Technical Note pubs.acs.org/ac
Associated Analysis of DNA Methylation for Cancer Detection Using CCP-Based FRET Technique Jiangyan Zhang,† Baoling Xing,‡ Jinzhao Song,† Feng Zhang,‡ Chenyao Nie,† Lian Jiao,‡ Libing Liu,† Fengting Lv,† and Shu Wang*,† †
Beijing National Laboratory for Molecular Sciences, Key Laboratory of Organic Solids, Institute of Chemistry, Chinese Academy of Sciences, Beijing 100190, P. R. China ‡ Department of Pathology, Changzhou Women and Children Health Hospital, Changzhou 213003, P. R. China S Supporting Information *
ABSTRACT: This paper describes an associated analysis method of DNA methylation for the detection of cancer using an optically amplifying cationic conjugated polymer (CCP, poly{(1,4-phenylene)-2,7-[9,9-bis(6′-N,N,N-trimethyl ammonium)-hexyl fluorene] dibromide)}. Genomic DNA is digested by methylation-sensitive restriction endonuclease, followed by PCR amplification to incorporate fluorescein-labeled dNTP. Only methylated DNA can be amplified by PCR, and the methylation level is detected through fluorescence resonance energy transfer (FRET) between CCP and fluorescein that is incorporated into the PCR product. The methylation levels of RASSF1A, OPCML, and HOXA9 promoters of 35 ovarian cancer samples and 11 normal samples were assayed. In accordance with the degree of methylation levels, they are clustered to three sections and assigned a value. Through an associated analysis, we acquired a threshold for cancer detection with a sensitivity of 85.7%. The assay takes about 20 h to obtain the detection results and shows great potential as a useful tool for diagnostic and screening of cancer.
D
labeled probe. There are also many gene-specific methylation detection methods that are not based on MSP. Among them, MS-SnuPE (methylation-sensitive single nucleotide primer extension) has the disadvantage of requiring the use of radioactive materials.14 COBRA (combined bisulfite restriction analysis) needs labor-intensive separation and hybridization.15 Pyrosequencing offers a highly reliable, quantitative, and highthroughput method for analysis, but the read length is limited by the thermal instability of enzymes.16 In recent years, water-soluble cationic conjugated polymers (CCP) have attracted much attention in sensitive biosensor applications.17,18 CCP is highly fluorescent in aqueous media and can affine with fluorescein-labeled DNA to amplify the fluorescence intensity through an effective energy transfer pathway.19 Our group has established two methylation detection methods based on CCP-based FRET (fluorescence resonance energy transfer).20,21 In one method, the genomic DNA was treated with bisulfite to induce C/T polymorphism, followed by single nucleotide base extension;20 in the other method, the genomic DNA was treated with HpaII to digest unmethylated DNA, followed by two-round PCR.21 Compared with the bisulfite treatment method, the HpaII treatment
NA methylation is a major epigenetic mechanism and plays an important role in human cancer.1−3 Aberrant promoter methylation of certain gene has attracted considerable interest as a potential biomarker for cancer detection.4,5 Up to date, few single genes that are methylated in a high proportion of cancer have been identified. It is likely that determining the methylation status of a panel of genes, rather than individual gene, will provide a more sensitive and specific diagnosis. Many methylation patterns have been reported in various cancers.6−9 For an ideal pattern, an accurate detection method and logical data analysis are necessary. With the development of research on DNA methylation, numerous detection methods have been established. For DNA methylation panel, methods to detect gene-specific methylation are needed. In consideration of the low amount of sample, most PCR-assisted methods are of practical importance because of their high sensitivity. MSP (methylation-specific PCR) is a very sensitive methylation detection method, but it is purely a qualitative approach and delivers just a “yes” or “no” answer for the methylation status.10 Then quantitative approaches are developed based on MSP, such as MethyLight,11 MS-FLAG (methylation-specific fluorescent amplicon generation),12 and SMART-MSP (sensitive melting analysis after real time methylation-specific PCR).13 A special real-time PCR instrument was used for all three methods. MethyLight and MSFLAG are not cost-effective because of the use of a double© 2013 American Chemical Society
Received: August 26, 2013 Accepted: December 10, 2013 Published: December 10, 2013 346
dx.doi.org/10.1021/ac402720g | Anal. Chem. 2014, 86, 346−350
Analytical Chemistry
Technical Note
Table 1. Primers for the PCR gene
forward primer
reverse primer
RASSF1A OPCML HOXA9
5′-GGAGGCGCTGAAGTCGG-3′ 5′-GCCAGTGTCAGTTTTCAGTTTG-3′ 5′-TGGACTCGTTCCTGCTGG-3′
5′-GCCCAGCGGGTGCCA-3′ 5′-ATCCCTGACCGCCACTTT-3′ 5′-TGGTGGTGATGGTGGTGGTA-3′
OPCML, and HOXA9 were selected, according to the published literature. All of the gene sequences are available at the following Web site: http//:www.ensembl.org. The region of gene to be amplified should cover two or three 5′-CCGG-3′ sites, which should not occupy the primer positions. PCR primers were designed using DNAMAN (Lynnon Biosoft, Canada), with a similar melting temperature (Tm) close to or higher than 60 °C, obtained by adjusting the length of primers. All the primers used are listed in Table 1. Detection of Methylation Level. The methylation levels of selected genes were detected by using the CCP-based FRET method. The digested DNA was amplified with two-round PCR. The first-round PCR reaction mixture consisted of 1 μL of digested DNA sample, primers (each 0.4 μM), 50 μM dNTPs, 0.02 units GoTaq Hot Start DNA polymerase (Promega), 1 × GoTaq Flexi buffer, and 0.6 mM MgCl2 in a reaction volume of 25 μL. The reaction was carried out with a hot start of 95 °C for 4 min, followed by 15 cycles of 95 °C for 30 s, an appropriate annealing temperature for 30 s, and 72 °C for 40 s. A volume of 1 μL of the first-round PCR product was transferred to the second-round PCR reaction mixture with a final volume of 20 μL. The second-round PCR reaction mixture included 0.5 μM each primer, 1 × dNTP/Fl-dNTP mix (4 μM dCTP, 4 μM dATP, 3.5 μM dGTP, 3.5 μM dTTP, 0.5 μM FldGTP, and 0.5 μM Fl-dUTP), 0.025 units HotMaster Taq DNA polymerase (TIANGEN), and 1 × HotMaster Taq buffer or 1 × GC buffer (TaKaRa). The PCR conditions were as follows: 95 °C for 4 min, followed by 30 cycles at 95 °C for 30 s, an appropriate annealing temperature for 30 s, and 65 °C for 60 s. For degradation of excess dNTPs and Fl-dNTPs, 3 μL of a mixture containing 1.5 units of shrimp alkaline phosphatase (SAP) (TaKaRa) in 1.5 μL of 10× SAP buffer [500 mM TrisHCl (pH 9.0), 50 mM MgCl2] were added to 12 μL of PCR product and incubated at 37 °C for 30 min, followed by heat inactivation of the enzyme at 85 °C for 10 min. After SAP digestion, the product can be detected by using CCP. The CCP working solution was prepared by adding 30 μL of CCP store solution (1 mM) to 485 μL of dimethyl sulfoxide (DMSO) and 485 μL of HEPES buffer (25 mM, pH 8.0). Ninety microliters of CCP working solution and 10 μL of digested sample were mixed in a 96-well plate. The 96-well plate was then read with the Multi-Mode Microplate Reader (BioTek) with an excitation of 380 nm and emissions at 424 and 530 nm. Emission filter 1 (440/30 nm) was used to detect emissions at 424 nm, and emission filter 2 (528/20 nm) was used to detect emissions at 530 nm. Calculate FRET ratios (I530 nm/I424 nm) to obtain the methylation level “E” in equation 1:
method can give a quantitative methylation level of a sample. Later, the HpaII treatment method was applied to analyze the relationship of cancer and methylation levels. In 2012, our group has developed a cumulative methylation alteration panel of multiplex genes and stepwise discriminant analysis on the basis of the methylation levels of a set of candidate genes in cancer samples and normal samples.22 Three genes were chosen to compose the panel and deduce a discriminant equation with 86.3% accuracy and 86.7% sensitivity for cancer detection. Compared with previous cumulative detection, this method takes into account the degree and contribution of promoter methylation in different candidate genes. However, the diagnostic result depends on a statistics analysis of a large number of samples, thus this method is difficult to be used for point-of-care diagnostics. In this paper, we describe an associated analysis of DNA methylation using a CCP-based FRET technique. Through an associated analysis of the methylation levels, a threshold can be set to diagnose cancer. Herein, we improve the data analysis mode based on the methylation level tested by the CCP-based FRET assay. The methylation level is clustered to three sections and assigned a value. Counting and comparing the “SUM” of the values of the candidate genes between cancer samples and normal samples, a threshold can be set to diagnose cancer. The sample with the SUM exceeding the threshold can be diagnosed as a cancer sample and the conclusion of “cancer” is definite. One important improvement has been incorporated into this protocol compared to our original publication:22 a threshold is used in this protocol instead of statistics and cumulative analysis to make it possible for point-of-care diagnostics.
■
EXPERIMENTAL SECTION
Tissue Samples and DNA Preparation. For this study, we used 35 ovarian cancer samples and 11 normal samples from Changzhou Women and Children Health Hospital. All tissue samples were obtained after receiving informed consent. Genomic DNA samples of formalin-fixed and paraffinembedded tissue sections were isolated and extracted following the instructions provided in a TIANamp Genomic DNA kit (QIAGEN). DNA concentration was quantified by absorbance at 260 nm using Nanodrop spectrophotometers (Thermo Scientific, ND2000). The genomic DNA was divided into two identical sections: section A for HpaII treatment and section B for non-HpaII treatment. The section A reaction system comprised 200 ng of genomic DNA, 20 units of HpaII (New England Biolabs), 1 × NEBuffer I (10 mM bis-tris-propane− HCl (pH 7.0), 10 mM MgCl2, and 1 mM dithiothreitol), and deionized water. Section B reaction system was comprised of 200 ng of genomic DNA, 1 × NEBuffer I, and deionized water. Both reaction systems were incubated at 37 °C for 12 h, followed by heat inactivation of HpaII at 85 °C for 15 min. The digested DNA was used for PCR amplification or stored at −80 °C. Gene Selection and Primer Design. The promoter of the selected gene should be unmethylated in normal cases and methylated in cancer cases. In this study, three genes RASSF1A,
E=
FRET ratioHpaII − FRET ratio blank FRET rationon‐HpaII − FRET ratio blank
(1)
Associated Analysis. Through analyzing the methylation status of three genes of ovarian cancer samples and normal samples, the methylation level was clustered to three sections and each section was assigned a value 1, 2, and 3, respectively. Among them, value 1 represented a low methylation level 347
dx.doi.org/10.1021/ac402720g | Anal. Chem. 2014, 86, 346−350
Analytical Chemistry
Technical Note
below 0.3, value 2 represented a moderate methylation level between 0.3 and 0.7, and value 3 represented high methylation level above 0.7. For each sample, three values are added up to a sum. A threshold was set after comparing the difference between the sum of cancer samples and normal samples.
Scheme 1. Principle of Methylation Level Analysis of Cancer-Related Genes Using CCP-Based FRET Technique, Together with the Chemical Structures of CCP and Fluorescein Used in the Detection
■
RESULTS AND DISCUSSION CCP can form a complex with negatively charged fluoresceinlabeled DNA through electrostatic interactions. As an energy donor with high fluorescence in aqueous media, CCP transfers energy to the acceptor fluorescein through FRET, leading to a significant optical amplification of fluorescein. CCP displays absorption maxima at approximately 380 nm and emission maxima at 424 nm. The emission of CCP overlaps the absorption of fluorescein, which meets the requirement for FRET.20,21 On the basis of CCP-based FRET, fluoresceinlabeled DNA can be detected. The principle of associated analysis of DNA methylation depends on the analysis of the methylation level of cancer-related genes using a CCP-based FRET technique with our previous method (Scheme S1 of the Supporting Information).22 First, the genomic DNA is divided into two identical sections: section A is used to detect methylated DNA and section B is used to detect all DNAs including methylated and unmethylated DNAs. In section A, genomic DNA is digested with restriction endonuclease HpaII that can cleave the 5′-CCGG-3′ site of the unmethylated DNA, whereby keeping the methylated ones unaffected. The higher the methylation level, the more genomic DNA are left to be the template of PCR; in section B, genomic DNA is not treated with HpaII and so both methylated and unmethylated DNAs are intact. All genomic DNA are left to be the template of PCR. Second, two-round PCR is performed and Fl-dNTPs are incorporated into the PCR amplicons in the second-round PCR. In this step, methylated DNA of section A and both methylated and unmethylated DNAs of section B are amplified. Third, the PCR product is treated with alkaline phosphatase to degrade excess dNTPs. Finally, the PCR product is mixed with CCP solution for the FRET signal from CCP to fluorescein to occur due to strong electrostatic interactions (Scheme 1, the chemical structures of CCP and fluorescein are also shown). The ratio of emission intensities of CCP to fluorescein is measured. Methylated DNA in section A will produce FRET ratioHpaII; both methylated and unmethylated DNA in section B will produce FRET rationon‑HpaII; non-template control will produce FRET ratioBlank. Parameter E is defined for viewing the methylation level with a subtracted background. For each gene, the methylation level is clustered in three sections: low, moderate, and high. A low methylation level is assigned as 1; a moderate methylation level is assigned as 2, and a high methylation level is assigned as 3. The values of three genes for each sample are added up, and the SUM is used as the threshold to diagnose cancer (Scheme 1). In this work, we examined ovarian cancer based on gene methylation. Three ovarian cancer-related genes were selected from the published literature, including RASSF1A (RASassociation domain factor 1A gene),23−25 OPCML (opioid binding protein/cell adhesion molecule-like),26,27 and HOXA9 (homeobox A9).28 To avoid false positives or negatives of the method, it is necessary to ensure the sensitivity and specificity of PCR. Thus the PCR condition should be carefully optimized. For effect minimization of the potential false priming when the genomic DNA is amplified, the first-round PCR is limited to no more than 25 cycles and the dNTP concentration is reduced
from 100 to 10 μM in the second-round PCR. With concern for the addition of Fl-dNTP, one labeled purine and one labeled pyrimidine should be in equal amounts. In this protocol, FldUTP and Fl-dGTP are used. The simultaneous incorporation of Fl-dUTP and Fl-dGTP, not using them separately, can eliminate the bias arising from the difference in dNTP fractions because the (dT + dG) value is identical in double-stranded DNA. Formation of secondary structure might cause PCR failure because of high GC fractions. It is an alternative to optimize the PCR buffer for lowering the annealing temperature and facilitating dissociation of the DNA template. The control experiments were also performed. In the first-round PCR, water instead of the genomic DNA, is added to the reaction tube. The corresponding FRET ratio, defined as the ratio of emission intensity at 530 and 424 nm, is obtained as a background signal when analyzing FRET data. If the FRET ratio of the negative control is too high, we exclude the possibility of PCR contamination or elevate the annealing temperature or redesign the primers to avoid the formation of hairpins, dimers, and cross-dimers. RASSF1A is a tumor suppressor gene that can regulate proliferation, induce apoptosis, and stabilize microtubules.29 The methylation levels of the RASSF1A gene are detected using the CCP-based FRET technique. The fluorescence spectra of different scenarios are shown in Figure 1, taking the low, moderate, and high methylation levels as representative. The methylation levels E calculated in eq 1 are 0.21, 0.46, and 0.89, respectively. All the methylation levels of the RASSF1A gene of 11 normal samples and 35 ovarian cancer samples are assayed 348
dx.doi.org/10.1021/ac402720g | Anal. Chem. 2014, 86, 346−350
Analytical Chemistry
Technical Note
Figure 1. Fluorescence spectra of different scenarios, including a (a) low methylation level, (b) moderate methylation level, and (c) high methylation level. Fluorescence spectra were obtained using a Hitachi F-4500 fluorometer equipped with a xenon lamp. Excitation wavelength was 380 nm.
similarly. As shown in Figure 2, the methylation level of all 11 normal samples are below 0.3; 8 out of 35 ovarian cancer
Figure 3. Values of RASSF1A, OPCML, and HOXA9 genes and the sum of three genes. Thirty-five cancer samples are labeled from 1 to 35, and 11 normal samples are labeled from N1 to N11. The SUM of each sample is listed at the end of the respective row. Blue zone represents low methylation level 1; orange zone represents moderate methylation level 2, and red zone represents high methylation level 3. Green zone represents SUM that is not more than 4, and pink zone represents that exceeding 4.
Figure 2. Methylation level of RASSF1A promoter of 35 cancer samples and 11 normal samples measured by the CCP-based FRET technique.
samples are above 0.7; 9 out of 35 are between 0.3 and 0.7; and 18 out of 35 are below 0.3. OPCML expression inhibits ovarian cancer cell growth, enhances intercellular attachment, and abrogates both subcutaneous and intraperitoneal tumorigenicity.30 In the OPCML gene, 2 out of 11 normal samples are between 0.3 and 0.7, 9 out of 11 are below 0.3; 12 out of 35 ovarian cancer samples are above 0.7; 15 out of 35 are between 0.3 and 0.7; and 8 out of 35 are below 0.3. HOXA9 is a member of the HOX genes that plays a definitive role in the determination of cell fate during embryogenesis and hematopoiesis.31 In the HOXA9 gene, all 11 normal samples are below 0.3; 17 out of 35 ovarian cancer samples are above 0.7; 10 out of 35 are between 0.3 and 0.7; and 8 out of 35 are below 0.3. To assess combined methylation status of three genes, the methylation level is clustered to three sections and assigned a value: low methylation level below 0.3 as 1, moderate methylation level between 0.3 and 0.7 as 2, and high methylation level above 0.7 as 3. For each sample, three values are added up to a sum. As shown in Figure 3, the SUM of all normal samples is not more than 4, and 31 out of 35 cancer samples exceed 4. So 4 can be set as a threshold. If the SUM of a sample exceeds 4, it is affirmative to diagnose it as cancer. The sensitivity of the protocol is calculated to be 85.7% (30/35), and the specificity is 100% (11/11).
■
CONCLUSIONS In this paper, we establish an associated analysis of the methylation level using the CCP-based FRET technique and the statistic threshold can be used for cancer diagnosis. The CCP-based FRET technique shows great potential as a useful tool for diagnostic and screening of cancer. The CCP-based FRET technique is sensitive, thus small DNA samples obtained from serum, plasma, or stool can be detected. It is also sensitive enough to detect low methylation levels of the gene promoter, thus it meets the requirement for most research purposes, including early cancer diagnosis. Our method is continuous and homogeneous, avoiding isolation and washing steps. The FldNTP is incorporated into the PCR product, without requiring fluorescent labels on the primer; thus, the cost is reduced and the time for synthesis of labeled DNA primers is saved. In our protocol, the degree and contribution of promoter methylation in different candidate genes has been taken into account and the diagnostic conclusion of cancer obtained by this protocol is definite. However, there are still some limitations of our method: only known methylation sites can be detected and the selection of methylation sites is also restricted by the methylation-sensitive restriction enzyme. Also because of the possible incomplete digestion of the restriction enzyme and the 349
dx.doi.org/10.1021/ac402720g | Anal. Chem. 2014, 86, 346−350
Analytical Chemistry
Technical Note
(21) Feng, F.; Liu, L.; Wang, S. Nat. Protoc. 2010, 5, 1255−1264. (22) Yang, Q.; Ying, D.; Wu, W.; Zhu, C.; Chong, H.; Lu, J.; Yu, D.; Liu, L.; Lv, F.; Wang, S. Nat. Commun. 2012, 3, 1206. (23) de Caceres, I. I.; Battagli, C.; Esteller, M.; Herman, J. G.; Dulaimi, E.; Edelson, M. I.; Bergman, C.; Ehya, H.; Eisenberg, B. L.; Cairns, P. Cancer Res. 2004, 64, 6476−6481. (24) Liggett, T. E.; Melnikov, A.; Yi, Q.; Replogle, C.; Hu, W.; Rotmensch, J.; Kamat, A.; Sood, A. K.; Levenson, V. Gynecol. Oncol. 2011, 120, 113−120. (25) Rathi, A.; Virmani, A. K.; Schorge, J. O.; Elias, K. J.; Maruyama, R.; Minna, J. D.; Mok, S. C.; Girard, L.; Fishman, D. A.; Gazdar, A. F. Clin. Cancer Res. 2002, 8, 3324−3331. (26) Cui, Y.; Ying, Y.; van Hasselt, A.; Ng, K. M.; Yu, J.; Zhang, Q.; Jin, J.; Liu, D.; Rhim, J. S.; Rha, S. Y.; Loyo, M.; Chan, A. T. C.; Srivastava, G.; Tsao, G. S. W.; Sellar, G. C.; Sung, J. J. Y.; Sidransky, D.; Tao, Q. PLoS One 2008, 3, e2990. (27) Teodoridis, J. M.; Hall, J.; Marsh, S.; Kannall, H. D.; Smyth, C.; Curto, J.; Siddiqui, N.; Gabra, H.; McLeod, H. L.; Strathdee, G.; Brown, R. Cancer Res. 2005, 65, 8961−8967. (28) Wu, Q.; Lothe, R. A.; Ahlquist, T.; Silins, I.; Tropé, C. G.; Micci, F.; Nesland, J. M.; Suo, Z.; Lind, G. E. Mol. Cancer 2007, 6, 45. (29) Agathanggelou, A.; Cooper, W. N.; Latif, F. Cancer Res. 2005, 65, 3497−3508. (30) Sellar, G. C.; Watt, K. P.; Rabiasz, G. J.; Stronach, E. A.; Li, L.; Miller, E. P.; Massie, C. E.; Miller, J.; Contreras-Moreira, B.; Scott, D.; Brown, I.; Williams, A. R.; Bates, P. A.; Smyth, J. F.; Gabra, H. Nat. Genet. 2003, 34, 337−343. (31) Samuel, S.; Naora, H. Eur. J. Cancer 2005, 41, 2428−2437.
bias of PCR amplification in independent tubes, the methylation level detected by the CCP-based FRET method is not absolutely quantitative.
■
ASSOCIATED CONTENT
S Supporting Information *
The assay mechanism of CCP-based FRET method for methylation detection. This material is available free of charge via the Internet at http://pubs.acs.org.
■
AUTHOR INFORMATION
Corresponding Author
*E-mail:
[email protected]. Author Contributions
The manuscript was written through the contribution of all authors. All authors have given approval to the final version of the manuscript. Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS The authors are grateful to the National Natural Science Foundation of China (Grants 21033010, 21273254, and 21021091) and the Major Research Plan of China (Grants 2012CB932600, 2011CB935800, and 2013CB932800).
■
REFERENCES
(1) Baylin, S. B.; Ohm, J. E. Nat. Rev. Cancer 2006, 6, 107−116. (2) Feinberg, A. P.; Ohlsson, R.; Henikoff, S. Nat. Rev. Genet. 2006, 7, 21−33. (3) Esteller, M. N. Engl. J. Med. 2008, 358, 1148−1159. (4) Duffy, M. J.; Napieralski, R.; Martens, J. W. M.; Span, P. N.; Spyratos, F.; Sweep, F. C. G. J.; Brunner, N.; Foekens, J. A.; Schmitt, M. Eur. J. Cancer 2009, 45, 335−346. (5) Mulero-Navarro, S.; Esteller, M. Crit. Rev. Oncol. Hematol. 2008, 68, 1−11. (6) Tsou, J. A.; Galler, J. S.; Siegmund, K. D.; Laird, P. W.; Turla, S.; Cozen, W.; Hagen, J. A.; Koss, M. N.; Laird-Offringa, I. A. Mol. Cancer 2007, 6, 70. (7) Worm, J.; Aggerholm, A.; Guldberg, P. Clin. Chem. 2001, 47, 1183−1189. (8) Jones, P. A.; Baylin, S. B. Nat. Rev. Genet. 2002, 3, 415−428. (9) Muller, H. M.; Fiegl, H.; Widschwendter, A.; Widschwendter, M. Ann. N.Y. Acad. Sci. 2004, 1022, 44−49. (10) Herman, J. G.; Graff, J. R.; Myohanen, S.; Nelkin, B. D.; Baylin, S. B. Proc. Natl. Acad. Sci. U.S.A. 1996, 93, 9821−9826. (11) Eads, C. A.; Danenberg, K. D.; Kawakami, K.; Saltz, L. B.; Blake, C.; Shibata, D.; Danenberg, P. V.; Laird, P. W. Nucleic Acids Res. 2000, 28, e32. (12) Bonanno, C.; Shehi, E.; Adlerstein, D.; Makrigiorgos, G. M. Clin. Chem. 2007, 53, 2119−2127. (13) Kristensen, L. S.; Mikeska, T.; Krypuy, M.; Dobrovic, A. Nucleic Acids Res. 2008, 36, e42. (14) Gonzalgo, M. L.; Jones, P. A. Nucleic Acids Res. 1997, 25, 2529− 2531. (15) Xiong, Z.; Laird, P. W. Nucleic Acids Res. 1997, 25, 2532−2534. (16) Colella, S.; Shen, L.; Baggerly, K. A.; Issa, J. P.; Krahe, R. Biotechniques 2003, 35, 146−150. (17) Feng, F.; He, F.; An, L.; Wang, S.; Li, Y.; Zhu, D. Adv. Mater. 2008, 20, 2959−2964. (18) Thomas, S. W., III; Joly, G. D.; Swager, T. M. Chem. Rev. 2007, 107, 1339−1386. (19) Gaylord, B. S.; Heeger, A. J.; Bazan, G. C. J. Am. Chem. Soc. 2003, 125, 896−900. (20) Feng, F.; Wang, H.; Han, L.; Wang, S. J. Am. Chem. Soc. 2008, 130, 11338−11343. 350
dx.doi.org/10.1021/ac402720g | Anal. Chem. 2014, 86, 346−350
