Fluorescent Strategy Based on Cationic Conjugated Polymer

Article pubs.acs.org/ac

Fluorescent Strategy Based on Cationic Conjugated Polymer Fluorescence Resonance Energy Transfer for the Quantification of 5‑(Hydroxymethyl)cytosine in Genomic DNA Tingting Hong,† Tianlu Wang,† Pu Guo,† Xiwen Xing, Fei Ding, Yuqi Chen, Jinjun Wu, Jingwei Ma, Fan Wu, and Xiang Zhou* College of Chemistry and Molecular Sciences, Key Laboratory of Biomedical Polymers of Ministry of Education, Wuhan University, Wuhan, Hubei 430072, People’s Republic of China S Supporting Information *

ABSTRACT: DNA methylation is dynamically reprogrammed during early embryonic development in mammals. It can be explained partially by the discovery of 5(hydroxymethyl)cytosine (5-hmC), 5-formylcytosine (5-fC), and 5-carboxylcytosine (5-caC), which are identified as key players involved in both active and passive demethylation pathways. As one of the ten−eleven translocation oxidation products, 5-hmC was found relatively abundant in neuron cells and embryonic stem cells. Herein we report a new method for 5-hmC quantification in genomic DNA based on CCP-FRET (cationic conjugated polymers act as the energy donor and induce fluorescence resonance energy transfer) assay combined with KRuO4 oxidation. 5-hmC in genomic DNA can be selectively transformed into 5-fC by the oxidation of KRuO4 and then labeled with hydroxylamine-BODIPY (BODIPY = 4,4difluoro-4-bora-3a,4a-diaza-s-indacene) fluorophore through the reaction between 5-fC and hydroxylamine-BODIPY. After the fluorescently labeled DNA was captured by CCP through electrostatic interactions, a significant FRET between CCP and hydroxylamine-BODIPY fluorophore was observed. This CCP-FRET-based assay benefits from light-harvesting, large Stokes shift, and optical signal amplification properties of the CCP. Furthermore, this CCP-FRET-based assay was quite successfully demonstrated for the 5-hmC quantification in three types of cells (mESc, HeLa, HEK 293T), providing a much more convenient choice for 5-hmC quantification in genomic DNA.

A

s a mark for transcriptional silencing, DNA methylation is central to many important cellular processes. The methylation state of DNA is found to be dynamically changed during mammalian early development,1,2 partially because of active demethylation. Additionally, the discovery of the ten− eleven translocation (Tet) oxidation products 5(hydroxymethyl)cytosine (5-hmC), 5-formylcytosine (5-fC), and 5-carboxylcytosine (5-caC) suggests a possible demethylation pathway.3−7 In contrast to 5-methylcytosine (5-mC), 5hmC varies substantially among different cell and tissue types (it is relatively abundant in neuron cells8−10 and embryonic stem cells11), and this variance indicates that 5-hmC plays a significant role in sculpting the epigenetic landscape.12,13 It is a challenge to develop a method to map the genomewide distribution of 5-hmC. However, many strategies capable of labeling and enriching 5-hmC in genomic DNA have been reported. Affinity enrichment based on a biotin tag is a major method for targeted sequencing of 5-hmC in genomic DNA;14−17 examples are enzymatic modification by βglucosyltransferase followed by the installation of a biotin tag through click chemistry14 and glycol oxidization with sodium periodate to generate aldehydes and then captured by ARP.15 © 2013 American Chemical Society

All these methods enrich the trace amounts of 5-hmC in genomic DNA and make 5-hmC available for high-throughput sequencing. Another attempt to measure 5-hmC with singlebase resolution combined with bisulfite sequencing has also been successful. Balasubramanian’s group found that 5-hmC could be selectively oxidized to 5-fC by KRuO4 in high yield;18 then 5-hmC was transformed into U after bisulfite treatment. Thus, 5-hmC in genomic DNA can be mapped through the comparison of oxidative bisulfite sequencing with KRuO4 and bisulfite-only sequencing. With β-glucosyltransferase, He et al. innovatively transformed 5-hmC into β-glucosyl-5(hydroxymethyl)cytosine, which could be protected from transformation into U after oxidization by Tet 1 and bisulfite treatment.19,20 However, the methods for direct fluorescence labeling of 5-hmC in genomic DNA are still limited, and many drawbacks exist. Received: July 7, 2013 Accepted: September 2, 2013 Published: September 2, 2013 10797

dx.doi.org/10.1021/ac4020676 | Anal. Chem. 2013, 85, 10797−10802

Analytical Chemistry

Article

Scheme 1. Principle of the CCP-FRET-Based Assay To Detect the 5-fC Modificationa

a Hydroxylamine-BODIPY can selectively interact with double-stranded DNA that contains a 5-fC modification. Then, with the addition of CCP, the negatively charged dsDNA can be captured by CCP through electrostatic interaction; the capture leads to significant FRET from CCP to the hydroxylamine-BODIPY label. No FRET occurred for normal dsDNA containing cytosine.

emission of the CCP; thus, hydroxylamine-BODIPY is suitable for FRET. Because of the negative charge of DNA, the fluorescence-modified DNA can be brought in close proximity to the positively charged CCP; this system satisfies the distance requirement for FRET through an electrostatic force. Thus, significant FRET from CCP to hydroxylamine-BODIPY-labeled DNA can be observed if the DNA contains the 5-fC modification. However, no FRET occurs if the DNA does not contain the 5-fC modification because the unmodified DNA cannot be labeled with hydroxylamine-BODIPY.

As we know, in contrast to 5-hmC, 5-fC is more active and can be directly modified by hydroxylamine or hydrazine derivatives that contain a biotin tag or a fluorescein group for further quantification.8,15,27 Thus, we try to develop a method capable of 5-hmC quantification in genomic DNA through fluorescence detection after KRuO4 oxidation.18 However, to our knowledge, in this method, the problem of a high background caused by fluorescent compounds is still unsolved. One way to lower background fluorescence is purification through HPLC;21 however, purification makes this method cumbersome and inefficient when being applied in genomic DNA. Thus, seeking a method to enhance the signal-to-noise ratio becomes the key factor in a fluorescence labeling method to quantify 5-hmC in genomic DNA. CCPs, cationic conjugated polymers, have emerged as promising nucleic acid biosensors because of their watersoluble property and extraordinary light-harvesting capability.22−25 The polymers can act as energy donors and transfer energy to an acceptor fluorophore by fluorescence resonance energy transfer (FRET) when the polymer and fluorophore are close enough. The strategy of CCP-FRET can significantly reduce spectral crosstalk and amplify the fluorescence signal23,26 because of the large Stokes shift and conjugated backbone, respectively; thus, this CCP-FRET assay facilitates the detection of low-concentration targets with minimal background. Here, we report the first demonstration of a CCP-FRETbased assay to detect 5-fC based on the electrostatic interactions between the CCP and the oppositely charged DNA; this assay can be further applied to the quantification of 5-hmC in genomic DNA after the oxidation by KRuO4. The principle is highlighted in Scheme 1. DNA that is modified with 5-fC can be captured by hydroxylamine-containing compounds as reported in previous studies.8,10,27 We specifically used hydroxylamine-BODIPY (reported in our previous study27) as an acceptor whose absorption overlaps with the characteristic

■

EXPERIMENTAL SECTION Materials. 2,7-Dibromo-9,9-bis(6′-bromohexyl)fluorene, 1,4-phenyldiboronic acid, and Pd(dppf)Cl2 were purchased from Synwit Technology Co., Ltd. (Beijing, China). p(chloromethyl)benzoyl chloride, 2,4-dimethylpyrrole, N-hydroxyphthalimide, and anisidine were purchased from SigmaAldrich. Oligomers were purchased from TaKaRa Biotech (China). The Thermo Scientific Slide-A-Lyzer G2 dialysis cassette is used to remove the KRuO4 through dialysis in 1 L of distilled water for 24 h. The final polyacrylamide gel electrophoresis products were scanned with a Pharos FX molecular imager (Bio-Rad, Hercules, CA). Methods. 5-Formylcytosine Selective Analysis by Polyacrylamide Gel Electrophoresis. Different modifications containing DNA were incubated with 20 μM hydroxylamineBODIPY in 100 mM NH4OAc (pH 5.0) buffer at 37 °C for 12 h in the presence of 100 mM anisidine. After addition of 10 μL of CH3COONa−CH3COOH buffer (1 M, pH 5.0) and 900 μL of ethanol, the mixture was frozen at −80 °C for 2 h and centrifuged for 20 min at 4 °C. The DNA samples were loaded into a 20% polyacrylamide gel for electrophoresis at 20 V/cm; this electrophoresis occurred at room temperature in 1× TBE (Tris/borate/EDTA) buffer. The gel was then analyzed using the Pharos FX molecular imager (Bio-Rad). 10798

dx.doi.org/10.1021/ac4020676 | Anal. Chem. 2013, 85, 10797−10802

Analytical Chemistry

Article

Fluorescence Detection Based on CCP. 5-fC-containing DNA was incubated with hydroxylamine-BODIPY under the optimal conditions described above. After reaction, the mixture solution was extracted with 500 μL of CHCl3, followed by vortexing and centrifugation. Then 2.5 μM CCP in repeat units was added to the extracted DNA solution, and the FRET ratio rested at a fixed value after several seconds of vortexing. Genomic DNA Oxidized by KRuO4. The genomic DNA extracted from different cells (about 10 μg) was denatured with 0.05 M NaOH at 37 °C for 30 min. Then the mixture was quickly put on 0 °C ice for about 5 min, followed by the addition of 2 μL of saturated KRuO4. Then it was held on ice for 1 h and dialyzed by the Thermo Scientific Slide-A-Lyzer G2 dialysis cassette in doubly distilled water (ddH2O) for 24 h to remove KRuO4. After that, the genomic DNA was incubated with hydroxylamine-BODIPY under the optimal conditions described above.

BODIPY are 492 and 518 nm, respectively. Thus, the emission peak of hydroxylamine-BODIPY is easily overlapped by its strong absorption peak, especially for the weak fluorescence emission at 518 nm, while this CCP-FRET-based assay can significantly reduce the spectral crosstalk for hydroxylamineBODIPY because of the large Stokes shift of CCP, which would allow the emission peak at 518 nm to be easily recognized. Figure 4b also demonstrates the optical amplification properties of the light-harvesting conjugated polymer CCP. Comparing the emission intensity for the direct excitation of hydroxylamine-BODIPY (λex = 492 nm) to the FRET emission from the excitation of the CCP at 380 nm, significant fluorescence amplification can be observed,23 and this amplification increases the sensitivity of this CCP-FRET-based assay. In the solution, there is excess fluorescent dye, which may aggregate around the CCP and generate a specific background signal through energy transfer from CCP to the surrounding fluorescent dye; therefore, the reaction solution was extracted by CHCl3 to remove the excess fluorescent dye. A comparison of the fluorescence from 5-fC-modified DNA and the control (without 5-fC modification) after the same treatment with fluorescent dye and extraction is demonstrated in Figure 4a, and only a small background can be observed because of the inefficient FRET between the CCP and the distant residual fluorescent dye after the extraction. However, to our knowledge, even a small amount of residual fluorescent dye can cause a background in direct excitation schemes, and this background hampers the detection of low fluorescence and makes purification by HPLC necessary for these methods.21 However, because of the inefficient FRET between the CCP and the residual fluorescent dyes, this cumbersome step of HPLC purification can be avoided in our CCP-FRET-based assay. Because genomic DNA is double-stranded (dsDNA) , the FRET ratios (I518/I420) of fluorescently modified singlestranded DNA (ssDNA) and dsDNA are compared in Figure 4a. A slight difference is observed because of the more complicated structure of double-stranded DNA. Additionally, because there are a comparable number of abasic sites that arise from the cleavage of the glycosidic bond and that contain an aldehyde functional group in genomic DNA, we also performed an experiment to test the efficiency of our method to selectively label 5-fC. The PAGE analysis in Figure 1 shows that no band for fluorescent-dye-labeled DNA occurred for the DNA containing abasic sites. The same result was confirmed through fluorescence measurement (Figure 4a, gray line) and indicated that the selectivity and accuracy of the detection of 5-fC would not be impacted by the existence of abasic sites in the genomic DNA. To further demonstrate this CCP-FRET-based assay in genomic DNA, a standard curve was established to test unknown samples. For the standard curve, different concentrations of 5-fC-modified double-stranded DNA were incubated with hydroxylamine-BODIPY under the optimal conditions described above. After the addition of 2.5 μM CCP in repeat units and several seconds of vortexing, the FRET ratio rested at a fixed value. The FRET ratios (I518/I420) were plotted as a function of the concentration of dsDNA, and a linear model fits the experimental data very well (shown in Figure 4c). Application of This CCP-FRET-Based Assay in Genomic DNA. The high sensitivity of the assay was demonstrated through the application of the assay to different cell lines. Because trace amounts of 5-fC and 5-formyluracil (5fU)28 exist in genomic DNA, virtually no FRET signal can be

■

RESULTS AND DISCUSSION Fluorescence Labeling of 5-fC-Containing DNA. The best results were observed when DNA was incubated with 20 μM hydroxylamine-BODIPY in 100 mM NH4OAc (pH 5.0) buffer at 37 °C for 12 h in the presence of anisidine, which acts as a catalyst. The formation of hydroxylamine-BODIPY-labeled DNA without byproducts was observed by HPLC (yield of approximately 80%, Figure 2). Additionally, this corresponding hydroxylamine-BODIPY-labeled DNA that precipitated through ethanol was further verified by MALDI-TOF mass spectrometry (Figure 3). The selectivity of our fluorescence labeling was first demonstrated through PAGE analysis (Figure 1). As expected, no fluorescence band was observed if the DNA had no 5-fC modification (if the DNA had normal cytosine or the 5-hmC modification in Figure 1), whereas an intense fluorescent band was seen if the DNA was modified by 5-fC. Sensitivity of the CCP-FRET-Based Assay. The maximum absorption and emission wavelengths of hydroxylamine-

Figure 1. PAGE analysis of different DNAs after treatment with hydroxylamine-BODIPY under the same conditions (0.5 μM DNA was incubated with 40 μM hydroxylamine-BODIPY and 100 mM anisidine in 100 mM NH4OAc pH 5.0 buffer at 37 °C for 12 h). The top gel was irradiated with 488 nm laser light in the fluorescence mode; the bottom gel was irradiated with UV light. Only DNA with 5fC or DNA with 5-hmC but oxidized to 5-fC by KRuO4 can be labeled by hydroxylamine-BODIPY dye, which shows the high selectivity of our fluorescence labeling method. 10799

dx.doi.org/10.1021/ac4020676 | Anal. Chem. 2013, 85, 10797−10802

Analytical Chemistry

Article

Figure 2. HPLC chromatograms of DNA: (A) control DNA without treatment with hydroxylamine-BODIPY, (B) 5-fC-modification-containing DNA that was reacted with hydroxylamine-BODIPY under the optimal conditions described above. Samples were analyzed by HPLC with a hyamine reversed-phase column equilibrated with buffer A (50 mM triethylammonium acetate (TEAA), pH 6.5) and buffer B (10% 50 mM TEAA, 90% CH3CN).

Figure 3. MALDI-TOF MS spectrum of hydroxylamine-BODIPY-labeled DNA. 5-fC-containing DNA was treated with hydroxylamine-BODIPY under the optimal conditions described above. After incubation, the mixture was ethanol-precipitated with the addition of 10 μL of 1 M sodium acetate (pH 5.0) and 900 μL of cold ethanol, followed by exsiccation and desalting. Hydroxylamine-BODIPY-labeled DNA: calcd, m/z 3710.1; found, m/z 3709.8. These results indicate the successful fluorescence labeling on 5-fC-DNA.

detected. Thus, we applied this CCP-FRET assay to indirectly quantify the relative abundance of 5-hmC in genomic DNA. Through the use of KRuO4, which has a high oxidation efficiency and is easy to handle, almost 95% of the 5-hmC can be selectively oxidized to 5-fC, even in genomic DNA. Thus, this oxidation step was employed before our CCP-FRET-based fluorescence detection. The PAGE analysis of 5-hmCcontaining DNA after labeling is shown in Figure 1; no fluorescent band can be found, similar to the result for the control (DNA without the modification). However, similar to

the 5-fC-containing DNA, the 5-hmC-containing DNA after oxidation can also produce a fluorescent DNA band after being labeled with hydroxylamine-BODIPY; this result demonstrates the feasibility of our method in the quantification of 5-hmC in genomic DNA. 5-hmC is relatively abundant in embryonic stem (ES) cells because of the highly expressed Tet proteins,29,30 which play important roles in the reprogramming of 5-mC and in the control of the differentiation potential in ES cells. Thus, we performed this CCP-FRET-based assay on the mouse 10800

dx.doi.org/10.1021/ac4020676 | Anal. Chem. 2013, 85, 10797−10802

Analytical Chemistry

Article

Figure 4. (a) Normalized fluorescence spectra of different DNAs with the same concentration (200 nM) under the optimal conditions described above: blue line, ssDNA with 5-fC modification; purple line, dsDNA with 5-fC modification; gray line, ssDNA containing an abasic site; green line, ssDNA without 5-fC modification. (b) Spectral separation and signal amplification by CCP. The red line corresponds to the spectrum of 50 nM fluorescently labeled DNA that was irradiated at 492 nm without the addition of CCP, and the black line corresponds to the spectrum of the same DNA as for the red line but irradiated at 380 nm after the addition of CCP. (c) Standard calibration curve: FRET ratio versus different concentrations of 5-fC-containing DNA (0, 25, 50, 75, 100, and 150 nM) (slope 0.47567, intercept 0.1052, r2 = 0.98784). (d−f) 5-hmC quantification in different cell lines: (d) mouse embryonic stem, (e) HeLa, (f) HEK 293T cells. The black line corresponds to the fluorescence spectrum of genomic DNA before oxidation, and the red line corresponds to the fluorescence spectrum of genomic DNA after oxidation. All the measurements were repeated three times.

cerebellum to further validate the utility of our fluorescence labeling method for biological samples. Genomic DNA was first extracted from mouse cerebellum cells and was subjected to sonication to obtain fragments of approximately 200 bp. Then these fragments were concentrated and oxidized by KRuO4, which selectively transformed the 5-hmC into 5-fC in the genomic DNA for further fluorescence labeling. To capture the DNA after the labeling of genomic DNA, 2.5 μM CCP in repeat units was added. A small 5-fC signal was detected before the oxidation because a trace amount of 5-fC compared with the amount of 5-hmC exists in ES cells. However, when the relatively abundant 5-hmC was transformed into 5-fC, a significant increase of the FRET signal was observed (Figure 4d). On the basis of the standard calibration curve obtained previously, approximately 0.04% of 5-hmC in all the nucleotides was detected (0.047% of the total nucleotides14 and about 1.3 × 103 5-hmC molecules per 106 C molecules with a quantitative mass spectrometric assay4 in the previous study). Another two types of human cell lines (HeLa, HEK 293T) that contain even lower levels of 5-hmC were tested to verify the sensitivity and the accuracy of our CCP-FRET-based assay. The 5-hmC level of HeLa cells was determined to be 0.0113% of the total nucleotides (Figure 4e); the result was comparable

to the results reported previously (approximately 0.01% of total nucleotides14). Little increase was observed in the FRET signal from HEK 293T cells after oxidation by KRuO4 (Figure 4f). This observation is consistent with the previous studies that indicated that virtually no 5-hmC exists in HEK 293T cells.14 However, Liu et al. found that the level of 5-hmC in genomic DNA from HEK 293T cells is about 32.5 molecules per 106 nucleosides based on LC−MS/MS/MS coupled with the stable isotope-dilution method.31The amount of 5-hmC in HEK 293T cells is too low for our assay to detect, which requires us to further lower the detection limit.

■

CONCLUSION The CCP-FRET-based assay presented here for the detection of the 5-fC modification benefits from the light-harvesting, large Stokes shift, and optical signal amplification properties of the CCP. Because of the inefficient FRET between the residual fluorescent dyes and the CCP, the assay avoids the cumbersome step of HPLC purification. Furthermore, this simple assay can be applied to the quantification of 5-hmC in genomic DNA after the oxidization from 5-hmC to 5-fC by KRuO4, and this assay does not require affinity enrichment by a biotin lag; these properties simplify the quantification of 5-hmC compared with other reported methods. 10801

dx.doi.org/10.1021/ac4020676 | Anal. Chem. 2013, 85, 10797−10802

Analytical Chemistry

■

Article

(17) Liutkeviciute, Z.; Kriukiene, E.; Grigaityte, I.; Maservicius, V.; Klimasauskas, S. Angew. Chem., Int. Ed. 2011, 50, 2090. (18) Booth, M. J.; Branco, M. R.; Ficz, G.; Oxley, D.; Krueger, F.; Reik, W.; Balasubramanian, S. Science 2012, 336, 934. (19) Yu, M.; Hon, G. C.; Szulwach, K. E.; Song, C. X.; Zhang, L.; Kim, A.; Li, X.; Dai, Q.; Shen, Y.; Park, B.; Min, J. H.; Jin, P.; Ren, B.; He, C. Cell 2012, 149, 1368. (20) Schuler, P.; Miller, A. K. Angew. Chem., Int. Ed. 2012, 51, 10704. (21) Hu, J.; Xing, X.; Xu, X.; Wu, F.; Yan, S.; Xu, Z.; Xu, J.; Weng, X.; Zhou, X. Chem.Eur. J. 2013, 19, 5836. (22) Gaylord, B. S.; Heeger, A. J.; Bazan, G. C. J. Am. Chem. Soc. 2003, 125, 896. (23) Gaylord, B. S.; Massie, M. R.; Feinstein, S. C.; Bazan, G. C. Proc. Natl. Acad. Sci. U.S.A. 2005, 102, 34. (24) Duan, X.; Li, Z.; He, F.; Wang, S. J. Am. Chem. Soc. 2007, 129, 4154. (25) Feng, F.; Liu, L.; Wang, S. Nat. Protoc. 2010, 5, 1255. (26) Feng, F.; Wang, H.; Han, L.; Wang, S. J. Am. Chem. Soc. 2008, 130, 11338. (27) Guo, P.; Yan, S.; Hu, J.; Xing, X.; Wang, C.; Xu, X.; Qiu, X.; Ma, W.; Lu, C.; Weng, X.; Zhou, X. Org. Lett. 2013, 15, 3266. (28) Hong, H.; Wang, Y. Anal. Chem. 2007, 79, 322. (29) Ito, S.; D’Alessio, A. C.; Taranova, O. V.; Hong, K.; Sowers, L. C.; Zhang, Y. Nature 2010, 466, 1129. (30) Koh, K. P.; Yabuuchi, A.; Rao, S.; Huang, Y.; Cunniff, K.; Nardone, J.; Laiho, A.; Tahiliani, M.; Sommer, C.; Mostoslavsky, G.; Lahesmaa, R.; Orkin, S. H.; Rodig, S. J.; Daley, G. Q.; Rao, A. Cell 2011, 8, 200. (31) Liu, S.; Wang, J.; Su, Y.; Guerrero, C.; Zeng, Y.; Mitra, D.; Brooks, P. J.; Fisher, D, E.; Song, H.; Wang, Y. Nucleic Acids Res. 2013, 41, 6421.

ASSOCIATED CONTENT

S Supporting Information *

Synthesis of CCP and the DNA sequences used in this study. 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 †

T.H., T.W., and P.G. contributed equally to this work.

Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS X.Z. thanks the National Basic Research Program of China (973 Program) (Grants 2012CB720600 and 2012CB720603), the National Science Foundation of China (Grant 91213302), the National Grand Program on Key Infectious Disease (Grant 2012ZX10003002-014), and the Program for Changjiang Scholars and Innovative Research Team in University (Grant IRT1030).

■

REFERENCES

(1) Iqbal, K.; Jin, S. G.; Pfeifer, G. P.; Szabo, P. E. Proc. Natl. Acad. Sci. U.S.A. 2011, 108, 3642. (2) Wossidlo, M.; Nakamura, T.; Lepikhov, K.; Marques, C. J.; Zakhartchenko, V.; Boiani, M.; Arand, J.; Nakano, T.; Reik, W.; Walter, J. Nat. Commum. 2011, 2, 241. (3) Maiti, A.; Drohat, A. C. J. Biol. Chem. 2011, 286, 35334. (4) Ito, S.; Shen, L.; Dai, Q.; Wu, S. C.; Collins, L. B.; Swenberg, J. A.; He, C.; Zhang, Y. Science 2011, 333, 1300. (5) Inoue, A.; Shen, L.; Dai, Q.; He, C.; Zhang, Y. Cell Res. 2011, 21, 1670. (6) Guo, J. U.; Su, Y.; Zhong, C.; Ming, G. L.; Song, H. Cell 2011, 145, 423. (7) Iabal, K.; Jin, S. G.; Pfeifer, G. P.; Szabo, P. E. Proc. Natl. Acad. Sci. U.S.A. 2011, 108, 3642. (8) Raiber, E. A.; Beraldi, D.; Ficz, G.; Burgess, H. E.; Branco, M. R.; Murat, P.; Oxley, D.; Booth, M. J.; Reik, P.; Balasubramanian, S. Genome Biol. 2012, 13, 69. (9) Pfaffeneder, T.; Hackner, B.; Truß, M.; Munzel, M.; Muller, M.; Deiml, C. A.; Hagemeier, C.; Carell, T. Angew. Chem., Int. Ed. 2011, 50, 7008. (10) Pastor, W. A.; Pape, U. J.; Huang, Y.; Henderson, H. R.; Lister, R.; Ko, M.; McLoughlin, E. M.; Brudno, Y.; Mahapatra, S.; Kapranov, P.; Tahiliani, M.; Daley, G. Q.; Liu, X. S.; Ecker, J. R.; Milos, P. M.; Agarwal, S.; Rao, A. Nature 2011, 473, 394. (11) Mellen, M.; Ayata, P.; Dewell, S.; Kriaucionis, S.; Heintz, N. Cell 2012, 151, 1417. (12) Szulwach, K. E.; Li, X.; Li, Y.; Song, C. X.; Han, J. W.; Kim, S.; Namburi, S.; Hermetz, K.; Kim, J. J.; Rudd, M. K.; Yoon, Y. S.; Ren, B.; He, C.; Jin, P. PLoS Genet. 2011, 7, 1002154. (13) Szulwac, K. E.; Li, X.; Li, Y.; Song, C. X; Wu, H.; Dai, Q.; Hasan, I.; Upadhyay, A. K.; Gearing, M.; Levey, A. I.; Vasanthakumar, A.; Godley, L. A.; Chang, Q.; Cheng, X.; He, C.; Jin, P. Nat. Neurosci. 2011, 14, 1607. (14) Song, C. X.; Szulwach, K. E.; Fu, Y.; Dai, Q.; Yi, C. Q.; Li, X. K.; Chen, C. H.; Zhang, W.; Jian, X.; Wang, J.; Looney, T.; Zhang, B.; Godley, L. A.; Hicks, L. M.; Lahn, B. T.; Jin, P.; He, C. Nat. Biotechnol. 2011, 29, 68. (15) Pastor, W. A.; Huang, Y.; Henderson, H. R.; Lister, R.; Agarwal, S.; Rao, A. Nat. Protoc. 2012, 7, 1909. (16) Song, C. X.; Clark, T. A.; Lu, X. Y.; Kislyuk, A.; Dai, Q.; Turner, W. S.; He, C.; Korlach, J. Nat. Methods 2011, 9, 75. 10802

dx.doi.org/10.1021/ac4020676 | Anal. Chem. 2013, 85, 10797−10802