Hydrophilic Material for the Selective Enrichment of 5

May 16, 2013 - 5-mC has been commonly recognized as “the fifth base” for decades .... (MRM) using the mass transitions (precursor ions → product...
2 downloads 0 Views 2MB Size
Article pubs.acs.org/ac

Hydrophilic Material for the Selective Enrichment of 5‑Hydroxymethylcytosine and Its Liquid Chromatography−Tandem Mass Spectrometry Detection Yang Tang, Jie-Mei Chu, Wei Huang, Jun Xiong, Xi-Wen Xing, Xiang Zhou, Yu-Qi Feng,* and Bi-Feng Yuan* Key Laboratory of Analytical Chemistry for Biology and Medicine (Ministry of Education), Department of Chemistry, Wuhan University, Wuhan 430072, P.R. China S Supporting Information *

ABSTRACT: 5-Methylcytosine (5-mC), an important epigenetic modification involved in development, can be converted enzymatically to 5-hydroxymethylcytosine (5-hmC). 5-hmC is considered an intermediate of active DNA cytosine demethylation and makes itself serve as an epigenetic mark. 5-hmC content in most mammalian cells is low and the quantification of 5-hmC by liquid chromatography−mass spectrometry (LC−MS) frequently suffers from ion suppression by the presence of unmodified nucleosides. To circumvent this problem, we developed a method to selectively transfer a glucosyl group to the hydroxymethyl moiety of 5-hmC and form a more hydrophilic residue (βglucosyl-5-hydroxymethyl-2′-deoxycytidine, 5-gmdC) by using T4 β-glucosyltransferase. The more hydrophilic 5-gmdC can be selectively enriched by using NH2-silica via hydrophilic interaction prior to liquid chromatography−tandem mass spectrometry (LC−MS/MS) analysis, which eliminates the ion suppression and significantly improves the detection sensitivity and accuracy. Using this method, we successfully quantified 5hmC content in genomic DNA of three human cell lines and seven yeast strains. To the best of our knowledge, this is the first report about the existence of 5-hmC in the model organism of yeast. In addition, the contents of 5-hmC in two yeast strains of Schizosaccharomyces pombe are even higher than those of 5-mC, indicating that 5-hmC may play important roles on the physiological functions of yeast.

C

that the mutations and decreased expression levels of TET genes displayed lower levels of 5-hmC in tumor tissues compared to healthy controls.9−11 Our recent study also revealed that the 5-hmC content in hepatocellular carcinoma tissues were 4−5-fold lower than that in tumor adjacent tissues and demonstrated that the 5-hmC level was highly correlated with tumor stages.12 These findings suggested that 5-hmC in genomic DNA might also be associated with tumor development. 5-hmC content in mammalian cells can be as low as 0.009% of cytosine (molar ratio of 5-hmC/cytosine in 293T cells);13 therefore, highly sensitive detection method is required for the quantitative analysis of 5-hmC content in mammalian genomes. Some methods have been developed for the detection of 5hydroxymethyl-2′-deoxycytidine (5-hmdC) in genomic DNA, including radioactive labeling followed by thin layer chromatography detection,3 immunohistochemistry,14 single-molecule real-time sequencing,15 and LC−mass spectrometry (LC− MS).16,17 The thin layer chromatography method involves

ytosine methylation (5-methylcytosine, 5-mC) at the CpG dinucleotide site is one of the best-characterized epigenetic modifications that play vital roles in a variety of cellular processes, including embryogenesis, regulation of gene expression, genomic imprinting, and X-chromosome inactivation.1 Many human diseases are caused by aberrant DNA cytosine methylation, and properly established and maintained DNA methylation patterns are crucial for the normal functions of cells.2 DNA methylation may undergo dynamic changes and can be reversible in a genome-wide or locus-specific manner. The mechanism of DNA demethylation in mammalian cells remains elusive until the recent discovery of 5-hydroxymethylcytosine (5-hmC) in genomic DNA.3,4 These reports also showed that the Ten−Eleven Translocation (TET) proteins were capable of catalyzing the oxidation of 5-mC to 5-hmC,3,4 indicating that active DNA demethylation may be achieved through the oxidation of 5-mC with the generation of 5-hmC as an intermediate. 5-mC has been commonly recognized as “the fifth base” for decades, and 5-hmC is considered as the sixth base of DNA since its discovery in mammalian cells.5 However, the biological significance of 5-hmC is unclear, though some studies showed that 5-hmC may play important roles on cellular differentiation6 and epigenetic regulation.7,8 Previous reports showed © 2013 American Chemical Society

Received: April 14, 2013 Accepted: May 16, 2013 Published: May 16, 2013 6129

dx.doi.org/10.1021/ac4010869 | Anal. Chem. 2013, 85, 6129−6135

Analytical Chemistry

Article

Figure 1. (A) The hydroxyl group of 5-hmdC in double-stranded DNA can be glucosylated by β-GT to form β-glucosyl-5-hydroxymethyl-2′deoxycytidine (5-gmdC) using UDP-glucose as a cofactor. (B) The schematic diagram for the quantification of 5-hmC content in genomic DNA by SPE-LC−MS/MS method.

2′-deoxycytidine, 5-gmdC) (Figure 1A), to develop a novel method for effective enrichment and detection of 5-hmC. After glucosylation, the 5-gmdC-containing DNA was enzymatically hydrolyzed to nucleosides. The 5-gmdC in the nucleoside mixture was then selectively enriched from other nonglucosylated nucleosides using NH2-silica via hydrophilic interaction followed by LC−MS/MS analysis, which eliminates the ion suppression from other bulk normal nucleosides and therefore significantly improved the detection sensitivity and accuracy during mass spectrometry analysis (Figure 1B). Using this method, we successfully determined the content of 5-hmC in genomic DNA of three human cell lines (HeLa, 293T, and Jurkat-T cell lines) and seven yeast strains.

labeling with radioactive isotope, and the accuracy is not comparable to those of other available methods. Immunohistochemical staining method is tedious and, to some extent, less quantitative. HPLC analysis relies heavily on the chromatographic separation to avoid coelution with other components. The measurement of 5-hmC by single-molecule real-time sequencing is possible, but the technology still needs to be improved for accurate quantitation. Reversed-phase liquid chromatography (RPLC) coupled with tandem mass spectrometry (MS/MS) has been employed for the analysis of 5-hmC.13,16,17 However, the analysis of targets by MS is frequently accompanied with interference, especially those arising from ion suppression effects induced by other compounds.18 Chromatographic separation prior to MS analysis can reduce the interference between analytes, but the baseline separation of normal and modified nucleosides (rA, rC, rG, rU, dA, dC, dG, T, 5-mdC, 5-hmdC) by a C18 column is not always satisfactory and these nucleosides frequently coelute or overlap with each other. In addition, a large amount of DNA sample is typically required for the quantification of a low content of 5-hmC, which will broaden the peaks of nucleosides and further deteriorate the separation of nucleosides. In this respect, effective extraction of a low level of 5-hmdC from bulk normal nucleosides before LC−MS/MS analysis will benefit the accurate quantification. In the current study, we take advantage of T4 phage βglucosyltransferase (β-GT), which can selectively transfer a polar glucosyl group to the hydroxymethyl group of 5-hmC and form a more hydrophilic residue (β-glucosyl-5-hydroxymethyl-



EXPERIMENTAL SECTION Chemicals and Reagents. 2′-Deoxycytidine (dC), 2′deoxyguanosine (dG), 2′-deoxyadenosine (dA), thymidine (T), cytidine (rC), guanosine (rG), adenosine (rA), uridine (rU), and 5-azacytidine (5-azaC) were purchased from Sigma-Aldrich (Beijing, China), and 5-hydroxymethyl-2′-deoxycytidine (5hmdC) was purchased from Berry & Associates (Dexter, MI). Chromatographic grade methanol was purchased from Merck (Darmstadt, Germany). All other solvents and chemicals used were of analytical grade. Acetonitrile (ACN), glucose, formic acid, aqueous ammonia (25%, by weight), sodium acetate, sodium chloride (NaCl), ethylenediaminetetraacetic acid (EDTA), sodium dodecylsulfate (SDS), and chloroform were purchased from Sinopharm Chemical Reagent Co., Ltd.

6130

dx.doi.org/10.1021/ac4010869 | Anal. Chem. 2013, 85, 6129−6135

Analytical Chemistry

Article

complementary strand, 5′-GGTACCTGCGGCTTAAG-3′) according to a previously reported protocol24 followed by treatment with S1 nuclease to remove single-stranded ODN. The double-stranded DNA was then desalted with a 3 kDa cutoff Ultra-0.5 membrane (Amicon, Millipore) and quantified with spectrophotometer B-500. The prepared double-stranded DNA was mixed with different amounts of single-stranded ODN (5′-GGTACCTGCGGCTTAAG-3′) to render the molar ratios of 5-gmdC/dG ranging from 0.0002% to 0.005%. The above DNA mixture was incubated with β-GT (10 U) in a 40-μL reaction solution containing 2 nmol of UDPglucose at 37 °C for 1 h followed by digestion with S1 nuclease and alkaline phosphatase following previously reported procedures.12 As for the analysis of 5-hmC in genomic DNA, the genomic DNA (50 μg) was fragmented into several hundred base pairs by sonication and then treated with β-GT (10 U) in a 40-μL reaction solution containing 2 nmol of UDP-glucose at 37 °C for 1 h. The obtained DNA was subsequently digested to nucleosides by S1 nuclease and alkaline phosphatase according to a previous report.12 The resulting nucleoside mixture was extracted with phenol/chloroform once and chloroform once to remove enzymes followed by drying under vacuum at 37 °C. The sample was reconstituted in water and then passed through the SPE filled with NH2-silica to extract 5-gmdC. The content of washing and desorption solution were optimized to obtain the best selectivity and recovery toward 5-gmdC. 5-azaC (internal standard) was added to the extracted 5-gmdC and then analyzed by LC−MS/MS. Analysis of the Genome-Wide 5-hmC in Mammalian and Yeast Cells. The nucleoside mixture from DNA digestion was reconstituted in water and then divided into two fractions (1/10 and 9/10, v/v). The minor fraction was directly subjected to HPLC analysis to quantify dG, and the major fraction was analyzed by the SPE-LC−MS/MS method to quantify 5-gmdC. The genome-wide 5-hmC can be calculated through the measurement of molar ratios of 5-gmdC/dG.25 Analysis of nucleosides was performed on the HPLC−ESIMS/MS system consisting of an AB 3200 QTRAP LC−MS/ MS (Applied Biosystems, Foster City, CA) with an electrospray ionization source (Turbo Ionspray) and a Shimadzu LC-20AD HPLC (Tokyo, Japan) with two LC-20AD pumps, a SIL-20A autosampler, a CTO-20AC thermostatted column compartment, and a DGU-20A3 degasser. Data acquisition and processing were performed using AB SCIEX Analyst 1.5 Software (Applied Biosystems, Foster City, CA). The HPLC separation was performed on a Hisep C18-T column (150 mm × 2.1 mm i.d., 5 μm, Weltech Co., Ltd., Wuhan, China) with a flow rate of 0.2 mL/min at 35 °C. Formic acid in water (0.01%, v/v, solvent A) and a mixture of 0.01% formic acid in methanol (v/v, solvent B) were employed as mobile phase. A gradient of 5 min 5% B, 20 min 5−50% B, 5 min 50% B, and 10 min 5% B was used. The mass spectrometry detection was performed under positive electrospray ionization mode. The target nucleosides were monitored by multiple reaction monitoring (MRM) using the mass transitions (precursor ions → product ions) of dC (228.4 → 112.2), T (243.3 → 127.2), dA (252.4 → 136.2), dG (268.4 → 152.4), rC (244.4 → 112.2), rU (245.4 → 113.1), rA (268.4 → 136.2), rG (284.5 → 152.2), 5-hmdC (258.2 → 142.1), 5-gmdC (420.2 → 142.2), 5-azaC (245.2 → 113.1). The MRM parameters of all nucleosides were optimized to achieve maximal detection sensitivity.

(Shanghai, China). Triton X-100 was purchased from Aladdin Reagent Co., Ltd. (Shanghai, China). Tris buffer-saturated phenol, Tris-HCl, and glass beads (425−600 μm) were purchased from Wenhan Technology Co., Ltd. (Wuhan, China). Yeast extract, peptone, and malt extract were purchased from BD Company (Franklin Lakes, NJ). Recombinant T4 phage β-glucosyltransferase (β-GT) was obtained from the New England Biolabs (Ipswich, MA). S1 nuclease and alkaline phosphatase (CIAP) were from Takara Biotechnology Co., Ltd. (Dalian, China). Phosphodiesterase I was purchased from Sigma-Aldrich (Beijing, China). The water used throughout the study was purified by using a Milli-Q apparatus (Millipore, Bedford, MA). Stock solutions for the aforementioned nucleosides were prepared in Milli-Q water at a concentration of 200 μg/mL. Cell Culture. HeLa (cervical carcinoma), Jurkat-T, and 293T cell lines were obtained from the China Center for Type Culture Collection (CCTCC). HeLa and 293T cells were maintained in DMEM medium at 37 °C under 5% CO2 atmosphere, and Jurkat-T cells were maintained in RPMI 1640 medium at 37 °C under 5% CO2 atmosphere. These media were all supplemented with 10% fetal bovine serum, 100 U/mL penicillin, and 100 μg/mL streptomycin (Gibco). The Saccharomyces cerevisiae strains used in this study include the YEF473A (MATa his3 leu2 lys2 trp1 ura3)19 in S288C background and the W1588-4C (MATa leu2-3,112 ura3-1 his3-11,15 trp1-1 ade2-1 can1-100)20 in W303 background. The two yeast strains Debaryomyces hansenii (CGMCC 2.494) and Yarrowia lipolytica (CGMCC 2.1718) were purchased from the China General Microbiological Culture Collection Center (CGMCC). The five yeast strains Kluyveromyces lactis (CICC 1773), Kluyveromyces marxianus (CICC 1953), Pichia pastoris (CICC 1958), Schizosaccharomyces pombe (CICC 1056), and Schizosaccharomyces pombe (CICC 1372) were purchased from the China Center of Industrial Culture Collection (CICC). All the yeast strains were cultured for 12 h according to a previously reported procedure.21 DNA Extraction. Mammalian genomic DNA from the cultured cells was extracted according to the protocol G described in a previous report.22 Yeast genomic DNA was extracted according to a previously described method with slight modification.23 Briefly, yeast cells were suspended in 1 mL of cell lysis buffer (10 mM Tris-HCl, pH 8.0, 1 mM EDTA, 100 mM NaCl, 1% SDS, and 2% Triton X-100) plus an equal volume of phenol/chloroform (1:1, v/v, saturated with 10 mM Tris, pH 8.0, 1 mM EDTA) and 1.5 g of glass beads (425−600 μm) and rigorously vortexed for 10 min. To the solution was added 1 mL of TE buffer (10 mM Tris-HCl, 1 mM EDTA, pH 8.0). The resulting solution was centrifuged at 13 500g for 5 min, and the aqueous layer was transferred into a clean tube. An equal volume solvent mixture of phenol/chloroform was added, and the sample solution was vortexed and centrifuged at 13 500g for 5 min. The aqueous layer was subsequently transferred into a clean tube, and genomic DNA was precipitated by isopropanol. DNA concentrations were determined using spectrophotometer B-500 (Metash Instruments Co., Ltd., Shanghai, China). Sample Preparation and Solid-Phase Extraction (SPE). The quality control (QC) samples with the molar ratios of 5gmdC/dG ranging from 0.0002% to 0.005% were prepared to validate the method. Briefly, a double-stranded DNA was prepared by annealing two complementary ODNs (5′CTTAAGCCGXAGGTACC-3′, X represents the 5-hmdC; 6131

dx.doi.org/10.1021/ac4010869 | Anal. Chem. 2013, 85, 6129−6135

Analytical Chemistry

Article

The quantification of dG in the digestion mixture was performed on a Hisep C18-T column (250 mm × 4.6 mm i.d., 5 μm, Weltech) using an Agilent 1200 HPLC system (Agilent Technologies, Palo Alto, CA). Ammonium acetate in water (10 mM, pH 6.5, solvent A) and methanol (solvent B) were employed as the mobile phase. An isocratic elution with 90% A and 10% B was used, and the flow rate was set at 1 mL/min. Calculation of the Percentage of 5-hmC. The genomewide content of 5-hmC was calculated using the following formula: 5‐hmC% =

M5‐gmdC MdG

× 100%

where M5‑gmdC is the molar quantity of 5-gmdC, and MdG is the molar quantity of 2′-deoxyguanosine determined in the DNA sample.



RESULTS AND DISCUSSION Effect of Nucleoside Mixture on MS Response of 5hmdC. In this study, we employed RPLC to separate 10 nucleosides. Shown in Figure 2A is the MRM chromatogram for the analysis of a mixture of 10 standard nucleosides. Because of the similar chemical structures, 5-hmdC, dC, rC, and rU exhibit very similar retention time. Considering that the content of 5-hmdC can be as low as 0.009% of cytosine,13 the microgram level of genomic DNA is typically used for the quantification of genome-wide 5-hmdC with the LC−MS-based method. However, peak tailing is normally present with a large amount of sample loading, and ion suppression from the coelutes can dramatically affect the accurate quantification of target analytes using the LC−MS method.18 Thus, we first investigated the potential ion suppression of 5-hmdC resulting from the presence of other nucleosides by adding different amounts of nucleoside mixture to 5-hmdC. The results showed that the signal response of 5-hmdC was significantly suppressed in the presence of 5−50 μg of unmodified nucleosides (Figure 2B,C). In addition, we mixed 5-hmdC together with other nucleosides (rG, rA, dG, dA, T, and 5-azaC) with retention times longer than 5-hmdC to examine whether these nucleosides will suppress the ionization of 5-hmdC. Our results showed that these nucleosides with longer retention times than 5-hmdC did not suppress its ionization (data not shown), suggesting that the ion suppression is mainly from the nucleosides eluting before 5-hmdC. To eliminate the potential ion suppression of 5-hmdC from other nucleosides, others employed off-line HPLC to isolate the 5-hmdC from the nucleoside mixture prior to LC−MS/MS analysis,13 which was however laborious and time-consuming. Isotope dilution mass spectrometry can compensate for ion suppression, but the isotope-labeled 5-hmdC is not commercially available and its synthesis is tedious and challenging. SPE-LC−MS/MS Method. T4 phage β-glucosyltransferase (β-GT) can selectively catalyze the transfer of a glucose moiety from uridine diphosphoglucose (UDP-glucose) to the hydroxyl group in the nucleobase portion of 5-hmdC, yielding βglucosyl-5-hydroxymethyl-2′-deoxycytidine (5-gmdC) in double-stranded DNA (Figure 1A). We took advantage of this property of β-GT and developed a SPE-LC−MS/MS method for the accurate and sensitive quantification of genome-wide 5hmC (Figure 1B). In this respect, the glucosyl group was transferred to 5-hmdC to form 5-gmdC, which can be selectively enriched by NH2-silica via hydrophilic interaction.

Figure 2. Effect of ion suppression from normal nucleosides on 5hmdC: (A) MRM chromatogram of 10 nucleoside standards, (B) MRM chromatogram of 5-hmdC spiked in various amounts of nucleoside mixture, (C) calculated peak area ratio of 5-hmdC spiked in various amounts of nucleoside mixture over 5-hmdC control (without adding nucleoside mixture) from part B.

The extracted 5-gmdC was then subjected to LC−MS/MS analysis. We evaluated the conversion rate of 5-hmdC to 5-gmdC by analyzing the substrates and products of the enzymatic reaction with matrix-assisted laser desorption ionization-time-of-flightmass spectrometry (MALDI-TOF-MS). The results showed that β-GT can convert 5-hmdC residues to 5-gmdC at nearly 100% efficiency (Figure S1 of Supporting Information), which is consistent with previous report showing that β-GT can quantitatively glucosylate 5-hmdC in double-stranded DNA.24 We further assessed the extraction selectivity and efficiency of NH2-silica toward 5-gmdC by hydrophilic interaction. To this end, 5-gmdC standard was prepared and the detailed procedures were described in the Supporting Information. In total, 100 fmol of 5-gmdC was added into 50−100 μg of the nucleoside mixture and then loaded onto NH2-silica SPE cartridges (200 mg). After washing with 2 mL of H2O/ACN (8:2, v/v) containing 0.25% ammonia (by weight), the captured nucleosides were eluted using 0.6 mL of H2O/ACN (3:7, v/v). The results showed that the recoveries of 5-gmdC were all 6132

dx.doi.org/10.1021/ac4010869 | Anal. Chem. 2013, 85, 6129−6135

Analytical Chemistry

Article

higher than 80% in different amounts of nucleoside mixture (Figure 3). However, the recoveries of other nucleosides

Figure 3. Recoveries of nucleosides by NH2-silica SPE. In total, 100 fmol of 5-gmdC was added into 50−100 μg of nucleoside mixture and then extracted by NH2-silica SPE. The recoveries were calculated by comparing the amount of nucleosides before and after SPE enrichment. Figure 4. Evaluation of the performance of NH2-silica SPE on the extraction of 5-gmdC. (A) Sample (100 fmol of 5-gmdC and 50 μg of nucleosides mixture) was directly analyzed by LC−MS/MS; (B) sample (100 fmol of 5-gmdC and 50 μg of nucleosides mixture) was enriched by NH2-silica SPE prior to LC−MS/MS analysis. The 5gmdC standard represents the direct analysis of 100 fmol of 5-gmdC by LC−MS/MS.

carrying no glucosyl group were below 7% except for rG and dG. This does not influence the LC−MS/MS detection of 5gmdC since the retention times of rG and dG are very different from that of 5-gmdC (Figure 2A). In addition, the MRM chromatogram showed that the signal intensity of 5-gmdC after enrichment by NH2-silica was almost the same as the control sample (without adding the nucleoside mixture); however, the signal intensity of 5-gmdC was much weaker without extraction by NH2-silica (Figure 4). The results demonstrated that the conversion of 5-hmdC to 5-gmdC followed by extraction with NH2-silica could eliminate the ion suppression from other nucleosides during LC−MS/MS analysis. In addition, we examined other two types of hydrophilic materials including silica and PSA (primary secondary amine-silica) for the enrichment of 5-gmdC. Under their optimal conditions, NH2silica showed the best performance for the selective enrichment of 5-gmdC among silica, PSA, and NH2-silica (data not shown); therefore, we chose NH2-silica to enrich 5-gmdC. Method Validation. The calibration curve of 5-gmdC was constructed by plotting the mean peak area ratio of 5-gmdC/5azaC (I.S.) versus the mean molar amount of 5-gmdC based on data obtained from triplicate measurements using the SPE-LC− MS/MS method. As for dG, the calibration curve was constructed by plotting the mean peak area of dG versus the mean molar amount of dG. The results showed that good linearities within the range of 5−2000 fmol of 5-gmdC and 0.015−15 nmol dG were obtained with a coefficient values (R) being great than 0.99 (Tables S1 and S2, Supporting Information). The limit of detection (LOD) and the limit of quantification (LOQ), defined as the amounts of the analytes at a signal-to-noise ratio (S/N) of 3 and 10, respectively, were 1.5 and 5.0 fmol for 5-gmdC and 6.7 and 20.0 fmol for 5-hmdC (Table S2, Supporting Information). After glucosylation, the LOD of 5-gmdC was improved by more than 4-fold compared to 5-hmdC, which can further enhance the detection sensitivity. The reproducibility of the developed method was evaluated by the measurement of intra- and interday precisions. The intra- and interday relative standard deviations (RSDs) were calculated with different amounts of 5-gmdC spiked in 50 μg of nucleoside mixture (Table 1). Three parallel SPE treatments of

samples over a day gave the intraday RSDs, and the interday RSDs were determined by treating samples independently for 3 consecutive days. The results showed that the intra- and interday RSDs were less than 4.5% and 7.5%, respectively (Table 1), demonstrating that good reproducibility was achieved. The accuracy of the proposed method was assessed using the QC samples by comparing the measured 5-gmdC content to the theoretical 5-gmdC content. Three different molar ratios of 5-gmdC/dG ranging from 0.0002 to 0.005% were measured (Table 2). The results showed that good reproducibility and accuracy could be achieved, which are manifested by RSDs and relative errors being less than 11.3% and 13.0%, respectively (Table 2). Quantification of Genome-Wide 5-hmC in Human Cell Lines and Yeast Strains. We determined the genome-wide 5hmC in human cell lines and yeast strains by employing the SPE-LC−MS/MS method. The signal intensity of 5-gmdC obtained by the SPE-LC−MS/MS method is much stronger (by ∼50-fold) than that of 5-hmdC obtained by direct analysis of the nucleoside mixture from 50 μg of genomic DNA (Figure 5). The stronger signal intensity obtained with the SPE-LC− MS/MS method is due to the fact that ion suppression from normal nucleosides was removed by extracting 5-gmdC from the nucleoside mixture. In addition, the LOD of 5-gmdC is approximately 4.5-fold lower than that of 5-hmdC (Table S2, Supporting Information), which also contributes to the increase of the MS response. Using this method, we quantified the levels of 5-hmC in genomic DNA of HeLa, 293T, and Jurkat-T cells, which were found to be 0.0083%, 0.014%, and 0.0004% (relative to dG), respectively (Table 3). The 5-hmC content measured in 293T 6133

dx.doi.org/10.1021/ac4010869 | Anal. Chem. 2013, 85, 6129−6135

Analytical Chemistry

Article

Table 1. Precisions (Intra- and Inter-Day) and Recoveries for the Determination of 5-gmdC at Three Different 5-gmdC/dG Molar Ratios Using 50 μg of Nucleoside Mixturea recovery (%, n = 3)

a

intraday (RSD %, n = 3)

interday (RSD %, n = 3)

analyte

low 0.0001%

medium 0.0005%

high 0.0010%

L

M

H

L

M

H

5-gmdC

84.9 ± 2.6

83.8 ± 3.7

88.2 ± 1.35

3.1

4.5

1.5

5.3

4.1

7.5

Recoveries are given as average value ± standard deviation of results from triplicate analysis.

Table 2. Accuracy of the SPE-LC−MS/MS Method for the Determination of 5-gmdC QCs

theoretical value (%)

measured value (%)

RSD % (n = 3)

relative error (%)

low medium high

0.000 20 0.001 00 0.005 00

0.000 22 0.001 12 0.005 36

7.9 4.6 11.3

13.0 10.4 6.7

Table 3. 5-hmC Contents in Genomic DNA of Human Cell Lines and Yeast Strainsa cells HeLa 293T Jurkat-T Debaryomyces hansenii Kluyveromyces lactis Kluyveromyces marxianus Pichia pastoris Schizosaccharomyces pombe (CICC 1056) Schizosaccharomyces pombe (CICC 1372) Yarrowia lipolytica

5-gmdC/dG % 0.0083 0.0140 0.0004 0.0088 0.0004 0.0031 0.0750 0.3447 0.2587 0.0642

± ± ± ± ± ± ± ± ± ±

0.0011 0.0034 0.0001 0.0003 0.0001 0.0002 0.0051 0.0491 0.0165 0.0199

Values are given as average value ± standard deviation from triplicate measurements; 5-hmC content in genomic DNA was expressed as 5gmdC/dG %.

a

(0.25−0.34% of dG) were found in the two Schizosaccharomyces pombe strains than in other yeast strains. The 5-hmC contents are even higher than 5-mC contents21 in these two strains, which may indicate that 5-hmC plays important roles in the physiological functions of yeast.



CONCLUSION In the current study, we developed a SPE-LC−MS/MS method for the sensitive and accurate determination of 5-hmC content in genomic DNA. 5-hmdC was converted to 5-gmdC using βGT followed by selectively extracting 5-gmdC from the nucleoside mixture with NH2-silica SPE. With this method, ion suppression was eliminated during LC−MS/MS analysis, which enables sensitive and accurate quantification of a low content of 5-hmC in the presence of bulk normal nucleosides. We then successfully quantified 5-hmC content in genomic DNA of three human cell lines as well as seven yeast strains, which is the first report about the existence of 5-hmC in the model organism of yeast.

Figure 5. Analysis of 5-hmC content in genomic DNA of HeLa cells using two different methods: (A) nucleoside mixture from digestion of 50 μg of DNA was directly analyzed by LC−MS/MS and (B) nucleoside mixture from digestion mixture of 50 μg of DNA was enriched by NH2-silica SPE prior to LC−MS/MS analysis.



cells is similar with the previous report of 0.009%.13 However, the tedious off-line HPLC purification of 5-hmdC, which may take days, is required before LC−MS/MS analysis.13 Song et al. reported that 5-hmC content in HeLa cells and 293T cells was 0.007% (versus total nucleosides) and 0.009% (versus total nucleosides) using the dot-blot assay.26 The higher values they obtained may be due to the different methods used. In addition, the sensitive SPE-LC−MS/MS method allows us to quantify an extremely low content of 5-hmC (0.0004%) in Jurkat-T cells. We recently reported the widespread presence of 5-mC in yeast strains.21 Nevertheless, it has not been examined whether 5-hmC exists in yeast DNA. Here, we further examined the genome-wide 5-hmC in yeast using SPE-LC−MS/MS method. The results showed that 5-hmC is also present in yeast strains in the levels ranging from 0.0004% to 0.3447% (relative to dG) (Table 3). Interestingly, much higher contents of 5-hmC

ASSOCIATED CONTENT

S Supporting Information *

Additional information as noted in text. This material is available free of charge via the Internet at http://pubs.acs.org.



AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected] (B.-F.Y.); [email protected] (Y.-Q.F.). Phone: +86-27-68755595. Fax: +86-27-68755595. Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS The authors thank the financial support from the National Basic Research Program of China (973 Program) (Grants 2012CB720601, 2013CB910702, and 2012CB720603), the 6134

dx.doi.org/10.1021/ac4010869 | Anal. Chem. 2013, 85, 6129−6135

Analytical Chemistry

Article

National Natural Science Foundation of China (Grants 91217309, 91017013, 31070327, 21205091, and 21228501), Ph.D. Programs Foundation of Ministry of Education of China (Grant 20120141120037), the Fundamental Research Funds for the Central Universities, and the Natural Science Foundation of Hubei Province (Grant 2011CDB440). We thank Prof. Yinsheng Wang (University of California, Riverside) for helpful comments on our manuscript.



REFERENCES

(1) Bird, A. Gene. Dev. 2002, 16, 6−21. (2) Rottach, A.; Leonhardt, H.; Spada, F. J. Cell. Biochem. 2009, 108, 43−51. (3) Kriaucionis, S.; Heintz, N. Science 2009, 324, 929−930. (4) Tahiliani, M.; Koh, K. P.; Shen, Y.; Pastor, W. A.; Bandukwala, H.; Brudno, Y.; Agarwal, S.; Iyer, L. M.; Liu, D. R.; Aravind, L.; Rao, A. Science 2009, 324, 930−935. (5) Münzel, M.; Globisch, D.; Carell, T. Angew. Chem., Int. Ed. 2011, 50, 6460−6468. (6) Ito, S.; D’Alessio, A. C.; Taranova, O. V.; Hong, K.; Sowers, L. C.; Zhang, Y. Nature 2010, 466, 1129−1133. (7) Kriukienė, E.; Liutkevičiu̅tė, Z.; Klimašauskas, S. Chem. Soc. Rev. 2012, 41, 6916−6930. (8) Branco, M. R.; Ficz, G.; Reik, W. Nat. Rev. Genet. 2012, 13, 7−13. (9) Yang, H.; Liu, Y.; Bai, F.; Zhang, J. Y.; Ma, S. H.; Liu, J.; Xu, Z. D.; Zhu, H. G.; Lin, Z. Q.; Ye, D.; Guan, K. L.; Xiong, Y. Oncogene 2012, 32, 663−669. (10) Ko, M.; Huang, Y.; Jankowska, A. M.; Pape, U. J.; Tahiliani, M.; Bandukwala, H. S.; An, J.; Lamperti, E. D.; Koh, K. P.; Ganetzky, R.; Liu, X. S.; Aravind, L.; Agarwal, S.; Maciejewski, J. P.; Rao, A. Nature 2010, 468, 839−843. (11) Kudo, Y.; Tateishi, K.; Yamamoto, K.; Yamamoto, S.; Asaoka, Y.; Ijichi, H.; Nagae, G.; Yoshida, H.; Aburatani, H.; Koike, K. Cancer Sci. 2012, 103, 670−676. (12) Chen, M. L.; Shen, F.; Huang, W.; Qi, J. H.; Wang, Y.; Feng, Y. Q.; Liu, S. M.; Yuan, B. F. Clin. Chem. 2013, 59, 824−832. (13) Ito, S.; Shen, L.; Dai, Q.; Wu, S. C.; Collins, L. B.; Swenberg, J. A.; He, C.; Zhang, Y. Science 2011, 333, 1300−1303. (14) Haffner, M. C.; Chaux, A.; Meeker, A. K.; Esopi, D. M.; Gerber, J.; Pellakuru, L. G.; Toubaji, A.; Argani, P.; Iacobuzio-Donahue, C.; Nelson, W. G.; Netto, G. J.; De Marzo, A. M.; Yegnasubramanian, S. Oncotarget 2011, 2, 627−637. (15) Song, C. X.; Clark, T. A.; Lu, X. Y.; Kislyuk, A.; Dai, Q.; Turner, S. W.; He, C.; Korlach, J. Nat. Methods 2012, 9, 75−77. (16) Le, T.; Kim, K. P.; Fan, G.; Faull, K. F. Anal. Biochem. 2011, 412, 203−209. (17) Jin, S. G.; Jiang, Y.; Qiu, R.; Rauch, T. A.; Wang, Y.; Schackert, G.; Krex, D.; Lu, Q.; Pfeifer, G. P. Cancer Res. 2011, 71, 7360−7365. (18) Annesley, T. M. Clin. Chem. 2003, 49, 1041−1044. (19) Bi, E.; Pringle, J. R. Mol. Cell. Biol. 1996, 16, 5264−5275. (20) Ko, N.; Nishihama, R.; Tully, G. H.; Ostapenko, D.; Solomon, M. J.; Morgan, D. O.; Pringle, J. R. Mol. Biol. Cell 2007, 18, 5139− 5153. (21) Tang, Y.; Gao, X. D.; Wang, Y.; Yuan, B. F.; Feng, Y. Q. Anal. Chem. 2012, 84, 7249−7255. (22) Ravanat, J.-L.; Douki, T.; Duez, P.; Gremaud, E.; Herbert, K.; Hofer, T.; Lasserre, L.; Saint-Pierre, C.; Favier, A.; Cadet, J. Carcinogenesis 2002, 23, 1911−1918. (23) Hoffman, C. S.; Winston, F. Gene 1987, 57, 267−272. (24) Robertson, A. B.; Dahl, J. A.; Vagbo, C. B.; Tripathi, P.; Krokan, H. E.; Klungland, A. Nucleic Acids Res. 2011, 39, e55. (25) Münzel, M.; Globisch, D.; Brückl, T.; Wagner, M.; Welzmiller, V.; Michalakis, S.; Müller, M.; Biel, M.; Carell, T. Angew. Chem., Int. Ed. 2010, 49, 5375−5377. (26) Song, C. X.; Szulwach, K. E.; Fu, Y.; Dai, Q.; Yi, C.; Li, X.; Li, Y.; Chen, C. H.; Zhang, W.; Jian, X.; Wang, J.; Zhang, L.; Looney, T. J.; Zhang, B.; Godley, L. A.; Hicks, L. M.; Lahn, B. T.; Jin, P.; He, C. Nat. Biotechnol. 2011, 29, 68−72. 6135

dx.doi.org/10.1021/ac4010869 | Anal. Chem. 2013, 85, 6129−6135