNA sensor consists of a gold electrode, a thiolgroup modified DNA probe designed for the targets, and an electrochemical analyzer. Upon recognition of target, the probe is induced to undergo a conformational alteration and this event is transduced to electrochemical signals that demonstrate the presence of the specific target. It has been employed as a popular platform for the applications such as clinical genetic analysis, environmental monitoring, and forensic identification. In this work, we have presented a single-step electrochemical BLM detection assay based on the destruction of a stem-loop structure of signal probes under the oxidative effect of BLMs with Fe(II) ion as a cofactor. This assay is designed such that, in the absence of BLMs and metal ion of Fe(II), the ferrocene (Fc)modified signal probe assumes a stem-loop structure that brings the ferrocenyl moiety in the close proximity to the gold electrode, producing a remarkable faradaic current. In the presence of Fe(II) · BLM, the activated metallobleomycin can mediate the oxidative signal probe destruction, which irreversibly cleaves the probe at the stem structure to release the Fc tag into the electrolyte solution, giving rise to a significant decrease of faradiac current. In this way, the amount of BLMs can be directly detected from the magnitude of current change. This method shows distinct advantages over conventional methods in terms of its potential sensitivity, specificity, low cost, and rapid response, making it a promising detection assay for use at the point of care. EXPERIMENTAL SECTION Chemicals and Reagents. Hexaammineruthenium(III) chloride ([Ru(NH3)6]3+, RuHex) and tris(2-carboxyethyl)phosphine hydrochloride (TCEP) were purchased from Alfa Aesar (Ward Hill, MA) and directly used as received. 6-Mercaptohexanol (MCH) and the fetal calf serum were purchased from Sigma(26) Pittillo, R. F.; Woolley, C.; Rice, L. S. Appl. Microbiol. 1971, 22, 564–566. (27) Fan, C.; Plaxco, K. W.; Heeger, A. J. Proc. Natl. Acad. Sci. U.S.A. 2003, 100, 9134–9137. (28) Xiao, Y.; Piorek, B. D.; Plaxco, K. W.; Heeger, A. J. J. Am. Chem. Soc. 2005, 127, 17990–17991. (29) Baker, B. R.; Lai, R. Y.; Wood, M. S.; Doctor, E. H.; Heeger, A. J.; Plaxco, K. W. J. Am. Chem. Soc. 2006, 128, 3138–3139. (30) Lai, R. Y.; Lagally, E. T.; Lee, S. H.; Soh, H. T.; Plaxco, K. W.; Heeger, A. J. Proc. Natl. Acad. Sci. U.S.A. 2006, 103, 4017–4021. (31) Xiao, Y.; Lubin, A. A.; Baker, B. R.; Plaxco, K. W.; Heeger, A. J. Proc. Natl. Acad. Sci. U.S.A. 2006, 103, 16677–16680. (32) Gorodetsky, A. A.; Ebrahim, A.; Barton, J. K. J. Am. Chem. Soc. 2008, 130, 2924–2925.
Aldrich (St. Louis, MO). Oligonucleotide probes (SH-ssDNA1: 5′-SH-(CH2)6-CGCTTTAAAAAAAGCG-(CH2)5-Fc-3′; SH-ssDNA2: 5′-SH-(CH2)6-CGCTTTAAAAAAAGCG-3′), with thiol group at 5′ terminus with a 6-carbon spacer or ferrocenyl moiety at 3′ terminus, were synthesized and purified by HPLC by Takara Biotechnology Co. Ltd. (Dalian, China). Bleomycins sulfate, in which the contents of A2 and B2 were up to 95%, were purchased from Melone Pharmaceutical Co., Ltd. (Dalian, China). The metal salts (MgCl2, CaCl2, Co(NO3)2, Ni(NO3)2, Cd(NO3)2, Cu(NO3)2, Pb(NO3)2, Zn(NO3)2, Mn(NO3)2, FeCl2 · 4H2O) and other reagents used in this work were of analytical grade and directly used without additional purification, purchased from Sinopharm Chemical Reagent Co. Ltd. (Shanghai, China). Unless otherwise noted, all buffers were prepared using distilled water purified by a Milli-Q water purification system (Millipore Corp., Bedford, MA) with an electrical resistance of 18.2 MΩ · cm. DNA immobilization buffer (I-buffer) was 10 mM PBS, 0.1 M NaCl, 0.1 M NaClO4, and 10 µM TCEP (pH 8.0). BLM detection buffer (D-buffer) was 10 mM PBS, 0.5 M NaCl, and 0.1 M NaClO4 (pH 8.0). The electrochemistry buffer in quantitation analysis (Q-buffer) of the thiolated ssDNA amount on the electrode surface was 10 mM Tis-HCl (pH 7.4) with 50 µM RuHex. Instrumentation. A CHI 830b electrochemical workstation (Chenhua Instrument Company, Shanghai, China) was employed for cyclic voltammetry (CV), chronocoulometry (CC), and square wave voltammogram (SWV) measurements. A IM6 Zennium electrochemical workstation (Zahner-elecktrik GmbH & Co., Germany) was employed for electrochemical impedance spectroscopy (EIS). The electrochemical workstations have a threeelectrode mode consisting of a gold working electrode, a platinum counter electrode, and a Ag/AgCl reference electrode (saturated with 3.0 M KCl), which was used for all electrochemical measurements at laboratory ambient room temperature (RT, 22-25 °C). E-DNA Sensor Preparation. The detailed E-DNA sensor fabrication procedures were reported in the literature.33,34 Briefly, gold electrode (2 mm in diameter, CH Instruments Inc.) was first polished sequentially with 0.3 and 0.05 µm alumina slurry for 5 min, followed by ultrasonic cleaning in ethanol and Milli-Q water for 5 min, respectively. Then, the electrode was washed thoroughly with Milli-Q water and dried in a nitrogen stream to obtain a clean gold surface. The cleanness of the electrode was checked by performing CV measurement in a fresh 0.5 M H2SO4 solution with the following parameters: potential range from -0.3 to 1.5 V versus Ag/AgCl and scan rate of 0.1 V/s. A typical single sharp reduction peak located at ∼0.88 V and multiple overlapping oxidation peaks in the range of 1.1-1.5 V were clearly observed (Figure S1, Supporting Information). Prior to immobilization onto the gold electrode, the thiolated ssDNA probe (1 µM) was dissolved in the I-buffer and incubated for 30 min at RT to reduce disulfide bonds. Then, an aliquot of 1 µM thiolated probe solution was dispensed on the clean gold surface in a dark place for 3 h at RT. After that, the electrode was washed thoroughly with Milli-Q water and dried in a nitrogen gas. Then, the electrode was incubated in the D-buffer containing 5 mM MCH for 2 h at RT to remove nonspecific DNA adsorption (33) Xiao, Y.; Lai, R. Y.; Plaxco, K. W. Nat. Protoc. 2007, 2, 2875–2880. (34) Zhang, J.; Song, S.; Wang, L.; Pan, D.; Fan, C. Nat. Protoc. 2007, 2, 2888– 2895.
Analytical Chemistry, Vol. 82, No. 19, October 1, 2010
8273
Scheme 1. Schematic Illustration of BLM Detection Using E-DNA Strategy via DNA Cleavagea
a A 16-nucleotide hairpin probe (SH-ssDNA1), capable of forming a stem-loop configuration, was designed to mimic the self-complementary structure. The Fe(II) · BLM-mediated cleavage of the stem structure of SH-ssDNA1 leads the concomitant Fc tag release away from the electrode, resulting in the significant faradaic current decrease.
on the gold surface. Subsequently, the electrode was rinsed thoroughly with Milli-Q water and dried with a stream of nitrogen gas again. The as-prepared electrode was incubated in the D-buffer for 1 h at RT to form the stem-loop structure before the BLM detection experiment. For the control experiments, the MCHmodified electrodes were fabricated as in the above procedures, just skipping the immobilization step of the thiolated ssDNA probe. Quantitation of ssDNA Probe on Gold Electrode. The surface density (the number of immobilized electroactive probe moles per unit area of the electrode surface) was quantified by chronocoulometric measurement according to a reported method.35 The determination of surface density of thiolated ssDNA probe immobilized on the gold is based on the assumption that redox active RuHex cation associates with anionic phosphate backbone of DNA. The charge corresponding to RuHex electrostatically bound to surface-confined ssDNA (Qss) can be calculated from the following equation: Qss ) Qtotal- Qdl, where Qtotal strands for the total charge flowing through the electrode, comprising both faradaic (redox) charges and nonfaradaic (capacitive) charges (Qdl). The amount of electroactive probe on the electrode surface (Γ) is calculated on the basis of the equation Γss ) (QssNA/nFA)/(z/m), where n is the number of electrons transferred in the reaction (n ) 1), F represents the Faraday constant (coulombs per equivalent), A is the effective surface area of gold electrode (square centimeters), m is the number of nucleotides in the DNA, z is the charge of the redox molecules, and NA is Avogadro’s number. In this work, chronocoulometric quantitation of DNA surface density was carried out in Q-buffer using SH-ssDNA2 as the model. CC measurement was carried out with an initial potential of +0.1 V and a final potential of -0.4 V. Electrochemical Detection of BLMs. The electrochemical measurements for BLM detection were performed in a series of single-compartment cells with 10 mL volume. The fabricated electrodes were incubated in a single-compartment cell filled with 8 mL of D-buffer for 1 h to allow the formation of a stem-loop structure. The BLM samples were prepared by mixing BLMs with Fe(II) ion in 1:1 molar ratio. Then, a series of a 10 µL Fe(II) · BLM solution with various concentrations were added to these singlecompartment cells with electrodes, respectively. These electrodes were incubated for 30 min, and then, the electrochemical measurements were carried out using CV, EIS, and SWV measurements. CV measurement was performed at a scan rate of 0.1 V/s between -0.1 and 0.5 V. EIS measurement was performed as follows: a frequency range of 4 kHz to 10 mHz, a bias potential (35) Steel, A. B.; Herne, T. M.; Tarlov, M. J. Anal. Chem. 1998, 70, 4670–4677.
8274
Analytical Chemistry, Vol. 82, No. 19, October 1, 2010
of +0.175 V, and an alternative voltage of 5 mV. SWV measurement was performed with the following parameters: initial potential of 0 V, final potential of 0.5 V, and frequency of 25 Hz. The timecourse experiments for the kinetic response upon Fe(II) · BLMmediated DNA cleavage were also investigated by monitoring the faradaic current change at each time point using SWV measurement. Unless otherwise noted, each electrochemical measurement was repeated at least three times with different electrodes of similar probe density under the same conditions. RESULTS AND DISCUSSION We have designed a simple E-DNA sensor for trace BLM detection based on the highly specific, metal-induced activation of BLMs, which is presented schematically in Scheme 1. The sensor contains a reported 16-nucleotide hairpin probe12,36-38 (SHssDNA1) with a thiol group at the 5′ end as anchor on a gold electrode surface and a ferrocenyl moiety at the 3′ end as the reporter. In this work, we chose Fc as an electroactive tag due to the good reversibility of its redox reaction and simplicity of the synthesis. This probe can be self-assembled onto the gold electrode surface via a thiol-gold linkage and self-hybridizes into a stem-loop structure with a 6-base-pair (bp) stem and 4-nt loop with Fc in the close proximity into the gold surface. As shown in Scheme 1, a significant voltammetric signal, arising from electron transfer between Fc redox tag and the gold electrode surface, is observed in the absence of Fe(II) · BLM. This observation is attributed to a one-step redox reaction of a surface-confined ferrocenyl moiety via a stem-loop structure. It is reported that BLMs binding with Fe(II) ion can exhibit sequence DNA strand scission, predominantly at 5′-GC/T-3′ site sequences. When treated with BLMs in the presence of Fe(II) and oxygen, the probe is found to undergo oxidative transformation to be selectively cut off to release the Fc tag. These cleaved fragments dissociate into the bulk solution, allowing the ferrocenyl moiety away from the gold electrode surface, resulting in a significant faradaic current decrease. Thus, this cleave event is translated to a measurable faradaic current decrease, which can be proportional to the concentration of BLMs. Prior to BLM detection, the fundamental study on the electrochemical characterization of prepared E-DNA sensor was performed. First, the experiment of the CV currents versus (36) Hashimoto, S.; Wang, B.; Hecht, S. M. J. Am. Chem. Soc. 2001, 123, 7437– 7438. (37) Akiyama, Y.; Ma, Q.; Edgar, E.; Laikhter, A.; Hecht, S. M. Org. Lett. 2008, 10, 2127–2130. (38) Van Atta, R. B.; Long, E. C.; Hecht, S. M.; van der Marel, G. A.; van Boom, J. H. J. Am. Chem. Soc. 1989, 111, 2722–2724.
Figure 1. (A) CV curves for the SH-DNA1 modified electrode before (red line) and after reaction with 100 nM (blue line) and 1 µM (black line) Fe(II) · BLM in the D-buffer. (B) SWV curves for the SH-DNA1 modified electrode before (red line) and after reaction with 100 nM (blue line) and 1 µM (black line) Fe(II) · BLM in the D-buffer.
different scan rates was carried out to investigate the controlled factor of the electrochemical process on the electrode surface in the absence or presence of Fe(II) · BLMs. As shown in Figure S2 (Supporting Information), the formal potential Eo of ferrocene redox was found to be 215 mV, and peak currents of the sensor were in good linear correlation to the scan rate in the range of 20-140 mV/s in the absence or presence of Fe(II) · BLMs. These results clearly demonstrated that the typical surfacebound electrochemical processes were obtained, in which Fc tags were well confined to the electrode surface by a stemloop structure, and also manifested the successful immobilization of thiolated ssDNA probe and efficient electron transfer. In addition, we also investigated the surface density of the prepared E-DNA sensor at fixed probe concentration. It is reported that the surface density of the probe at the electrode surface is an important parameter in affecting the E-DNA signaling and equilibration time,39,40 which is quantitatively determined on the basis of the electrostatic binding of the redox active RuHex cation to the anionic phosphate backbones of DNA by measuring charge passed during the reduction of RuHex cation. In order to eliminate the redox active effect from the ferrocenyl moiety of SH-ssDNA1, a probe of SH-ssDNA2 with the same nucleotide sequence as SHssDNA1 was employed as a model to obtain a signal response for SH-ssDNA2 modified electrode using CC measurement (Figure S3, Supporting Information). To allow both reproducible and quantitative determination of the amount of BLMs, it is important to maintain the similar probe density of the electrode. According to the reported data,41 the surface density can be well controlled by changing the immobilization probe concentration during the fabrication process, whereas above a certain probe concentration, the probe is saturated on the surface to give a best performance with an optimal signal gain. Consistent with this result, we observed that the signal gain reaches a plateau when 1 µM signaling probe was used to prepare the electrode (data not shown) and the experimental data indicated that the optimal surface density was reproducibly achieved around ∼6.36 × 1012 strand/cm2 (RSD < 5%, repeated six times with different electrodes), calculated according to equations as described in the Experimental Section. To explore the feasibility of this proposed signal-off E-DNA sensor, CV measurement was employed. In the absence of (39) Jayaraman, A.; Hall, C. K.; Genzer, J. J. Chem. Phys. 2007, 127, 144912– 144922. (40) Markham, N. R.; Zuker, M. Nucleic Acids Res. 2005, 33, W577–W588. (41) Ricci, F.; Lai, R. Y.; Heeger, A. J.; Plaxco, K. W.; Sumner, J. J. Langmuir 2007, 23, 6827–6834.
Fe(II) · BLM, a pair of a typical redox peak at 0.203 and 0.250 V of ferrocenyl moiety was observed in CV as shown in the red line in Figure 1A, indicating the stem-loop formation forces the redox reporter of Fc-labeled probe in close proximity to the electrode to generate the efficiency of electron transfer. After adding 100 nM Fe(II) · BLM in this system, substantially diminished redox peaks were observed (blue line). Adding Fe(II) · BLM at a higher concentration of 1 µM, the couple of redox peaks almost disappeared. These observations clearly indicated the decrease of electron transfer between the Fc redox moiety and gold electrode, due to the destruction of stem-loop structure, thereby releasing the Fc tag away from the electrode. That also implied the sensor was sensitively responsive to Fe(II) · BLM. It has been shown experimentally that SWV measurement can provide excellent resolution of the response with much higher sensitivity compared to the conventional sweep techniques such as CV; thus, we employed SWV measurement to further characterize the electron transfer, as shown in Figure 1B. It was observed that the proposed E-DNA sensor gave a remarkable reduction peak at 0.250 V vs SWV at the D-buffer in the absence of Fe(II) · BLM, and the peak current decreased from 1.260 to 0.548 µA (up to 56.5%) in response to 100 nM Fe(II) · BLM to 0.286 µA (up to 77.3%) in response to 1 µM Fe(II) · BLM. Moreover, a further control experiment upon the effect of Fe(II) ion and BLMs alone in the system was also investigated. We incubated the sensor in the D-buffer with Fe(II) ion alone. The experiment result indicated that the introduction of Fe(II) had a negligible effect on the redox peaks even in high concentration of 10 µM. Next, we challenged the sensor in the D-buffer with a series of only BLM solution with different concentrations, and no significant change in the E-DNA signal was observed. The above results give immediate evidence for the nucleic acid degradation by BLMs, in which BLMs critically depend on the metal ion binding to exhibit the effect of oxidative transformation of nucleic acid; otherwise, the BLMs alone have no oxidative effect on the nucleic acids. The sensitivity of the sensor under the optimized detection condition was investigated by SWV measurement. Figure 2A gives the SWV current responses of the sensor to Fe(II) · BLM at different concentrations. It clearly demonstrates that the addition of Fe(II) · BLM at different concentrations to the sensing system induced different decreases in peak current, associated with the remaining quantity of Fc redox moieties in close proximity to the gold electrode. As the concentrations of Fe(II) · BLM were increased, the resulting reduction currents in the SWV were Analytical Chemistry, Vol. 82, No. 19, October 1, 2010
8275
Figure 2. (A) SWV curves corresponding to the detection of different concentrations of BLMs, ranging from 0 to 1000 nM, in the D-buffer. The dash line presents the SWV curve for a MCH-modified electrode in the absence of BLMs. The voltammograms shown are averages of three different electrodes with similar probe density under the same experimental condition. (B) Plot of peak current as a function of the increasing concentrations of BLMs from 0 to 1 µM using the data of (A). Inset is the linear fit plot of peak current as a function of the logarithm of BLM concentrations from 100 pM to 1 µM. The error bars represent the standard deviation of three independent measurements as the data of (A).
Figure 3. Study of the effect of different metal ion coordination environments for BLM-mediated DNA cleavage. The error bars represent the standard deviation of three independent measurements at each metal ion with equal concentration of the BLMs.
intensified. Figure 2B appears that dynamically decreased SWV peak currents increased along with the increasing Fe(II) · BLM concentration ranging from 0 to 1 µM. An inset in Figure 2B exhibits a linear correlation to the logarithm of Fe(II) · BLM concentration across the four decade ranging from 100 pM to 1 µM with a linear correlation (R2) 0.9925). The experimental data showed that the proposed sensor is sensitively and specifically responsive to Fe(II) · BLMs. The directly measured detection limit is as low as 100 pM (0.293 ng/mL). To the best of our knowledge, this proposed method is superior in sensitivity and has wide linear range for BLM detection, compared to 8 ng/ mL of BLM sulfate by RIA,22 20 ng/mL of BLMA2 by HPLC,19 30 ng/mL of BLMA5 and 40 ng/mL of BLMA2 by fluorescence quenching.14 It is well-known that the BLMs require a reduced transition metal such as Fe(II) and oxygen to catalyze DNA lesions, which are responsible for tumor necrosis. We also carried out the experiment to study the effect of different metal ion coordination environments for BLM-mediated DNA cleavage. The BLM samples were prepared by mixing 1 µM BLMs with different metal ions in a 1:1 molar ratio. Figure 3 shows the SH-ssDNA1 modified electrodes upon treatment with a mixture of metal ion (Fe2+, Cu2+, Zn2+, Ca2+, Mg2+, Pb2+, Mn2+, Ni2+, Co2+, or Cd2+) and BLMs, respectively. Three independent measurements with different electrodes (having similar probe density) were carried out at each mixture containing BLMs and a metal ion under the same experimental condition, respectively. As expected, the data 8276
Analytical Chemistry, Vol. 82, No. 19, October 1, 2010
Figure 4. Time-course experiment for monitoring the kinetics of Fe(II) · BLM-mediated DNA cleavage adding Fe(II) · BLMs with different concentrations.
suggests that the combination of Fe(II) · BLM gave the best performance of a significant reduction of faradaic current (up to 76.2%). The time-course of oxidative cleavage of surface-confined SHssDNA1 probe in the presence of Fe(II) · BLM was also investigated. The decrease in current from Fe(II) · BLM-mediated DNA cleavage was monitored by SWV measurements. SWV measurements were performed after 0, 3, 5, 7, 10, 15, 20, 25, and 30 min of oxidation reaction. At each time point, the sensor was maintained in the D-buffer for direct detection. As shown in Figure 4, when treated with Fe(II) · BLM, the current was observed to be rapidly reduced within 10 min and then achieved equilibrium within 30 min, resulting in a significant decrease in faradiac current. The sensor equilibrium can also be achieved within 30 min as low as 100 pM Fe(II) · BLM (data not shown). To test the generality of our proposed sensor in the clinical sample, we challenged this E-DNA sensor in a complex sample matrix of 50% serum diluted 1:1 with the D-buffer. These measurements were conducted either in the simple D-buffer or in the 50% serum. We observe a similar initial signal increase when the sensor is introduced to the 50% serum, compared to the peak current observed in the D-buffer under the same experimental condition (Figure 5). In addition, a very small peak potential shift in the Fc redox potential was observed in the serum, it was ascribed to the minor differences in the pH of reaction buffer. More importantly, we observed a signal reduction from 1.26 to 0.48 µA (nearly 61.9% decrease) in 50% serum in the presence of 250 nM Fe(II) · BLM, which had similar current change (from 1.26
Figure 5. Generality of the E-DNA sensor in a complex sample matrix of 50% serum. SWV curves are shown in 50% (green curve), the D-buffer (red curve), and 50% serum with 250 nM Fe(II) · BLM (blue curve), respectively. The voltammograms shown are averages of three different electrodes with similar probe density under the same experimental conditions.
to 0.47 µA) in the D-buffer as shown in Figure 2, suggesting the strong stability and universality of the E-DNA sensor remains, independent of the reaction medium. Thus, this sensing platform performs well, to allow being well-suited for clinical applications. CONCLUSIONS We have demonstrated a simple, rapid, and convenient electrochemical detection of BLMs with high sensitivity and selectivity, based upon Fe(II) · BLM-mediated DNA strand scission. A reported motif was employed as substrate probe for BLMs to self-assemble onto the gold electrode, with a terminus tethered
on the electrode surface and the other terminus labeled with ferrocenyl moiety as signal reporter in stem-loop format. In the presence of Fe(II) · BLM, the E-DNA sensor undergoes the irreversible cleavage event, which can be transduced into a measurable current decrease. The sensor equilibrium is relatively rapid, and BLMs can be specifically detected within 30 min. Importantly, the proposed sensor achieves an excellent sensitivity as low as picomolar detection limits (100 pM), and it is unaffected from nonspecific contaminants, as demonstrated by the investigation of BLMs in serum. On the basis of the findings and results, we believe that this method shows distinct advantages over conventional methods in terms of high specificity, low detection limit, and simple-to-implement procedure and it could offer an interesting alternative approach for the determination of trace amounts of BLMs in clinical samples. ACKNOWLEDGMENT This work was supported by NSF (20776039, 21075040), Shanghai Project 09JC1404100, SKLBE Fund 2060204, and the Fundamental Research Funds for the Central Universities. SUPPORTING INFORMATION AVAILABLE Additional information as noted in text. This material is available free of charge via the Internet at http://pubs.acs.org. Received for review July 3, 2010. Accepted August 25, 2010. AC101761Q
Analytical Chemistry, Vol. 82, No. 19, October 1, 2010
8277
