Determination of DNA Methylation Using Electrochemiluminescence

Jan 22, 2012 - ... Immunosensor for Ultrasensitive Detection of Pseudorabies Virus Antibody .... Leila Syedmoradi , Fariba Esmaeili , Michael L. Norto...
1 downloads 0 Views 2MB Size
Download PDF
Letter pubs.acs.org/ac

Determination of DNA Methylation Using Electrochemiluminescence with Surface Accumulable Coreactant Ryoji Kurita,*,† Kumi Arai,† Kohei Nakamoto,†,‡ Dai Kato,† and Osamu Niwa†,‡ †

National Institute of Advanced Industrial Science and Technology (AIST), Tsukuba Central 6, 1-1-1 Higashi, Tsukuba, Ibaraki, Japan 305-8566 ‡ Graduate School of Pure and Applied Science, University of Tsukuba, 1-1-1 Tennoudai, Tsukuba, Ibaraki, Japan 305-8573 ABSTRACT: Cytosine methylation in DNA was determined by an enzyme linked immunosorbent assay (ELISA) with electrochemiluminescence (ECL) detection and employed for the DNA methylation assay of a long and real genomic sample for the first time. The developed method employed an antimethyl cytosine antibody labeled with acetylcholinesterase, which was added to recognize single methylated cytosine in a DNA oligomer. The acetylcholinesterase converted acetylthiocholine (substrate) to thiocholine (product), which was accumulated on a gold electrode surface via gold−thiol binding. This surface accumulated preconcentration made it possible to observe bright and distinctive ECL by applying a potential to the gold electrode in the presence of a tris(2,2-bipyridyl)ruthenium complex luminophore when the analyte DNA contained a methylation region. Methyl-cytosine was measured quantitatively in the 1−100 pmol range, which exhibits sufficiently high sensitivity to achieve real DNA measurements without amplification by a polymerase chain reaction (PCR). The proposed ECL method also exhibited high selectivity for methyl-cytosine against nonmethylated cytosine, guanine, thymine, and adenine nucleotides. Finally, original and methylated DNA samples were clearly distinguished with our method using a real DNA bacteriophage sample (48 502 base pairs).

D

an osmium complex.8,9 Unfortunately, thymine is also labeled with the fluorescent dye. This complicates the methyl-cytosine assay because a fluorescence resonance energy transfer technique is needed to distinguish methyl-cytosine from thymine. Some researchers have reported a noncleavage assay with an antimethyl cytosine antibody labeled with fluorescent dye. However, to obtain high sensitivity, they employed a huge and expensive excitation light source such as a Nd:YAG or He− Ne laser and combined this with capillary electrophoresis.10,11 We also proposed the electrochemical detection of methylcytosine by utilizing the oxidation potential difference between cytosine and methylated cytosine.12 This method is rapid and simple without any labeling; however, the detection limit should be improved for the practical detection of real DNA samples. Electrogenerated chemiluminescence (ECL) has attracted great interest over the past 2 decades in relation to converting electrical energy into radiative energy13−16 for a wide range of applications including biological analysis. A wide variety of tris(2,2′-bipyridyl)ruthenium(II) (Ru(bpy)32+) coreactants have been investigated to obtain a bright emission.15,17 Recently, our group proposed a way of realizing bright ECL by using

NA methylation is a well-known epigenetic modification mechanism that regulates gene expression and plays crucial roles in embryonic development.1 Cytosine methylation in CpG islands has received particular attention because it is thought to be involved in controlling genetic expression, including that in cancer,2 genomic imprinting,3 cellular differentiation, and Alzheimer’s disease.4 5-Methyl-cytosine is now recognized as the fifth DNA base containing heritable information. Therefore, highly sensitive, accurate, and quantitative information concerning cytosine methylation in DNA would be valuable with respect to genetic disease diagnosis. Two major cytosine methylation assay methods have been reported. One is hydrolysis and sequencing with a bisulfite salt,5,6 and the other is a cleavage assay with methyl-cytosine sensitive (or insensitive) restriction enzymes.7 A bisulfite based determination method is very widely used to distinguish between cytosine and methyl-cytosine. Treatment with bisulfite converts cytosine to uracil, while methyl-cytosine remains unaffected. Therefore, information about methyl-cytosine in DNA can be obtained by combining bisulfite treatment and a polymerase chain reaction (PCR). A restriction enzyme, which selectively catalyzes the scission of a specific sequence containing methyl-cytosine, can also be used to determine methylation in DNA. However, a DNA sample is damaged by the strand scission that occurs during the enzyme reaction. Recently, a noncleavage assay has been reported that was realized by labeling methyl-cytosine with a fluorescent dye via © 2012 American Chemical Society

Received: October 11, 2011 Accepted: January 21, 2012 Published: January 22, 2012 1799

dx.doi.org/10.1021/ac202692f | Anal. Chem. 2012, 84, 1799−1803

Analytical Chemistry

Letter

Figure 1. Schematic diagram of the principle of methyl-cytosine determination with the ECL method.

Methyl-Cytosine Analysis by ECL. Figure 1 is a schematic diagram of the proposed technique for methyl-cytosine analysis by ECL. The detailed protocol is as follows. To immobilize a real DNA sample, the DNA was annealed at 95 °C for 10 min and then a 250 ng sample was diluted with a commercial DNA coating solution (Pierce, Reacti-Bind, 17250) and immobilized on a 96-well microtiter plate (Nunc, 269620) by incubating it at room temperature for 2 h as previously reported.26 To immobilize a short synthesized oligomer, a 5′ amino-modified oligomer was immobilized on a microtiter plate with an active ester surface (Sumitomo Bakelite, BS-61602) by incubation at 80 °C for 2 h. After rinsing each well twice with 300 μL of washing buffer, the plate was coated with blocking buffer. Next, 50 μL of 500 ng/mL antimethyl cytosine antibody (Aviva Systems Biology) was added to each well and incubated at 37 °C for 1 h. After rinsing each well, 50 μL of 200 ng/mL biotinylated secondary antibody (Abcam) was added and incubated at 37 °C for 1 h. Then, 50 μL of avidinacetylcholinesterase conjugate (Cayman) was added and incubated for 2 h at room temperature. Next, 100 μL of 1 mM acetylthiocholine (Sigma), which was the substrate of the labeled enzyme, was added and incubated for 1 h at room temperature. The acetylcholinesterase reaction was halted by adding 100 μL of pH 2.0 glycine-HCl buffer. The fluids that constituted the assay results were diluted twice with 10 mM of phosphate buffer (pH 7.0, 0.53 g/L NaH2PO4, 0.867 g/L Na2HPO4), and then the thiocholine molecules in the diluted fluids were collected on a gold electrode (3 mm diameter circle) in our original electrochemical flow cell19 by injecting the diluted fluids for 10 min at a flow rate of 20 μL/min. Under this condition, about 36% of the injected thiocholine was adsorbed on the electrode, as confirmed in a previous study.19 Before the ECL measurement, the electrode was polished with 1 μm alumina slurry and then rinsed with pure water. Finally, 0.1 M of phosphate buffer (pH 7.0, 5.3 g/L NH2PO4, 8.67 g/L Na2HPO4) containing 2 mM Ru(bpy)32+ (Sigma) was injected into the flow cell, and the ECL intensity was measured with a photon counter (Hamamatsu Photonics) while the electrode potential was scanned from 0 to 2 V vs Ag/AgCl.

thiocholine (trimethylammonio ethanethiol) as a coreactant with Ru(bpy)32+,18 and it was applied to an ECL based sandwich immunoassay.19 Because thiocholine is a bifunctional molecule exhibiting both the ECL acceleration effect and surface accumulation via gold−thiol binding, the strongest emission yet reported appears in the low concentration region, which is particularly important for immunoassay applications designed to detect trace level analytes. Several researchers have focused on DNA determination by ECL, and they have mainly employed the Ru(bpy)32+ and tripropylamine system.20,21 For examples, DNA hybridization has been measured with a luminophore labeled probe DNA.22,23 Intercalations of aromatic structures in ruthenium complexes into DNA have been measured as an ECL emission.24,25 However, there has been no report regarding the determination of methyl-cytosine by ECL. This is because it is impossible to realize selective ECL emission derived solely from methyl-cytosine using conventional intercalation or hybridization techniques. In this paper, the determination of methyl-cytosine in DNA by using ECL is reported for the first time. The highly sensitive, selective, and quantitative detection of methyl-cytosine is achieved by measuring the ECL emission using the reaction of Ru(bpy)32+ and thiocholine, which is formed by an enzyme labeled antimethyl cytosine antibody. We employed our method to detect methyl-cytosine in a real DNA sample and succeeded in achieving a bisulfite free and PCR free determination.

■

MATERIALS AND METHODS

Reagents. DNA oligonucleotides were synthesized and purified with high-performance liquid chromatography by Hokkaido System Science. The DNA sequences used to obtain the calibration curve and selectivity were 5′-AAA AAA GYG AAA AAA-3′ (Y = thymine or cytosine or methyl-cytosine). The DNA of the bacteriophage (48 502 base pairs, MW is approximately 32 300 kDa) was obtained from Takara Bio Inc. A total of 10 μg of DNA was incubated with 12 units of CpG methyltransferase (New England Laboratories) in the presence of S- adenosylmethionine as a methyl donor at 37 °C for 60 min. Two buffer solutions were prepared using pH 7.4 phosphate buffered saline (PBS) (0.2 g/L KCl, 0.2 g/L KH2PO4, 8 g/L NaCl, 1.15 g/L Na2HPO4) for the assay. One was PBS containing 0.05 w/v % Tween 20 as a washing buffer, and the other was PBS containing 0.05 w/v% Tween 20 and 0.1 w/w% (bovine serum albumin) BSA as an incubation buffer.

■

RESULTS AND DISCUSSION Calibration Curve for Methyl-Cytosine. First, variations in the ECL emission were observed when amount of DNA containing a single methylated region changed. 15-mer DNA was used with the sequence AAAAAAG(mC)GAAAAAA-3′, mC = methyl-cytosine). 1800

the the was (5′-

dx.doi.org/10.1021/ac202692f | Anal. Chem. 2012, 84, 1799−1803

Analytical Chemistry

Letter

thiocholine preconcentration time would be useful because the ECL intensity is linear with the preconcentration time until the gold surface is fully covered. The detection limit, which is defined as 3 times the standard deviation of the negative control, was 0.18 pmol (1.8 nM). Previously, laser-induced fluorescence polarization detection combined with capillary electrophoresis (CE−LIFP),11 which is known to be a highly sensitive analytical method, was needed for measurements in this trace level methyl-cytosine range. The proposed method with a detection limit close to that of CE-LIFP is promising as a bisulfite free and PCR free measurement method. Moreover, our ECL based assay has the inherent advantages of high costperformance, low power and ease of miniaturization of the entire system. Selectivity to Methylated Cytosine. The selectivity of the methyl-cytosine assay was confirmed for other nucleotides by comparing the ECL signals from three kinds of DNA as shown in Figure 3a. The Y in the sequence in Figure 3a

Figure 2a shows the typical variations in ECL intensity when the potential of the gold electrode was scanned from 0 to 2 V in

Figure 2. (a) Variations in ECL intensity when measuring 0, 1, 10, 50, 100 pmol of DNA. The entire ECL emission was obtained in 2 mM Ru(bpy)32+ in 0.1 M phosphate buffer (pH 7.0) without an optical filter. (b) Calibration curve for methyl-cytosine containing DNA estimated from ECL peak intensity at 1.15 V in Figure 2 (a). The inset shows an expanded calibration curve in a low concentration range.

2 mM Ru(bpy)32+. The ECL intensity rapidly increased at 1.15 V and then gradually returned to the background level. This is because reductive radicals are produced from thiocholine at 1.15 V by the deprotonation of its alkyl chain,18 and then the radicals react with Ru(bpy)33+ as in previously described aminebased coreactants.27−29 Figure 2b shows a calibration curve for methyl-cytosine containing DNA oligomer. The Y-axis shows the peak ECL intensity at 1.15 V estimated from Figure 2a. Clear increases in ECL intensity were observed as the amount of methyl-cytosine increased. This is because the amount of thiocholine production increased due to the increase in enzyme (acetylcholinesterase) labeled anti methyl-cytosine antibody, and the thiocholine and Ru(bpy)32+ reacted to produce an emission when a potential was applied. An ECL emission with surface accumulable thiocholine is considered to be advantageous for trace level monitoring because the ECL emission only occurs at the electrode surface. Moreover, the number of thiocholine molecules can be increased by the enzymatic reaction. Therefore, methyl-cytosine was measured with high sensitivity in the 1−100 pmol range. Note that ECL intensity saturation was observed when the enzymatic product (thiocholine) concentration exceeded 10 μM for 10 min preconcentration because the gold electrode was completely covered with thiocholine. Therefore, the thiocholine concentration in the assay results should be less than 10 μM if we reduce the incubation time of the labeling enzyme or dilute the solution before the ECL measurement. In addition, tuning the

Figure 3. (a) Target sequences for investigating selectivity. Y means thymine or cytosine or methyl-cytosine. (b) Variations in ECL intensity on a gold electrode as the electrode potential increased when the three kinds of 100 pmol of DNA in part a were measured. (c) Selectivity for other nucleotides. Other experimental conditions were the same as in Figure 2.

indicates thymine or cytosine or methyl-cytosine. Figure 3b shows typical variations in ECL intensity when the three kinds of DNA were measured. An increase in the ECL emission at 1.15 V was clearly observed when only methyl-cytosine containing DNA was measured. In contrast, the ECL intensity was low when the DNA contained no methyl-cytosine. Osmium complex labeling for electrochemical detection or 1801

dx.doi.org/10.1021/ac202692f | Anal. Chem. 2012, 84, 1799−1803

Analytical Chemistry

Letter

discrepancy was observed between the calculated and experimental results. This would be because the ECL intensity has a sigmoidal rather than a linear relationship to the amount of thiocholine formed from methyl-cytosine. The real sample was relatively large, which initiated the saturation of the ECL intensity. Therefore, the methyl-cytosine amount estimated from the ECL intensity would be inaccurate in such concentration ranges. Dilution to a level suitable for ECL measurement is required for a more quantitative analysis. However, we could obtain a 4 times larger ECL with a methylated sample than with an unmodified real sample without bisulfite and PCR reactions. This demonstration proves the applicability of the ECL based methylation assay for determining methylation in genomic DNA.

fluorescent labeling via the osmium complex have been reported as methyl-cytosine labeling methods.8,30,31 However, these methods have no selectivity against thymine because the C5−C6 bond in thymine is also oxidized into thymine glycol.31 The result in Figure 3b clearly shows that the assay has selectivity for both cytosine and thymine. Obviously, our assay also has selectivity against adenine and guanine, since the DNA sequences contained both adenine and guanine. In other words, our assay using ECL has a high selectivity for methyl-cytosine against all four bases in DNA. Weak and broad ECL was observed at potentials above 1.1 V even when a methyl-cytosine free DNA and a negative control were measured as shown in Figure 3b. This emission is known to be derived from the reaction of Ru(bpy)32+ with dissolved oxygen or other impurities.32 However, it is easy to determine whether or not the ECL signal involves methyl-cytosine because a sharp and distinctive ECL peak appears at 1.15 V. Figure 3c shows the normalized ECL intensities at 1.15 V for each DNA sequence. By measuring the ECL intensity at a potential of 1.15 V, ECL emissions other than methyl-cytosine are negligible because the ECL emission for DNA containing no methyl-cytosine is the same as that of a negative control. Methyl-Cytosine Determination in Real DNA Sample. It is important to confirm whether our method can be applied to a real DNA sample with a much longer sequence. Two DNA samples were prepared, one was as is (Unmodified) and the other was incubated with methyltransferase (Methylated). Figure 4 shows the results when measuring 250 ng of real

■

CONCLUSIONS A new ECL based methyl-cytosine determination technique has been developed by using an antimethyl cytosine antibody labeled with acetylcholinesterase. The ECL emission derived from methyl-cytosine was successfully observed without any effect from adenine, thymine, guanine, or cytosine. We achieved bisulfite free and PCR free determination in a real DNA sample. The advantages of our method are that it provides simple and scission free analysis because it does not require bisulfite conversion, PCR amplification, DNA digestion, or separation using chromatography or electrophoresis techniques.

■

AUTHOR INFORMATION

Corresponding Author

*Phone: +81-29-861-6158. Fax: +81-29-861-6177. E-mail: r. [email protected]. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS Part of this study was supported by NEDO Japan and the Grant-in-Aid for Young Scientists (B), Grant No. 23710152. We also thank Mr. Meacock for revising the language of the manuscript.

■

REFERENCES

(1) Yoder, J. A.; Walsh, C. P.; Bestor, T. H. Trends Genet. 1997, 13, 335−340. (2) Jones, P. A. Cancer Res. 1996, 56, 2463−2467. (3) Monk, M. Dev. Genet. 1995, 17, 188−197. (4) Ledoux, S.; Nalbantoglu, J.; Cashman, N. R. Mol. Brain Res. 1994, 24, 140−144. (5) Clark, S. J.; Harrison, J.; Paul, C. L.; Frommer, M. Nucleic Acids Res. 1994, 22, 2990−2997. (6) Docherty, S. J.; Davis, O. S. P.; Haworth, C. M. A.; Plomin, R.; Mill, J. Methods 2010, 52, 255−258. (7) Rush, L. J.; Plass, C. Anal. Biochem. 2002, 307, 191−201. (8) Tanaka, K.; Tainaka, K.; Kamei, T.; Okamoto, A. J. Am. Chem. Soc. 2007, 129, 5612−5620. (9) Tanaka, K.; Tainaka, K.; Umemoto, T.; Nomura, A.; Okamoto, A. J. Am. Chem. Soc. 2007, 129, 14511−14517. (10) Proll, J.; Fodermayr, M.; Wechselberger, C.; Pammer, P.; Sonnleitner, M.; Zach, O.; Lutz, D. DNA Res. 2006, 13, 37−42. (11) Wang, X. L.; Song, Y. L.; Song, M. Y.; Wang, Z. X.; Li, T.; Wang, H. L. Anal. Chem. 2009, 81, 7885−7891. (12) Kato, D.; Sekioka, N.; Ueda, A.; Kurita, R.; Hirono, S.; Suzuki, K.; Niwa, O. J. Am. Chem. Soc. 2008, 130, 3716−3717. (13) Wei, H.; Wang, E. Luminescence 2011, 26, 77−85. (14) Richter, M. M. Chem. Rev. 2004, 104, 3003−3036.

Figure 4. Comparison of ECL intensities for real bacteriophage DNA samples (48 502 base pairs, 250 ng). One is as is (Unmodified) and the other is incubated with CpG methyltransferase for methylation to 5′- CpG-3′ sites (Methylated). Other experimental conditions were the same as in Figure 2.

bacteriophage DNA samples before and after an enzymatic methylation reaction. A larger ECL intensity was obtained in the methylated DNA sample than in unmethylated DNA because the methyltransferase methylated cytosine in the CpG sites in the DNA. The DNA sequence has 3112 CpG sites;33 therefore, a total of 48 pmol of cytosine is methylated in the double stranded DNA (250 × 10−9 (g)/32 300 (kDa) × 3112 × 2 = 48 × 10−12 (mol)). Although the amount of methylcytosine in the DNA was 48 pmol, the estimated value was 74 ± 3 pmol, which was obtained from the ECL intensity in Figure 4 and the calibration curve in Figure 2b. The same order of magnitude was obtained between a synthesized short oligomer and a long genomic DNA. Unfortunately, a 1.5 times 1802

dx.doi.org/10.1021/ac202692f | Anal. Chem. 2012, 84, 1799−1803

Analytical Chemistry

Letter

(15) Liu, X. Q.; Shi, L. H.; Niu, W. X.; Li, H. J.; Xu, G. B. Angew. Chem., Int. Ed. 2007, 46, 421−424. (16) Horiuchi, T.; Niwa, O.; Hatakenaka, N. Nature 1998, 394, 659− 661. (17) Leland, J. K.; Powell, M. J. J. Electrochem. Soc. 1990, 137, 3127− 3131. (18) Kurita, R.; Nakamoto, K.; Kato, D.; Niwa, O. Chem. Lett. 2009, 38, 804−805. (19) Kurita, R.; Arai, K.; Nakamoto, K.; Kato, D.; Niwa, O. Anal. Chem. 2010, 82, 1692−1697. (20) So, M. J.; Hvastkovs, E. G.; Schenkman, J. B.; Rusling, J. F. Biosens. Bioelectron. 2007, 23, 492−498. (21) Dennany, L.; Forster, R. J.; White, B.; Smyth, M.; Rusling, J. F. J. Am. Chem. Soc. 2004, 126, 8835−8841. (22) Zhang, L. Y.; Li, D.; Meng, W. L.; Huang, Q.; Su, Y.; Wang, L. H.; Song, S. P.; Fan, C. H. Biosens. Bioelectron. 2009, 25, 368−372. (23) Yao, W.; Wang, L.; Wang, H. Y.; Zhang, X. L.; Li, L. Microchim. Acta 2009, 165, 407−413. (24) Hu, L. Z.; Bian, Z.; Li, H. J.; Han, S.; Yuan, Y. L.; Gao, L. X.; Xu, G. B. Anal. Chem. 2009, 81, 9807−9811. (25) Rodriguez, M.; Bard, A. J. Anal. Chem. 1990, 62, 2658−2662. (26) Naramura, M.; Jang, I. K.; Kole, H.; Huang, F.; Haines, D.; Gu, H. Nat. Immunol. 2002, 3, 1192−1199. (27) Gross, E. M.; Pastore, P.; Wightman, R. M. J. Phys. Chem. B 2001, 105, 8732−8738. (28) Miao, W. J.; Choi, J. P.; Bard, A. J. J. Am. Chem. Soc. 2002, 124, 14478−14485. (29) Delaney, J. L.; Hogan, C. F.; Tian, J. F.; Shen, W. Anal. Chem. 2011, 83, 1300−1306. (30) Wrobel, K.; Figueroa, J. A. L.; Zaina, S.; Lund, G. J. Chromatogr., B 2010, 878, 609−614. (31) Chang, C. H.; Ford, H.; Behrman, E. J. Inorg. Chim. Acta 1981, 55, 77−80. (32) Chen, X.; Yi, C. Q.; Li, M. J.; Lu, X.; Li, Z.; Li, P. W.; Wang, X. R. Anal. Chim. Acta 2002, 466, 79−86. (33) http://blast.ddbj.nig.ac.jp/

1803

dx.doi.org/10.1021/ac202692f | Anal. Chem. 2012, 84, 1799−1803