Detection and separation of nucleoside-5'-monophosphates of DNA by conjugation with the fluorescent...

Michael Cornelius*Christian G. C. T. WörthHans-Christian KliemManfred WiesslerHeinz H. Schmeiser

Division of MolecularToxicology,German Cancer ResearchCenter,Heidelberg, Germany

Detection and separation of nucleoside-5’-monophosphates of DNA by conjugation with thefluorescent dye BODIPY and capillaryelectrophoresis with laser-induced fluorescencedetection

We investigated the separation and detection of the 5’-monophosphates of 2’-de-oxynucleosides selectively conjugated with 4,4-difluoro-5,7-dimethyl-4-bora-3a,4a-diaza-s-indacene-3-propionyl ethylene diamine hydrochloride (BODIPY FL EDA) atthe 5’-phosphate group using capillary electrophoresis with laser-induced fluores-cence detection (CE-LIF). BODIPY conjugates of the four common deoxynucleoside-5’-monophosphates (2’-deoxyguanosine-5’-monophosphate, 2’-deoxyadenosine-5’-monophosphate, 2’-deoxycytidine-5’-monophosphate, and thymidine-5’-monophos-phate) were prepared and subjected to CE-LIF to serve as standard compounds forpeak assignment and to develop separation conditions for the analysis of DNA.BODIPY conjugates were detected and resolved by CE-LIF after digestion of DNA oran oligonucleotide to 5’-monophosphates by nuclease P1 (NP 1) and fluorescencelabeling without further purification step. Comparative analyses of calf-thymus DNAdigested either with micrococcal nuclease/spleen phosphodiesterase to 3’-mono-phosphates or with NP 1 to 5’-monophosphates showed that both versions of thefluorescence postlabeling assay were equally efficient and sensitive. Moreover, usingthe same assay, 2’-deoxyuridine and 2’-deoxy-5methylcytidine were identified inbisulfite treated DNA after NP 1 digestion indicating that fluorescence postlabeling of2’-deoxyribonucleoside-5’-monophosphates with BODIPY FL EDA and detection byCE-LIF has the potential to determine DNA damage and genomic DNA methylation.

Keywords: BODIPY postlabeling / Capillary electrophoresis laser-induced fluorescence /Deoxynucleotides / DNA adducts / 5-Methylcytosine DOI 10.1002/elps.200410405

1 Introduction

Numerous studies have used CE methods for the analysisof nucleotides because they are easily separated due totheir negative charge in a pH range from 2 to 12 [1]. SinceUV detection was used for most CE separations ofnucleotides, detection limits were typically in the micro-molar range and thus the technique was not sensitiveenough for applications in biological matrices. Fluores-cence, especially laser-induced fluorescence (LIF) meth-ods, were found to be very successful in decreasingdetection limits for nucleotide analyses. Following thefluorescence postlabeling approach pioneered bySharma [2] and Giese [3, 4] we recently introduced thefluorescent dye 4,4-difluoro-5,7-dimethyl-4-bora-3a,4a-

Correspondence: Dr. Heinz H. Schmeiser, Division of MolecularToxicology, German Cancer Research Center, Im NeuenheimerFeld 280, D-69120 Heidelberg, GermanyE-mail: [email protected]: 149-6221-423375

Abbreviations: BODIPY FL EDA, 4,4-difluoro-5,7-dimethyl-4-bora-3a,4a-diaza-s-indacene-3-propionyl ethylene diamine hy-drochloride; CT-DNA, calf-thymus DNA; dAp, 2’-deoxyadeno-sine-3’-monophosphate; dCp, 2’-deoxycytidine-3’-monopho-sphate; dGp, 2’-deoxyguanosine-3’-monophosphate; dNp, 2’-deoxyribonucleoside-3’-monophosphates; EDC, 1-ethyl-3-(3’-N,N’-dimethyl-aminopropyl)-carbodiimide hydrochloride;5mdCp, 2’-deoxy-5-methylcytidine-3’-monophosphate; MN,micrococcal nuclease; NP 1, nuclease P1; pdA, 2’-deoxyadeno-sine-5’-monophosphate; pdC, 2’-deoxycytidine-5’-monopho-sphate; pdG, 2’-deoxyguanosine-5’-monophosphate; pdI, 2’-deoxyinosine-5’-monophosphate; pdN, 2’-deoxyribonucleo-side-5’-monophosphate; pdU, 2’-deoxyuridine-5’-monopho-sphate; p5mdC, 2’-deoxy-5-methylcytidine-5’-monophosphate;pT, thymidine-5’-monophosphate; SPD, spleen phosphodiester-ase; Tp, thymidine-3’-monophosphate

Electrophoresis 2005, 26, 2591–2598 2591

* Sponsored by the scholarship programme of the German Fed-eral Environmental Foundation (DBU)

2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

Nuc

leic

acid

s

2592 M. Cornelius et al. Electrophoresis 2005, 26, 2591–2598

diaza-s-indacene-3-propionyl ethylene diamine hydro-chloride (BODIPY FL EDA) as derivatization reagent forthe detection of modified and normal deoxynucleotides[5]. This dye binds covalently by its amino linker to thephosphate group of a deoxynucleotide once the latter hasbeen activated with a carbodiimide reagent. The dye-labeled deoxynucleoside-monophosphates then aredetected with high sensitivity (8.5 pmol/L) based on theirfluorescence by MEKC with LIF detection (CE-LIF). In thisway we have prepared and characterized BODIPY FLEDA conjugates of the four normal deoxyribonucleoside-3’-monophosphates (dNps) [5] and used as referencecompounds for the analysis of DNA by CE-LIF. Addi-tionally, 2’-deoxy-5-methylcytidine-3’-monophosphate(5mdCp) was synthesized, derivatized with BODIPY FLEDA [6], and used as standard for the determination of themethylation level in human tumor cells [7].

Here, we have extended the concept of fluorescencepostlabeling of normal and modified deoxynucleotideswith BODIPY FL EDA to 2’-deoxyribonucleoside-5’-monophosphates (pdNs). BODIPY FL EDA conjugates ofthe four normal pdNs were prepared and analyzed by CE-LIF to serve as standard compounds. Using the sameconditions for derivatization and slightly modified condi-tions for CE-LIF analysis as described previously [5] for the

analysis of DNA via BODIPY conjugates of dNps we showthat DNA and an oligodeoxynucleotide can be analyzedwith equal specificity and sensitivity when digested topdNs. As shown in Scheme 1, fluorescence postlabeling ofnormal and modified deoxynucleotides with BODIPY FLEDA can be performed either by digestion with nucleaseP1 (NP 1) or with a mixture of micrococcal nuclease (MN)and spleen phosphodiesterase (SPD). Having the oppor-tunity to choose between different enzymatic digestionprocedures widens the scope for assaying DNA damageby this method, especially for those modified nucleotideswhich show differential resistance to enzymatic excision.

pdNs obtained after enzymatic digestion of DNA havebeen labeled with a fluorescent tag, separated, anddetected using MEKC-LIF before by the group of Sharma[8, 9]. These authors reported the preparation of dansy-lated pdNs, which allowed the determination of modifiednucleotides in the presence of normal nucleotides at theamol level (10218 mol) by CE-LIF demonstrating that CE-LIF is a useful analytical tool for assaying DNA damage.However, enrichment by HPLC prior to fluorescencelabeling (dansylation) was required to enhance the assaysensitivity to detect 1 adduct/107 nucleotides per micro-gram of DNA [8].

Scheme 1. Scheme of the developed single-step DNA analysis by CE-LIF: after enzymatic digestionof DNA to 2’-deoxyribonucleoside-3’-monophosphates (dNps) or pdNs, chemical derivatization withBODIPY FL EDA and EDC leads to fluorescent phosphoramidate conjugates (dNp-Bodipy and pdN-Bodipy), which are analyzed by CE-LIF.


Electrophoresis 2005, 26, 2591–2598 Analysis of DNA by CE-LIF 2593

2 Materials and methods

2.1 Instrumentation

CE analysis was performed on a PACE MDQ system withan LIF detector (argon-ion laser with lem = 488 nm) fromBeckman Coulter (Munich, Germany). The separation wascarried out at 257C on an untreated fused-silica capillary(59 cm, effective length 48.5 cm, 50 mm ID) from CS-Chro-matography Service (Langerwehe, Germany); the appliedvoltage was 20 kV. The samples were injected hydro-dynamically with 2.5 psi?s. Outlet was the cathode for allruns. The capillaries were conditioned by rinsing with 1 M

sodium hydroxide (15 min), 1 M HCl (15 min), 1 M sodiumhydroxide (15 min), bidistilled water (5 min), and electrolyte(15 min). Between each run, the capillary was rinsed for60 s with 200 mM SDS, for 90 s with 1 M sodium hydroxide,for 60 s with bidistilled water, and for 120 s with electrolyte.Separations were achieved with the following electrolytes:(a) Electrolyte I for pdNs: 75 mM SDS in a solution of 80% v/v sodium phosphate buffer (20 mM, pH 9.0) and 20% v/vmethanol as organic modifier. For simultaneous separa-tion of 2’-deoxyinosine-5’-monophosphate (pdI) and2’-deoxyuridine-5’-monophosphate (pdU), glucose wasadded as additional modifier. (b) Electrolyte II for dNps:90 mM SDS in a solution of 90% v/v sodium phosphatebuffer (20 mM, pH 9.0) and 10% v/v methanol as organicmodifier. HPLC separations were performed on a JascoPU-980 intelligent pump, Jasco LG 980–02 ternary gra-dient unit, Jasco DG-980–50 3-line degaser (Groß-Umstadt, Germany), and a Rheodyne injector. MS (ESI)was performed on a Finnigan MAT TSQ 7000.

2.2 Reagents

1-Ethyl-3-(3’-N,N’-dimethyl-aminopropyl)-carbodiimidehydrochloride (EDC) was obtained from Fluka (Germany).BODIPY FL EDA was purchased from MoBiTec (Göttin-gen, Germany). MN, nuclease P1 (NP1), calf-thymusDNA (CT-DNA), MES, HEPES, and the dNps and pdNs(2’-deoxyadenosine-3’-monophosphate (dAp), 2’-deoxy-guanosine-3’-monophosphate (dGp), thymidine-3’-mono-phosphate (Tp), 2’-deoxycytidine-3’-monophosphate(dCp), 2’-deoxyadenosine-5’-monophosphate (pdA), 2’-deoxyguanosine-5’-monophosphate (pdG), thymidine-5’-monophosphate (pT), 2’-deoxycytidine-5’-mono-phosphate (pdC), pdU, and pdI) were obtained fromSigma (Steinheim, Germany). SPD was from Calbiochem(Darmstadt, Germany) and the oligonucleotide (21-mer)containing 5-methylcytosine was purchased from Euro-gentec (Liège, Belgium). All other chemicals were of thehighest purity available.

2.3 Synthesis and characterization of standardcompounds

One hundred microliters of 2’-deoxynucleoside-5’-phos-phate solution (1 mM in 100 mM HEPES, pH 6.5) was mixedwith 100 mL of a 300 mM EDC solution (in 100 mM HEPES,pH 6.5) and 100 mL of a BODIPY FL EDA solution (25 mM in100 mM HEPES, pH 6.5) and reacted overnight at 377Cunder constant shaking in the dark. The reaction mixtureswere separated on an ODS Hypersil RP 18 column (length250 mm, ID 20 mm, particle size 5 mm, Bischoff, Leonberg,Germany) with a lineargradient from 5 to 65% ACN in 0.1 M

triethylammonium acetate (TEAA), pH 7.0, in 30 min with aflow rate of 5 mL/min. The product fractions (2’-deoxy-nucleotide-5’-(bodipy)-phosphoramidates) were detect-ed with a Jasco FP 920 intelligent fluorescence detector(Groß-Umstadt) at lex = 500 nm, lem = 510 nm, collectedand lyophilized. 2’-Deoxyadenosine-5’-(bodipy)-phos-phoramidate: yield = 10%, MS (ESI negative) m/z = 646.0(anion) 100%, MS (ESI negative DAU 646.0) m/z 625.7[M - HF]2 (30%); m/z = 605.8 [M - 2HF]2 (100%). 2’-de-oxycytosine-5’-(bodipy)-phosphoramidate: yield: = 24%,MS (ESI negative) m/z = 622.1 (anion) 100%, MS (ESInegative DAU 622.2) m/z 602.0 [M - HF]2 (60%);m/z = 581.9 [M - 2HF]2 (100%). 2’-deoxyguanosine-5’-(bodipy)-phosphoramidate: yield = 17%, MS (ESI negative)m/z = 662.6 (anion) 100%, MS (ESI negative DAU 662.0)m/z 642.0 [M - HF]2 (40%); m/z = 622.0 [M - 2HF]2 (100%).2’-Deoxythymidine-5’-(bodipy)-phosphoramidate: yield =24%, MS (ESI negative) m/z = 637.1 (anion) 100%, MS(ESI negative DAU 637.1) m/z 616.9 [M - HF]2 (50%);m/z = 596.9 [M - 2HF]2 (100%).

2.4 DNA hydrolysis

(a) To pdNs: 10 mg of CT-DNA or oligonucleotide was dis-solved in6 mL ofbuffer (MES 50mM, pH 5.5) and 2 mL ofzincchloride (2 mM in water) and 2 mL NP 1 (0.266 mg = 0.2 U inwater) were added. The mixture was incubated for 2 h at377C. (b) To dNps: 10 mg of DNA was dissolved in water(5 mL) and hydrolyzed by incubation for 3 h at 377C with 5 mLof an enzyme mixture consisting of 4.2 mL of MN(150mU/mL) and SPD (12.5 mU/mL) and 0.8 mL buffer(0.25 M HEPES, 100 mM calcium chloride pH 6.0).

2.5 Derivatization

(a) For 10 mg DNA hydrolysate ordNps: 1.8 M EDC, (30 mL; in0.8 M HEPES buffer, pH 6.5), 27mM BODIPY FL EDA (30 mL;in 0.8 M HEPES buffer, pH 6.5), and 0.8 M HEPES buffer(20 mL) at pH 6.5 were added to the hydrolysate and incu-bated for 25h at 257C in the dark [6]. (b) For 10 mg DNA oroligonucleotide hydrolysate or pdNs: 1.8 M EDC (30 mL; in



0.35 M MES, pH 6.5), 27 mM BODIPY FL EDA (30 mL; in0.35 M MES, pH 6.5), and 20 mL MES, pH 6.5 were added tothe hydrolysate and incubated for 25h at 257C in the dark.

2.6 Bisulfite treatment of CT-DNA

DNA was denatured and treated with sodium bisulfite aspublished [10, 11]. Briefly, 2 mg of CT-DNA in a volume of50 mL H2O was denatured by sodium hydroxide (finalconcentration 0.2 mM) for 20 min at 377C. Thirty micro-liters of 10 mM hydroquinone (Sigma) and 520 mL of 3 M

sodium bisulfite (Sigma) at pH 5, both freshly preparedprior use, were added, mixed, and incubated undermineral oil layer at 557C for 14 h. Modified DNA was puri-fied using the Gene Clean Kit (Qbiogene) according to themanufacturer manual and eluted into 100 mL of water.Modification was completed by sodium hydroxide treat-ment for 20 min at 377C, followed by ethanol precipitation.Modified DNA was resuspended in water and stored at2207C (for no longer than 2 wk) until analysis.

3 Results and discussion

3.1 Synthesis and CE-LIF analyses of BODIPYconjugates of pdNs

BODIPY conjugates of the four normal pdNs were pre-pared by incubating each of the pdNs (pdA, pdG, pdC,and pT) with the water soluble carbodiimide EDC and

BODIPY FL EDA in HEPES buffer in the dark at 377C for16 h. Each reaction mixture was subjected to HPLCmonitored by fluorescence detection (excitation at500 nm, emission at 510 nm) and individual peaks werecollected and identified by MS. The yields of the obtainedBODIPY conjugates were determined by measurement at503 nm using the molar extinction coefficient of BODIPY[12] and were between 10 and 25%. These conjugateswere used as reference compounds in CE-LIF analyses.As shown in Fig. 1 baseline separation of a mixture ofthese standards was achieved within 25 min using slightlymodified electrolyte conditions as described before forthe 3’-monophosphate conjugates [13]. Detection andidentification of 2’-deoxy-5-methylcytidine-5’-monopho-sphate (p5mdC) were achieved by digesting an oligonu-cleotide site specifically modified with 5-methylcytosine(sequence: 5’-TGC C*AC GTC AGC CTG GAT ACC-3’)with NP 1, and then the digest was fluorescence labeledwith BODIPY FL EDA and injected into a CE column withCE-LIF. The resulting electropherogram (data not shown)showed a base-line separation from the four commonpdN conjugates with p5mdC eluting just before pdC. Thetime-corrected individual peak areas (mean of three runs)were determined and standardized to that of p5mdC. Theratios obtained were 5mdC:C:T:A:G = 1:7.7:2.7:4.1:1.5roughly representing the composition of the oligonucleo-tide (note that the 5’-T can not be labeled due to themissing phosphate group), except that the peak area forthe pdG conjugate was far too low. This is inline withresults obtained with the BODIPY conjugate of 2’-deoxy-

Figure 1. Electropherogram ofa mixture prepared by mixingstandard compounds of the fournormal pdNs according to theiroccurrence in CT-DNA (60% A/T, 40% G/C) with LIF detection.Peaks were assigned by addi-tion of characterized standardcompounds. Electrolyte I: 75 mM

SDS in 16 mM sodium phos-phate buffer pH 9.0 containing20% v/v methanol; fused-silicacapillary: total length, 59 cm;length to the detection window,48.5 cm; inner diameter, 50 mm;injection, 2.5 psi?s; tempera-ture, 257C; applied voltage,20 kV. The sample was diluted1000-fold with water prior toinjection.



Figure 2. CE-LIF analysis of amixture of fluorescence labeledpdNs after optimizing theseparation conditions by addi-tion of glucose to the buffer.Electrolyte: 75 mM SDS in 18 mM

sodium phosphate buffer pH 9.0containing 8% v/v methanol and0.5 M glucose; all other condi-tions see Fig. 1.

guanosine-3’-monophosphate (dGp) [5, 6, 13] arisingfrom a quenching effect, which is dependent on the base[14]. The pdNs of the bases uracil (pdU) and hypoxanthine(pdI), deamination products of cytosine and adenine,were subjected to the fluorescence postlabeling assayand reaction products detected by CE-LIF were tenta-tively identified as their BODIPY conjugates. By addingglucose to the separation buffer, a mixture consisting ofseven standard compounds was fully resolved by CE-LIF(Fig. 2), however, a longer separation time and a reversalof the elution order of p5mdC and pdC was observed. Asa consequence, the migration order was pdA , p-dU , pT , pdI , pdC , p5mdC , pdG. The addition ofglucose has been shown to improve the resolution andseparation selectivity in MEKC, thus allowing the separa-tion of nucleosides [15].

3.2 Comparative CE-LIF analysis of CT-DNAdigested to pdNs or to dNps

In order to compare the two digestion versions of our flu-orescence postlabeling assay CT-DNA (10 mg) wasdigested either with NP 1, which forms pdNs, or with amixture of MN and SPD in HEPES buffer, which formsdNps as described [5, 6, 7, 13]. NP 1 is widely used fordigesting DNA [16], however, the routinely used digestionmixture contains sodium acetate pH 5.5, which is subjectto interferences since carboxylic acid groups also will belabeled. Therefore, MES buffer pH 5.5 was used. As out-lined in Scheme 1, both digests (340 mM nucleotides) were

treated under identical conditions with an excess of car-bodiimide (340 mM EDC) and BODIPY FL EDA (9 mM) inthe dark at 257C for 25 h. After derivatization, equalamounts of both reaction mixtures were mixed and sub-jected to CE-LIF analysis under the conditions estab-lished for separating 3’-monophosphate conjugatesdescribed previously [7, 13]. The electropherogramobtained is shown in Fig. 3 and signals were character-ized by cochromatography with synthesized standards.Although the pdC and p5mdC containing conjugatesmigrated as a single peak (20.8 min), the others wereresolved. However, separation of pdC and p5mdC wasachieved by changing the SDS and methanol concentra-tion of the electrolyte (conditions as in Fig. 1). The profileresulting from the five dye-labeled 3’-monophosphateswas consistent with results published earlier by Wirtz etal. [13]. It is clear from Fig. 3 that 5’-monophosphates ofindividual nucleosides migrate faster than the corre-sponding 3’-monophosphates, with the exception of themonophosphates of deoxyguanosine (dGp faster thanpdG). The time-corrected total peak area of the 3’-mono-phosphate conjugates was comparable to that of the5’-monophosphate conjugates indicating that both ver-sions of our fluorescence postlabeling assay were equallyefficient. Therefore, we assume that the efficiencies of thedifferent DNA digestions and the derivatization reactionsare not significantly different. Since signal areas of indi-vidual corresponding nucleoside monophosphates werecomparable, the fluorescence quantum yields, whichwere determined for the 3’-monophosphate conjugates[5, 6], might be similar too.



Figure 3. CE-LIF analysis of amixture of pdNs and dNps.Equal amounts of digests eachobtained from 10 mg CT-DNAseparately digested with eitherMN/SPD or NP 1 and deriva-tized with BODIPY FL EDA weremixed and analyzed by CE-LIF.Electrolyte II, all other conditionssee Fig. 1.

3.3 Reproducibility of the analysis

In order to test the reproducibility of the assay, based onnucleoside-5’-monophosphate BODIPY conjugates, andto compare it with the established fluorescence post-labeling assay, based on nucleoside-3’-monophosphateBODIPY conjugates, we analyzed commercially availableCT-DNA. Aliquots from an identical source were hydro-lyzed by NP 1 and derivatized in four parallel, independ-ent reactions. Precision determination of the method wascarried out on one of these aliquots. The sample was an-alyzed in six consecutive runs by CE-LIF, peak areas ofthe 5 nucleoside-5’-monophosphate BODIPY conjugateswere time corrected and results were expressed as per-centage of the total peak area. The mean values obtainedwere: pdA 36.58% 6 0.25, pT 28.79% 6 0.11, pdC24.42% 6 0.08, pdG 8.43% 6 0.09, p5mdC 1.56%6 0.03 with an RSD below 3%. These results reflect thedifferent derivatization efficiencies for each nucleotideand the different fluorescence quantum yields of theBODIPY conjugates, confirming results reported for the2’-deoxynucleoside-3’-phosphate BODIPY conjugates[6, 13]. Assuming the fluorescence quantum yield andderivatization efficiency for 2’-deoxycytidine-5’-phos-phate and p5mdC are the same, the methylation level ofCT-DNA was found to be 6.02% 6 0.09, which is in goodaccordance with published data (6.47%) [13]. Measure-ments of the four aliquots each analyzed five to seventimes gave similar results: pdA 37.11% 6 0.62, pT

28.42% 6 0.56, pdC 24.35% 6 0.47, pdG 8.28% 6 0.35,p5mdC 1.55% 6 0.05 with an RSD below 5%. Thisrevealed a high degree of reproducibility between indi-vidual derivatizations as well as between individualmeasurements. Similarly, variations in the amount ofinjected nucleotide conjugates by diluting samples withwater (1:500, 1:1000, 1:5000, 1:10 000) before analysisalso had no influence on the determined nucleotide con-tent of CT-DNA, indicating that the relation between fluo-rescence and nucleotide content was linear over a broadrange of dilutions (data not shown). The LOD for pdNswas determined by diluting derivatized samples of CT-DNA (10 mg; 30.8 nmol nucleotides) to obtain concentra-tions of nucleotides in the picomolar range. The methodyielded a lower detection limit of 0.5 amol (,100 pM) at anS/N of 3 for each conjugate of pdA, pdC, and pT, and ap-proximately 1.0 amol (,200 pM) of pdG when a 5 nL(2.5 psi?s) volume of diluted samples was injected. Byinjecting a tenfold higher volume (,50 nL) we achieved aconcentration LOD for pdA of 10 pM, which is in goodagreement with the LOD for the 3’-monophosphate dApfound by Schmitz et al. [5] using essentially the samemethod. These authors have reported an LOD of 850 pM

for dAp, however, the sample had to be diluted 100-foldbefore injecting ,48 nL on a BioFocus 3000 CE. Thus, the48 nL volume contained 0.4 amol of dAp conjugate, cor-responding to 8.5 pM as the lowest amount or concentra-tion that has been injected and detected. These resultsshow that the two versions of this fluorescence post-



Figure 4. CE-LIF analysis of2 mg CT-DNA after bisulfitetreatment, digestion with NP 1,and subsequent derivatizationwith BODIPY FL EDA. Opti-mized electrolyte: 75 mM SDS in18 mM sodium phosphate bufferpH 9.0 containing 8% v/v meth-anol and 0.5 M glucose. Thesample was diluted 1000-foldwith water prior to injection.

labeling assay are equally sensitive. Using the same con-ditions as for 10 mg of CT-DNA for hydrolysis, derivatiza-tion, and analysis, 1 mg of CT-DNA was enough to obtainmeaningful electropherograms (data not shown). How-ever, a further reduction of the amount of DNA resulted inan increase of interfering signals. After reducing the con-centration of enzymes and fluorescence marker, Wirtz etal. [13] showed that the determination of the methylationlevel of 100 ng of CT-DNA was possible when analyzed asnucleoside-3’-monophosphates.

3.4 Analysis of bisulfite-treated DNA

Having established the reproducible experimental condi-tions, we sought to determine the uracil content in DNAtreated with bisulfite by CE-LIF. This reaction is commonlyused to determine the methylation status of every cyto-sine in genomic DNA [10, 11]. Such information is relevantto studies in developmental processes and in cancerwhere time-dependent changes in methylation patternsoccur [17–19]. In this method, sodium bisulfite is used toconvert cytosine residues to uracil residues in singlestranded DNA, under conditions whereby 5-methylcyto-sine remains unreactive. After treatment with bisulfite, CT-DNA (2 mg) was digested by NP 1, derivatized and ana-lyzed by CE-LIF. As expected and shown in the electro-pherogram in Fig. 4, the pdU BODIPY conjugate wasresolved and identified by spiking with standard. Peak

areas determined as percentage of the total peak areafrom two independent reactions analyzed in triplicatewere as follows: pdA 34.91% 6 0.08, pT 32.49% 6 0.15,pdU 17.88% 6 0.18, pdC 2.28% 6 0.09, pdG10.85% 6 0.31, p5mdC 1.59% 6 0.08. These resultswere slightly different from those obtained for 10 mg of CT-DNA (see above). The 5-methylcytosine content (1.59%)was inline with reported results and showed no change bybisulfite treatment [13 and present paper]. However, asmall amount of pdC (2.28%) was still detectable, indi-cating that the chemical conversion to uracil was around90%. In contrast, others [20] have shown that full cytosinedeamination was achieved under conditions similar tothose used by us. In addition the total peak area wasdetermined and found to be only about 30% of thatexpected for 2 mg of untreated CT-DNA, indicating a largeloss of DNA during the treatment. Using HPLC analysis ofpdNs Grunau et al. [20] reported that most of the DNA isdegraded during bisulfite treatment.

4 Concluding remarks

In this work we demonstrated that DNA can be analyzedfor genomic cytosine methylation levels and damage in asingle step, and with high sensitivity by fluorescencepostlabeling of 2’-deoxynucleoside-5’-monophosphateswith BODIPY FL EDA followed by use of CE-LIF. This ver-sion of the method is equally efficient and sensitive as that



reported before using dNps and therefore can be used forthe analysis of DNA especially in applications wheredigestion to deoxynucleoside-5’-monophosphates ismore advantageous than digestion to deoxynucleoside-3’-monophosphates (e.g., enzymatically resistant mod-ified nucleotides). Both versions of the assay have thepotential to monitor DNA modifications in human sam-ples.

We would like to thank Dr. B. Brueckner (German CancerResearch Center) for bisulfite treatment of CT-DNA. Thiswork was sponsored by the scholarship program of theGerman Federal Environmental Foundation (DBU) (toM.G.C.).

Received December 15, 2004

5 References

[1] Geldart, S. E., Brown, P. R., J. Chromatogr. A 1998, 828, 317–336.

[2] Kelman, D. J., Liliga, K. T., Sharma, M., Chem.-Biol. Interact.1988, 66, 85–100.

[3] Wang, P., Giese, R. W., J. Chromatogr. A 1998, 809, 211–218.

[4] Lan, Z. H., Qian, X., Giese, R. W., J. Chromatogr. A 1999, 831,325–330.

[5] Schmitz, O. J., Wörth, C. C. T., Stach, D., Wiessler, M.,Angew. Chem. Int. Ed. 2002, 41, 445–448.

[6] Stach, D., Schmitz, O. J., Stilgenbauer, S., Benner, A., Döhner,H., Wiessler, M., Lyko, F., Nucleic Acids Res. 2003, 31, 1–6.

[7] Lyko, F., Stach, D., Brenner, A., Stilgenbauer, S., Döhner, H.,Wirtz, M., Wiessler, M., Schmitz, O., Electrophoresis 2004,25, 1530–1535.

[8] Sharma, M., Jain, R., Ionescu, E., Slocum, H. K., Anal. Bio-chem. 1995, 228, 307–311.

[9] Lee, T., Yeung, E. S., Sharma, M., J. Chromatogr. 1991, 565,197–206.

[10] Herman, J. G., Graff, J. R., Myohanen, S., Nelkin, B. D.,Baylin, S. B., Proc. Natl. Acad. Sci. USA 1996, 93, 9821–9826.

[11] Frommer, M., McDonald, L. E., Millar, D. S., Collis, C. M.,Watt, F., Grigg, G., Molloy, P. L., Paul, C. L., Proc. Natl. Acad.Sci. USA 1992, 89, 1827–1831.

[12] Haugland, R. P., in: Spence, M. T. Z. (Ed.), Handbook of Flu-orescent Probes and Research Chemicals, 9th edn., Molec-ular Probes Inc., Eugene OR, 2002, 115–116.

[13] Wirtz, M., Stach, D., Kliem, H.-C., Wiessler, M., Schmitz, O.,Electrophoresis 2004, 25, 839–845.

[14] Seidel, C. A. M, Schulz, A., Sauer, M. H. M., J. Phys. Chem.1996, 100, 5541–5553.

[15] Kaneta, T., Tanaka, S., Taga, M., Yoshida, H., J. Chromatogr.1992, 609, 369–374.

[16] Li, G., Shimelis, O., Zhou, X., Giese, R. W., BioTechniques2003, 34, 908–909.

[17] Gaudet, F., Hodgson, J. G., Eden, A., Jackson-Grusby, L.,Dausman, J., Gray, J. W., Leonhardt, H., Jaenisch, R., Sci-ence 2003, 300, 489–492.

[18] Eden, A., Gaudet, F., Waghmare, A., Jaenisch, R., Science,2003, 300, 455.

[19] Lee, P. P., Fitzpatrick, D. R., Beard, C., Jessup, H. K., Lehar,S., Makar, K. W., Perez-Melgosa, M., Sweetser, M. T.,Schlissel, M. S., Nguyen, S., Cherry, S. R., Tsai, J. H., Tucker,S. M., Weaver, W. M., Kelso, A., Jaenisch, R., Wilson, C. B.,Immunity 2001, 15, 763–774.

[20] Grunau, C., Clark, S. J., Rosenthal, A., Nucleic Acids Res.2001, 29, E65.


Detection and separation of nucleoside-5'-monophosphates of DNA by conjugation with the fluorescent...

Documents

Transcript of Detection and separation of nucleoside-5'-monophosphates of DNA by conjugation with the fluorescent...