Electrochemical Study on the Effects of Epigenetic Cytosine

Dec 17, 2012 - Ian M. Huffnagle , Alyssa Joyner , Blake Rumble , Sherif Hysa , David Rudel , and Eli G. Hvastkovs. Analytical Chemistry 2014 86 (16), ...
0 downloads 0 Views 941KB Size
Download PDF
Article pubs.acs.org/ac

Electrochemical Study on the Effects of Epigenetic Cytosine Methylation on Anti-Benzo[a]pyrene Diol Epoxide Damage at TP53 Oligomers Jennifer E. Satterwhite, Caitlin M. Trumbo, Allison S. Danell, and Eli G. Hvastkovs* East Carolina University, Department of Chemistry, 300 Science and Technology Building, Greenville, North Carolina 27858, United States S Supporting Information *

ABSTRACT: Anti-benzo[a]pyrene-r-7,t-8-dihydrodiol-t-9,10epoxide (anti-BPDE) is a known carcinogen that damages DNA, and this damage is influenced by the DNA sequence and epigenetic factors. The influence of epigenetic cytosine methylation on the reaction with anti-BPDE at a known hotspot DNA damage site was studied electrochemically. Gold electrodes were modified with thiolated DNA oligomers spanning codons 270−276 of the TP53 gene. The oligomers exhibited 5-carbon cytosine methylation at the codon 273 location on the bound probe, the acquired complementary target, or both. Redox active diviologen compounds of the form C12H25V2+C6H12V2+C12H25 (V2+ = 4,4′-bipyridyl or viologen, C12-Viologen) were employed to detect antiBPDE damage to DNA. DNA was exposed to racemic (±)- or enantiomerically pure (+)-anti-BPDE solutions followed by electrochemical interrogation in the presence of C12-Viologen. Background subtracted square wave voltammograms (SWV) showed the appearance of two peaks at approximately −0.38 V and −0.55 V vs Ag/AgCl upon anti-BPDE exposure. The acquired voltammetry is consistent with singly reduced C12-Viologen dimers bound at two different DNA environments, which arise from BPDE damage and are influenced by cytosine methylation and BPDE stereochemical considerations. UV spectroscopic and mass spectrometric methods employed to validate the electrochemical responses showed that (+)-anti-BPDE primarily adopts a minor groove bound orientation within the oligomers while selectively targeting the nontranscribed ssDNA sequence within the duplexes.

G

utilizing DNA hybridization interfaces based on alterations in the DNA π-stack. We previously described a DNA damage detection approach using redox active diviologen molecules that bind in a structurally specific manner to DNA.23 C12H25V2+C6H12V2+C12H25 (V2+ = 4,4′-bipyridyl or viologen, termed C12-Viologen from here on) utilizes DNA as a sort of template, binding in different orientations based on DNA structure and providing electrochemical responses based on these environments.24−26 Genotoxicity from anti-r7,t8,t9,10epoxy-benzo[a]pyrene (anti-BPDE) at DNA oligomers was detected based on positive-shifted (vs formal potential) C12Viologen reduction peaks upon exposure to the xenobiotic. C12-Viologen voltammetry essentially elucidated DNA structural alterations induced by covalent BPDE adducts. Our anti-BPDE assay was effective in detecting DNA damage, and it elucidated a clear preference for BPDE to bind to a guanine within a “hotspot” codon site of the TP53

enotoxicity, or DNA damage, can occur through exposure to xenobiotic chemicals, or those originating outside the body.1,2 Because of the human health risk, genotoxicity must be monitored in chemical development industries. Several highthroughput assays have been developed to detect xenobiotic related genotoxicity.3−5 Emerging electrochemical genotoxicity detection approaches6−8 are attractive based on cost and throughput considerations.3 Several electrochemical genotoxicity detection approaches have been described including those that couple mass spectrometry analysis to probe adduct structure.3,9−12 One electrochemical approach to detect genotoxicity is to employ a DNA hybridization interface. Electrochemical hybridization sensors utilizing an electrode-bound probe strand hybridized to its target provide a convenient DNA analysis platform for detection or other biochemical analysis techniques.13−16 DNA can be detected directly or indirectly utilizing a bevy of strategies. The benefits of hybridization sensors lie in their ability to provide DNA structural information. For instance, mismatches,17,18 protein-binding,19−21 and DNA damage from cisplatin22 have been electrochemically detected © 2012 American Chemical Society

Received: October 25, 2012 Accepted: December 17, 2012 Published: December 17, 2012 1183

dx.doi.org/10.1021/ac303077h | Anal. Chem. 2013, 85, 1183−1191

Analytical Chemistry

Article

gene. Anti-BPDE metabolites have been shown to preferentially attack guanines within codons 157, 248, and 273 within this gene.27 TP53 codes for the p53 protein, which is involved in cellular apoptosis,28−31 and mutations at these particular hotspot codons are prevalent in many cancers.32−34 However, in vivo, genotoxicity is nuanced and complex. DNA sequence,35−37 epigenetic modifications,38−40 stereochemical considerations,41 and repair efficiency42,43 all influence ultimate potential genotoxin carcinogenicity. The aforementioned hotspots all feature 5′-CpG-3′ dinucleotide sequences, and within coding region exons 5−8 of the p53 gene, the cytosines within these dinucleotide sequences are often methylated at the 5-C location. Cytosine methylation is an epigenetic process involved in cell signaling,44−47 and its effects are often seen in the promoter region of genes, which may exhibit CpG islands.40 However, methylation in the coding region has also been shown to influence transcription in animal and plant cells.48 5C methylated cytosines located 5′ to guanine have been shown to influence the eventual structure of the BPDE-adduct,38,49,50 which in turn affects repair rates and eventual toxicity.51 Methods used to study the role of DNA sequence, stereoselectivity, and epigenetic factors from in situ-generated metabolite exposure have typically involved cellular or DNA exposure to a xenobiotic of interest, followed by purification, enzymatic digestion, specialized PCR amplification, gel separation, and sequencing.27,36,52−57 NMR and circular dichroism (CD) spectroscopy have been utilized to study the structure of specific DNA adducts.35,49,51,58,59 Here, we again utilize DNA hybridization sensors with oligomers from the hotspot TP53 codon 273 sequence, but we electrochemically elucidate the impacts of epigenetic cytosine methylation at the codon 273 site as well as probe the effects of anti-BPDE stereochemistry. We show that the obtained C12Viologen voltammetry is dependent on cytosine methylation and anti-BPDE stereoisomer exposed to the DNA oligomers. We employ both racemic (±)- and enantiomerically pure (+)-anti-BPDE (Scheme 1) to show that the SWV signal is due

Institute (Kansas City, MO). All DNA oligomers (see below) were purchased from IDT DNA Technologies (Coralville, IA). Benzo[a]pyrene (BP), Mercaptohexanol (MCH), Tris-HCl, Tris base, KH2PO4, K2HPO4, and THF were obtained from Sigma Aldrich. All other chemicals were reagent grade and used as received. Solution Preparation. DNA oligomer sequences used in this study were (a) meP.273, 5′-TTT GAG GTG meCGT GTT TGT GCC-3′ (me = 5-methyl cytosine, P = probe strand); (b) wt.273, 5′-TTT GAG GTG CGT GTT TGT GCC-3′ (wt = wild type, no methylation); (c) meC.273 complement, 5′-GGC ACA AAC AmeCG CAC CTC AAA-3′ (C = complementary strand); (d) wt.273 complement, 5′-GGC ACA AAC ACG CAC CTC AAA-3′. wt.273 and meP.273 oligomers were both purchased unmodified and with 5′-thiol modifications. DNA was prepared according to a previously published protocol.23 Details are summarized in the Supporting Information. BPDE is a known carcinogen and was handled using proper personal protective equipment. All BPDE chemicals were distributed into amber vials under a N2 atmosphere and kept sealed at −20 °C before use. Fresh stock solutions were prepared in the amber vials with THF for each day performing damage studies. Fresh BP stock solutions in THF were also prepared each day. Electrode Preparation. Detailed preparation protocol is summarized in the Supporting Information. Briefly, dsDNA was formed on cleaned gold 2-mm diameter electrodes by exposing to thiolated DNA, MCH, and unmodified complementary sequence ssDNA. Electrochemical Experiments. All electrochemical measurements were performed on a CH Instruments 660A workstation (Austin, TX). DNA modified electrodes were connected to the potentiostat and placed in 10 mL of 10 mM Tris, 10 mM NaCl, pH 7.4 (electrochemical buffer) along with Ag/AgCl reference (saturated KCl) and Pt counter electrodes. The electrochemical buffer was aggressively purged with Ar before all electrochemical runs. DNA surface coverage (Γ) was verified employing a previously established procedure using ruthenium hexamine (Ru(NH3)62+/3+) that binds to the electrode immobilized DNA.60 The electrode preparation procedure routinely provided dsDNA Γ ∼ 1.5−2.0 × 1012 dsDNA oligomers cm−2. For voltammetry with C12-Viologen, the following parameters were employed for the following techniques: cyclic voltammetry (scan from 0.1 to −0.7 V, 100 mV/s), square wave voltammetry (−0.2 V to −0.75 V scan, 4 mV step, 25 mV amplitude, 2 Hz frequency), chronocoulometry (0 V to −0.7 V pulse, 0.25 s pulse width, 0.01 sample interval). DNA Damage with BPDE. BPDE damage was performed following an established protocol.23 The detailed damage protocol is summarized in the Supporting Information. Briefly, BPDE was diluted from a THF stock solution in electrochemical buffer and exposed to DNA modified electrodes at 37 °C for a desired period of time. The electrode was rinsed with electrochemical buffer and placed back into the C12-Viologen electrochemical solution cell. Solution damage experiments were performed in ammonium acetate (50 mM, pH 6.8) with a 4× excess of BPDE at 4 °C in the dark. The reaction products were filtered and collected using Amicon Ultra Centrifugal filter units following the manufacturer instructions. nanoElectrospray Ionization Mass Spectrometry (nanoESI-MS) Analysis. DNA samples were diluted to 10 μM in

Scheme 1. anti-BPDE Stereoisomers

to the methylation influence on the binding of the two antiBPDE stereoisomers on the DNA. Overall, we present a bioanalytical technique that has the power to provide answers to basic DNA damage questions in addition to providing insight into epigenetic influences and stereochemical implications relating to DNA damage from an environmental carcinogen.

■

EXPERIMENTAL SECTION Materials. C12-Viologen was synthesized according to a previously published procedure.24 Racemic benzo[a]pyrene-r7,t-8-dihydrodiol-t-9,10-epoxide (±), (anti) (anti-BPDE, BPDE, item L0137) and the optically pure (+)-antibenzo[a]pyrene-trans-7,8-dihydrodiol-9,10-expoxide ((+)-anti-BPDE, item P0702) were obtained from the NCI Chemical Carcinogen Reference Standards Repository, Midwest Research 1184

dx.doi.org/10.1021/ac303077h | Anal. Chem. 2013, 85, 1183−1191

Analytical Chemistry

Article

exposed to meP-meC.273 up to 600 s. Background subtraction involved obtaining a baseline SWV before BPDE exposure (t = 0 s) and subtracting this SWV from subsequent SWVs obtained after BPDE exposure.23 Raw data are shown in the Supporting Information, Figure S1. For comparison purposes, the inset of Figure 1 shows the response obtained when (±)-anti-BPDE was exposed to the nonmethylated, wild type codon 273 sequence (wt.273, black) vs meP-meC.273 (red) for the same exposure time period. The main plot also features a control showing the negligible current response from exposure to 200 μM benzo[a]pyrene, suggesting that the current increases are due to the BP metabolite adducting the DNA, rather than a benzo[a]pyrene intercalative noncovalent interaction. This response is consistent with our previous report.23 Regarding alternate DNA sequence controls, we previously showed that obtained electrochemical responses were significantly muted if guanine at the codon 273 site was swapped for an alternate base; therefore, the significant time based current change seen in the Figure 1 data suggests a reaction between BPDE and the electrode immobilized DNA.23 Finally, it should be noted that this differential method of data representation shows current change as a function of BPDE exposure. We have previously shown that C12-Viologen does show minor nonspecific binding to the underlying MCH layer based on hydrophobic interactions;25 however, the plot here accounts for this binding by subtracting its contribution from the subsequent BPDE exposure runs. The introduction of cytosine methylation clearly influences the resulting C12-Viologen voltammetry compared to the nonmethylated case as seen in the inset plot of Figure 1. Using nonmethylated DNA, we previously showed that the primary current change seen upon (±)-anti-BPDE exposure was at −0.38 V with a minor current increase noted at −0.55 V (all potentials reported vs Ag/AgCl).23 The negative shifted potential was negligible in the nonmethylated case and was possibly influenced by the differential SWV subtraction method. In the presence of meP-meC.273, significant peak current increases at both −0.38 V and −0.55 V were seen, with the latter much more significant compared to wt.273. The Figure 1 inset shows that the −0.38 V peak current is roughly halved vs wt.273, while the −0.55 V current at methylated DNA-modified electrodes is approximately equal to the −0.38 V peak current. The (±)-anti-BPDE concentration and time responses at meP -meC.273 were studied and the results are plotted in Figure 2. Figure 2a shows that the peak current intensity at both −0.38 V and −0.55 V was dependent on (±)-anti-BPDE concentration. Figure 2b shows a plot of the −0.55 V peak current as a function of BPDE exposure time for the different BPDE concentrations. The plot shows that the peak current increased more rapidly when higher concentrations of BPDE were exposed to me.273. From the initial Figure 2b slopes, the apparent rate constant (k′) for the reaction between (±)-antiBPDE and DNA was calculated to be 2.8 × 10−3 s−1. This rate is approximately 5 times slower than what was determined for BPDE reacting at wt.273 (k′ = 0.015 s−1), but this reflects the more complex reaction process occurring at the electrode in the presence of meP-meC.273 involving competing reactions between the two stereoisomers and the methylated DNA. From the average background subtracted peak currents at the various 5 min (±)-anti-BPDE exposure times (n = 3), the linear range of the sensor provided an approximate limit of detection of 500 nM BPDE at 3:1 S/N. This value is lower than our

50:50 methanol−water. nanoESI-MS, using a custom-built source, was conducted on an Esquire 3000plus quadrupole ion trap mass spectrometer (Bruker Daltonics, Billerica, MA), which has unit mass resolution. Ions were created in negative mode nanoESI using ∼1 kV potential difference between the emitter and source aperture. Heated nitrogen (250 °C) was used to aid in nebulization and desolvation. dsDNA was ensured by performing thermal melting UV measurements before MS samples were diluted with the 50:50 methanol− water mixture. UV Spectroscopy. DNA solutions (5 μM) in 50 mM ammonium acetate (50 μL total) were added to thermostatted quartz cells (Starna Cells, Valencia CA) and placed in a Varian Cary 300 Bio UV−vis spectrophotometer with a temperature controller. For thermal melting, the samples were heated from 35 to 75 °C at a rate of 1.00 °C/min with a data collection interval of 0.5 °C. Data Analysis. OriginPro 8 graphing software was used for all data analysis.

■

RESULTS (±)-anti-BPDE Exposure. DNA oligomers spanning codons 270−276 were employed as this particular DNA sequence includes the known codon 273 5′-CGT-3′ BPDE damage hotspot on the nontranscribed coding strand in midsequence.27 The main focus of this contribution is the inclusion of methylated cytosines at the codon 273 site in both the gold electrode-immobilized probe (P) strand, which corresponded to the nontranscribed coding strand, and the complementary target (C) (termed meP-meC.273 from here on, where me = 5-methyl cytosine).61 The cytosines 5′ to guanines at the codon 273 site in both the P and C strands within this oligomer were methylated at the 5-C location. 5-C methylation produces a hydrophobic methyl group that protrudes into the major groove of the double helix.49 Initially, meP-meC.273 was exposed to racemic anti-BPDE solutions, consisting of both (+)-anti-BPDE and (−)-antiBPDE. Figure 1 shows background subtracted square wave voltammograms (SWV) obtained when (±)-anti-BPDE was

Figure 1. Background subtracted C12-Viologen SWV response at meP-meC.273 after 200 μM (±)-anti-BPDE exposure at 15 s (red), 30 s (green), 60 s (cyan), 180 s (blue), 300 s (pink), and 600 s (dark yellow). Black dash shows response after exposure to 200 μM benzo[a]pyrene for 600 s. Inset shows 200 μM (±)-anti-BPDE 300 s exposure comparison at wt.273 (black) and me.273 (red). 1185

dx.doi.org/10.1021/ac303077h | Anal. Chem. 2013, 85, 1183−1191

Analytical Chemistry

Article

coverage (Γsat) of 5.0 × 10−12 mol cm−2, which corresponded to an additional 2.1 (±0.6) C12-Viologen molecules bound per dsDNA duplex as a function of (±)-anti-BPDE exposure. C12Viologen binding isotherms for the two different redox potentials are shown in Figure S2 in the Supporting Information. The significance of these data is discussed further below. (+)-anti-BPDE Exposure. (+)-anti-BPDE and (−)-antiBPDE enantiomers have been shown to bind to nonmethylated DNA in similar, yet opposite, ways, aligning primarily in the minor groove postguanine adduction.51 Methylation at the 5′ cytosine is known to alter the morphology of the resulting antiBPDE-DNA adduct based on which enantiomer damages the guanine.49 To study the effects of methylation on BPDE adduction and subsequent C12-Viologen voltammetry, enantiomerically pure (+)-anti-BPDE was exposed to electrodeimmobilized methylated DNA. Additionally, the influence of methylation on either ssDNA forming the immobilized oligomer was studied. Figure 3 shows background subtracted SWV plots comparing the standard response when 100 μM (+)-anti-BPDE or 200 μM (±)-anti-BPDE was exposed to the three different methylated DNA oligomer combinations for 180 s. The inset of Figure 3 shows the response for meP-meC.273 (probe + complementary strands both methylated) exposed to (+)-anti-BPDE from 15 s to 300 s similar to Figure 1. It is clear that exposure to (+)-anti-BPDE produced higher current responses vs exposure to the racemic solution containing the same overall concentration of BPDE, which suggests a kinetically faster reaction with the pure enantiomer vs the racemate. In addition to higher overall peak current, the more interesting aspect of the Figure 3 data is the −0.38 V vs −0.55 V peak current differences upon exposure to (+)-antiBPDE. Figure 4 shows the average (n = 3) peak current (ip) ratios (−0.38 V/−0.55 V) at 300 s anti-BPDE exposure for each of the methylated DNA combinations. The ratio approaches ∼2:1 for all methylation combinations exposed to (+)-anti-BPDE vs ∼1:1 for (±)-anti-BPDE. Higher ip ratios indicate that a larger fraction of C12-Viologen was reduced at the more positive −0.38 V potential upon exposure to (+)-antiBPDE. Reduction at this potential is consistent with destabilization of the oxidized viologen (V2+) on the DNA based on the positive shift from its formal potential (Ef) of approximately −0.45 V. When DNA was exposed to the

Figure 2. (a) Background subtracted C12-Viologen SWV response on meP-meC.273 modified electrode at 300 s (±)-anti-BPDE exposure. (b) Peak current at −0.55 V over time for the (±)-anti-BPDE concentrations shown in part a. Error bars represent the standard deviation for n = 3 replications.

previous report23 but reflects the sluggish kinetic reaction between (±)-anti-BPDE and the methylated oligomer. The appearance of two peaks as BPDE was exposed to the DNA suggests that two main DNA-bound C12-Viologen redox active populations change upon BPDE exposure. A twoelectron reduction of the same viologen population from V2+ to V0 (V = viologen) is unlikely for two reasons. First, the V+1/0 reduction potential is negative shifted (approximately −0.8 V vs Ag/AgCl) from the V2+/1+ potential. Second, the neutral molecule is insoluble and precipitates onto the electrode, which is accompanied by a lack of voltammetric response at more negative potentials.24 Additionally, integration of CV reduction waves produced a total C12-Viologen saturated surface

Figure 3. (a) Background subtracted C12-Viologen SWV response on (a) meP-C.273 (i.e., no methylation on complement), (b) P-meC.273 (i.e., no methylation on probe), and (c) meP-meC.273 modified electrodes due to 180 s (±)-anti-BPDE exposure (200 μM, red) or (+)-anti-BPDE (100 μM, blue). Inset within part c shows the full SWV overlay from (+)-anti-BPDE exposure at meP-meC.273 from 15 to 300 s. 1186

dx.doi.org/10.1021/ac303077h | Anal. Chem. 2013, 85, 1183−1191

Analytical Chemistry

Article

few nanometers from the BP-Tetrol in solution (black spectrum in Figure 5). This slight shift toward longer wavelengths is indicative of BPDE-DNA adducts oriented in a solvent exposed environment.51,62,63 Similar results were obtained for all combinations of methylated oligomers exposed to (+)-anti-BPDE (Figure S2 in the Supporting Information). meP-meC.273 exposure to (±)-anti-BPDE resulted in λmax of 346 nm, an additional 3-nm red-shift from that seen upon exposure to (+)-anti-BPDE. The red-shift in indicates that adducts from (±)-anti-BPDE exposure are not exclusively located in a solvent-exposed environment. The importance of this data as it relates to the electrochemical results is discussed further below. Mass Spectrometry. Nano-ESI MS was employed in order to obtain a more complete picture of the reactions occurring on the electrode and to identify the primary reaction products involved in the BPDE-DNA reaction. Figure 6 shows the 7−

Figure 4. ip ratio (−0.38 V/−0.55 V) comparison for different methylated oligomer combinations (n = 3) exposed to racemic (200 μM) or enantiomerically pure anti-BPDE (100 μM) solutions for 300 s.

racemic solution, a larger fraction of C12-Viologen was reduced at −0.55 V resulting in lower ip ratios. This is consistent with stabilization of the oxidized viologen on the DNA based on the negative shift from Ef. Overall, these data suggest that cytosine methylation influences the anti-BPDE-DNA reaction. In turn, this affects the binding of C12-Viologen to the damaged complex. Methylation impacts are discussed in more detail below, but to gain additional insight into the BPDE-DNA reaction, additional analysis techniques were employed. UV−Visible Spectroscopy. UV−vis spectroscopic measurements were obtained in order to glean information related to the BPDE adduct orientation on the DNA helix. Figure 5 shows UV absorbance data between 310 and 380 nm for mePmeC.273 reacted with (±)-anti-BPDE or (+)-anti-BPDE as well as a BP-Tetrol spectrum obtained only in buffer. Absorbance in this region by BPDE corresponds to the S0 → S2 pyrenyl residue absorption band.62 The figure shows that DNA reacted with (+)-anti-BPDE exhibits absorbance maxima red-shifted a

Figure 6. nanoESI-MS of the 7− charge state for (+)-anti-BPDE reactions with meP-C.273 oligomers. Single BPDE-adducted ssDNA oligomers and unreacted oligomers are denoted with closed and open symbols, respectively. meP (▲,Δ) and C (◆,◇) ssDNA are shown.

charge state reaction products after (+)-anti-BPDE was exposed to double-stranded meP-C.273 oligomers. The spectrum shows four peaks representing both unreacted and singly adducted BPDE-ssDNA that formed the double stranded oligomer in solution. Experimental conditions favor dehybridization upon entry into the gas phase. On the basis of DNA thermal melting analysis data, the oligomers did exist as duplexes before MS analysis (data not shown). The spectrum shows that BPDE exhibited binding selectivity toward the probe, or coding, strand, as the BP+meP (black triangle) adduct peak exhibits approximately double the intensity compared to the BP+C (black diamond) adduct peak. BP denotes that the adduct is not the solution phase damage agent; it is no longer an epoxide. The Figure 6 spectrum is consistent with a single BP adduct per duplex, which dehybridizes upon gas phase entry, as well as attack at the codon 273 hotspot within the sequence.23Several other groups have also shown that DNA oligomers exposed to BPDE in excess up to 1:12 DNA to BPDE ratios also produce singly adducted oligomers at a specific hotspot guanine.64,65 Similar spectra were obtained for each of the methylated oligomer combinations, and mass spectral intensities were normalized based on the concentration ratios of unreacted ssDNA to BPDE-adducted ssDNA (more details found in the Supporting Information). The data from this analysis are shown in Table 1. The table shows that the relative MS intensities for the adducted BP-probe sequence are always significantly higher

Figure 5. UV−vis absorbance spectra for meP-meC.273 exposed to (±)-anti-BPDE (blue) or (+)-anti-BPDE (red). BP-Tetrol spectrum (black) was obtained in buffer only. 341 nm is marked with a dashed line. 1187

dx.doi.org/10.1021/ac303077h | Anal. Chem. 2013, 85, 1183−1191

Analytical Chemistry

Article

imparted to C12-Viologen by dsDNA, and geometric as well as spectroscopic evidence showed that the minor groove was a likely binding site for reduced C12-Viologen dimers. The emergence of the negative-shifted peak in SWV of BPDEexposed DNA suggests the emergence of a C12-Viologen population that is stabilized by the DNA structure. Peak growth at this potential was negligible in experiments using nonmethylated wt.273.23 The significant growth of −0.55 V peak here suggests that cytosine methylation within the codon 273 oligomer alters anti-BPDE binding in such a way to stabilize C12-Viologen binding. Methylated cytosines located 5′ to BPDE-adducted guanine alter the eventual orientation of the adduct depending on the BPDE stereoisomer. Either BPDE isomer initially intercalates at the adduct site before covalent attachment at the guanine N2 exocyclic amine creating a trans-anti-BP-Gua adduct. Added hydrophobic overlap imparted by the methylated cytosine stabilizes intercalation for (−)-trans-anti-BP-Gua, but this effect is minimized for the resulting (+)-trans-anti-BP-Gua. The lack of hydrophobic stabilization from the 5′ flanking methylated cytosine forces (+)-trans-anti-BP-Gua to reorient and eventually adopt a minor groove orientation. Intercalated (−)-transanti-BP-Gua forces base displacement of the planar guanine ring structure to face the minor groove.38,49,51,58 This base displacement and the accompanying minor groove alterations seemingly provide a stabilizing environment for oxidized C12Viologen. A possible stabilizing force may be π-system overlap between the viologen acceptor and purine donating effects from the displaced guanine. Similar stabilization has been reported for viologen coupled to an electron donating group within a dendrimer, which caused negative reduction potential shifts for the viologen.66 The result is a negative shifted reduction current seen upon (±)-anti-BPDE exposure to methylated cytosine DNA hybrids. Overall, on the basis of previous data using nonmethylated wt.273, the −0.38 V current decrease concomitant with −0.55 V current increase is consistent with the intercalation of a significant portion of the minor groove bound adducts seen in previous nonmethylated DNA studies. This is consistent with the reassignment of the (−)-trans-antiBP-Gua adducts to the base-displaced intercalated type. UV− vis data was consistent with intercalated adducts arising from exposure to (±)-anti-BPDE at meP-meC.273. The larger ∼5 nm absorbance red shift suggests a higher level of base stacking π−π* interactions, which would be expected of intercalating BP adducts arising from (−)-anti-BPDE damage at a guanine neighboring a methylated cytosine.50,63,67 Integration of CV reduction waves showed that approximately two additional C12-Viologen molecules per dsDNA duplex aggregate on the electrode surface as a function of (±)-anti-BPDE exposure over the time period shown in Figure 1. The binding isotherms generated from integration of each reduction peak within the CVs showed that both C12-Viologen populations saturated at a similar amount (Figure S1 in the Supporting Information). The speculative explanation for these saturation values is a near saturation of BPDE binding sites within the DNA layer featuring approximately half minor groove bound BPDE-damaged duplexes and half displaced guanine/intercalated BPDE-damaged duplexes. On the basis of the presumed π-interactions between displaced guanine and viologen and the ability of reduced C12-Viologen to dimerize,23−25 a saturation value of one C12-Violgen per dsDNA at −0.55 V might be expected if half of the damage sites featured an intercalating, base-displacing adduct. The ∼1.2

Table 1. Distribution of Reaction Products after (+)-antiBPDE Was Exposed to Denoted DNA Oligomersa relative MS peak intensity for BP-adducted ssDNAb oligomer

Probe ssDNA (coding)

Comp. ssDNA (noncoding)

meP-C.273 P-MeC.273 meP-meC.273

46.7 (±7.9) 71.8 (±1.0) 54.8 (±3.7)

24.9 (±10.0) 44.0 (±5.9) 31.9 (±11.3)

a

Each MS acquisition consisted of 20 spectra averaged, and acquisitions were carried out in triplicate. bReported as normalized average (±standard deviation).

than the corresponding complementary sequence. In general, intensities for BP-adducted probe sequence ssDNA are almost double those for the complementary sequence. Additionally, (+)-anti-BPDE reaction at nonmethylated sequences did show some strand preference for the probe strand (Figure S3 in the Supporting Information). The probe sequence corresponds to the coding, nontranscribed sequence in the TP53 gene, which contains the 5-CGT-3′ codon 273 sequence, and these data are consistent with (+)-anti-BPDE targeting this site within this particular ssDNA strand.64

■

DISCUSSION Multiple SWV reduction waves are indicative of multiple C12Viologen bound populations on the electrode-immobilized DNA. In our previous report, we postulated that the emergence of the positive shifted SWV −0.38 V peak in BPDE exposed dsDNA was due to the aggregation of C12-Viologen external to the DNA helix.23 Aggregation in this manner would be expected if a protruding hydrophobic site was created on the DNA helix upon reaction with anti-BPDE. Anti-BPDE is known to almost exclusively bind to guanine at the exocyclic N2 site at the 10-carbon BPDE position.51 Adduction of the racemic antiBPDE to guanine would be expected to primarily produce two different positional adducts aligned in opposite directions in the wt.273 minor groove depending on the stereoisomer. (−)-trans-anti-BP-Gua is known to align toward the 3′ side of the adducted DNA strand while (+)-trans-anti-BP-Gua aligns toward the 5′ side.51 The minor groove alignment of the BP adduct forces a slight protuberance of the hydrophobic pyrenyl system outside of the helix and exposure to the aqueous medium.51,58 This external hydrophobic location is a natural C12-Viologen aggregation site, as the dodecyl chains on C12Viologen endow it with extensive hydrophobicity. The current increase at −0.38 V is approximately 70 mV positive of the C12-Viologen formal redox potential (Ef) . This suggests that the viologen in this population is actually destabilized in the oxidized (V2+) form and that the driving force for reduction at that potential is the formation of face-to-face dimers via overlap of singly occupied π* orbitals. Peak growth at −0.38 V when methylated codon 273 oligomers were exposed to (±)-antiBPDE suggests that minor groove-located BPDE adducts were formed on the helix. The emergence of a peak approximately 100 mV negative shifted vs the C12-Viologen Ef suggests a more stabilized oxidized viologen form (V2+) on the DNA hybrids. In previous work studying C12-Viologen voltammetry differences in the presence of ssDNA vs dsDNA, it was shown that C12-Viologen binds to dsDNA in two distinct populations based on the DNA structure.24,25 Multiple-populations were not seen on ssDNA presumably because of the lack of a rigid helical structure. The negative shifted redox wave was explained by the stability 1188

dx.doi.org/10.1021/ac303077h | Anal. Chem. 2013, 85, 1183−1191

Analytical Chemistry

Article

BPDE binding selectivity toward the codon 157 and 158 guanines on the nontranscribed strand. Tretyakova et al. used isotope labeled-MS to demonstrate that BPDE reacted with higher selectivity at the nontranscribed codon 157 site if the complementary strand included methylated cytosine.65 In accordance with these reports, the MS data presented here is consistent with (+)-anti-BPDE selectively reacting at guanines in methylated cytosine codon sequences on the probe (nontranscribed coding) strand. Additionally, our data is consistent with previous reports demonstrating that methylation significantly influences the general coupling efficiency of anti-BPDE to DNA oligomers.50 The codon sequences where BPDE is likely bound within these oligomers are 5′-mCGT-3′ or 5′-mCGC-3′ for the meP or meC ssDNA, respectively. The sequence differences could account for the slight differences in electrochemical response when (+)-anti-BPDE was exposed to the partially methylated DNA oligomers. The more significant −0.55 V C12-Viologen reduction peak seen employing P-meC.273 could be due to more extensive adduct heterogeneity for anti-BPDE as it binds to this oligomer. It was shown that cytosine methylation in a variety of sequences forced eventual (+)-anti-BPDE-based adducts into both intercalative and minor groove bound orientations.50 DNA sequence contexts are vitally important in determining the eventual mutagenicity of the resulting adduct. The importance of these sequence effects are in the stability and the rigidity of the resulting adduct. BPDE adducts within CGC and CGT sequences have been shown to be more rigid and repaired less efficiently through nucleobase excision repair, while those in TGT sequences are more fluid and more easily repaired.35 Further studies will focus on the sequence effects to pinpoint the changes in voltammetric peak ratio between the two sequences, but it is clear based on the electrochemical and UV spectroscopic evidence that (+)-anti-BPDE exposure primarily results in minor groove bound (+)-trans-anti-BP-Gua adducts. The MS, spectroscopic, and electrochemical data all are consistent with methylation selectively driving (+)-anti-BPDE adduction onto the probe (nontranscribed coding) strand.

C12-Viologen per dsDNA ratio determined from the analysis here is consistent with slightly more than 50% intercalatingtype adducts within the film. Interaction between C12-Viologen and external hydrophobic BP-adduct sites would not necessarily be encumbered by minor groove-defined geometrical constraints but, based on the saturation level (∼0.9 C12-Viologen per dsDNA), the C12-Viologen presumably also interacts in the same manner at these sites. Larger −0.38 V peaks were detected upon exposure to pure enantiomer (+)-anti-BPDE (Figure 3). The reactivity of (+)-anti-BPDE has been shown to be much more active toward DNA compared to (−)-anti-BPDE or syn-BPDE.51,67 The accumulation of more BPDE adducts over this time period caused a larger accumulation of C12-Viologen onto the DNA and larger reduction peak currents. The orientation of (+)-trans-anti-BP-Gua has been shown to be relatively unaffected by 5′ flanking methyl cytosine, primarily forming minor groove bound adducts upon guanine attachment.49,58 A minor groove bound orientation for these BPDE adducts is consistent with the large observed electrochemical response at −0.38 V upon exposure to (+)-anti-BPDE. It is also possible that the activity of the enantiomerically pure BPDE could be kinetically enhanced due to the highly concentrated electrode surface reaction conditions favoring multiple BP adducts per dsDNA oligomer and leading to higher electrochemical signals. Although providing key insight, the solution phase conditions for MS cannot quite mimic these concentration conditions at the electrode surface. The larger −0.38 V signals do accompany smaller −0.55 V responses, but the larger −0.38 V peak does suggest that the majority of adducts from (+)-anti-BPDE exposure are located in the minor groove. UV spectroscopic data (Figure 5) supported this finding, as the slight 3 nm red shift from the BP-tetrol solution spectrum is indicative of solvent exposed BP within the DNA. The solvent exposed location has been shown to be the DNA minor groove where the BP pyrenyl edge protrudes slightly from the DNA helix.51,62,63 The −0.55 V peak did not entirely disappear when the methylated oligomers were exposed to (+)-anti-BPDE, however. This suggests that the BP adduct adopts multiple configurations on these methylated oligomers. When the complementary (noncoding) strand included the methylated cytosine, the −0.55 V peak was slightly larger compared to other methylation combinations, which is seen in the Figure 4 ratio plot. This ratio is indicative of the relative amounts of C12-Viologen bound presumably external to the DNA helix compared to within the minor groove, which is further indicative of minor groove vs intercalative BP adducts, respectively. It has been shown that sequence effects may influence the binding and eventual structural motif of (+)-transanti-BP-Gua.50 Our MS data showed that (+)-anti-BPDE showed binding selectivity toward the probe sequence, but the complementary sequence was damaged in appreciable amounts as well. Several groups have used various liquid chromatography−mass spectrometry (LC−MS) methods to show that cytosine methylation promotes BPDE binding to the 3′ guanine in DNA oligomers.50,64,65,68 MS methods designed to elucidate BPDE binding on DNA oligomers have typically monitored the nontranscribed strand, but previous reports have shown that methylation increases the selectivity of BPDE for a particular hotspot DNA sequence.64,65,68 Glick et al. used nanoLC−MS methods to show that methylation on either of the DNA strands in p53 oligomers spanning codons 155−159 increased

■

CONCLUSION

■

ASSOCIATED CONTENT

Overall, we have presented an electrochemical assay that can be used to study xenobiotic stereochemical and DNA sequence effects within a genotoxicity context. C12-Viologen is a convenient indirect redox reporter that produces voltammetry based on the structural nuances of DNA. Cytosine methylation is an important biological event that influences genotoxicity, and it was shown here that methylation may influence how a xenobiotic binds to the immobilized DNA producing a different voltammetric signature compared to a nonmethylated DNA sequence. Anti-BPDE was used in this study due to its wellstudied chemistry with DNA, but the results presented here suggest that the platform might be used to study myriad xenobiotics with known stereochemically important genotoxicity outcomes.

S Supporting Information *

Supplemental figures Figures S1−S4 as well as additional experimental protocols. This material is available free of charge via the Internet at http://pubs.acs.org. 1189

dx.doi.org/10.1021/ac303077h | Anal. Chem. 2013, 85, 1183−1191

Analytical Chemistry

■

Article

(26) Hvastkovs, E. G.; Buttry, D. A. Anal. Chem. 2007, 79, 6922− 6926. (27) Denissenko, M. F.; Pao, A.; Tang, M.; Pfeifer, G. P. Science 1996, 274, 430−432. (28) Vousden, K. H.; Lane, D. P. Nat. Rev. Mol. Cell Biol. 2007, 8, 275−283. (29) Vousden, K. H.; Lu, X. Nat. Rev. Cancer 2002, 2, 594−604. (30) Vogelstein, B.; Lane, D.; Levine, A. J. Nature 2000, 408, 307− 310. (31) Vogelstein, B.; Kinzler, K. W. Nat. Med. 2004, 10, 789−799. (32) Bullock, A. N.; Fersht, A. R. Nat. Rev. Cancer 2001, 1, 68−76. (33) Pfeifer, G. P.; Denissenko, M. F.; Olivier, M.; Tretyakova, N.; Hecht, S. S.; Hainaut, P. Oncogene 2002, 21, 7435−7451. (34) Pfeifer, G. P.; Besaratinia, A. Hum. Genet. 2009, 125, 493−506. (35) Cai, Y.; Patel, D. J.; Geacintov, N. E.; Broyde, S. J. Mol. Biol. 2007, 374, 292−305. (36) Feng, Z.; Hu, W.; Hu, Y.; Tang, M. Proc. Nat. Acad. Sci. U.S.A. 2006, 103, 15404−15409. (37) Arakawa, H.; Wu, F.; Costa, M.; Rom, W.; Tang, M. Carcinogenesis 2006, 27, 639−645. (38) Huang, X.; Colgate, K. C.; Kolbanovskiy, A.; Amin, S.; Geacintov, N. E. Chem. Res. Toxicol. 2002, 15, 438−444. (39) Denissenko, M. F.; Chen, J. X.; Tang, M.; Pfeifer, G. P. Proc. Nat. Acad. Sci. U.S.A. 1997, 94, 3893−3898. (40) Jones, P. A.; Baylin, S. B. Nat. Rev. Genet. 2002, 3, 415−428. (41) Cho, Y.; Kozekov, I. D.; Harris, T. M.; Rizzo, C. J.; Stone, M. P. Biochemistry 2007, 46, 2608−2621. (42) Delaney, J. C.; Essigmann, J. M. Chem. Biol. 1999, 6, 743−753. (43) Denissenko, M. F.; Pao, A.; Pfeifer, G. P.; Tang, M. Oncogene 1998, 16, 1241−1247. (44) Chen, Z. X.; Riggs, A. D. J. Biol. Chem. 2011, 286, 18347− 18353. (45) Ooi, S. K.; O’Donnell, A. H.; Bestor, T. H. J. Cell. Sci. 2009, 122, 2787−2791. (46) Zhao, Z.; Han, L. Biochem. Biophys. Res. Commun. 2009, 382, 643−645. (47) Kanai, Y.; Hirohashi, S. Carcinogenesis 2007, 28, 2434−2442. (48) Hohn, T.; Corsten, S.; Rieke, S.; Mueller, M.; Rothnie, H. Proc. Natl. Acad. Sci. U.S.A. 1996, 93, 8334−8339. (49) Zhang, N.; Lin, C.; Huang, X.; Kolbanovskiy, A.; Hingerty, B. E.; Amin, S.; Broyde, S.; Geacintov, N. E.; Patel, D. J. J. Mol. Biol. 2005, 346, 951−965. (50) Pradhan, P.; Graeslund, A.; Seidel, A.; Jernstroem, B. Chem. Res. Toxicol. 1999, 12, 816−821. (51) Geacintov, N. E.; Cosman, M.; Hingerty, B. E.; Amin, S.; Broyde, S.; Patel, D. J. Chem. Res. Toxicol. 1997, 10, 111−146. (52) Pfeifer, G. P.; Denissenko, M. F.; Tang, M. Toxicol. Lett. 1998, 102−103, 447−451. (53) Smith, L. E.; Denissenko, M. F.; Bennett, W. P.; Li, H.; Amin, S.; Tang, M.; Pfeifer, G. P. J. Natl. Cancer Inst. 2000, 92, 803−811. (54) Denissenko, M. F.; Tang, M.; Pfeifer, G. P. Biomark. Environ. Assoc. Dis. 2002, 139−157. (55) Pfeifer, G. P.; Riggs, A. D. Mol. Biotechnol. 1996, 5, 281−288. (56) Feng, Z.; Hu, W.; Chen, J. X.; Pao, A.; Li, H.; Rom, W.; Hung, M.; Tang, M. J. Natl. Cancer Inst. 2002, 94, 1527−1536. (57) Feng, Z.; Hu, W.; Rom, W. N.; Beland, F. A.; Tang, M. Carcinogenesis 2002, 23, 1721−1727. (58) Geacintov, N. E., Broyde, S., Eds. The Chemical Biology of DNA Damage; Wiley-VCH: Weinheim, Germany, 2010. (59) Broyde, S.; Wang, L.; Zhang, L.; Rechkoblit, O.; Geacintov, N. E.; Patel, D. J. Chem. Res. Toxicol. 2008, 21, 45−52. (60) Steel, A. B.; Herne, T. M.; Tarlov, M. J. Anal. Chem. 1998, 70, 4670−4677. (61) Hainaut, P. IARC TP53 Database. http://www-p53.iarc.fr/. (62) Geacintov, N. E.; Cosman, M.; Mao, B.; Alfano, A.; Ibanez, V.; Harvey, R. G. Carcinogenesis 1991, 12, 2099−2108. (63) Jernström, B.; Gräslund, A. Biophys. Chem. 1994, 49, 185−199. (64) Glick, J.; Xiong, W.; Lin, Y.; Noronha, A. M.; Wilds, C. J.; Vouros, P. J. Mass Spectrom. 2009, 44, 1241−1248.

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This work was financially supported by the East Carolina University Thomas Harriot College of Arts and Sciences (HCAS) and East Carolina University Office of Research and Graduate Studies. E.G.H. is especially thankful for the College Research Award from HCAS. We would like to thank Dr. Anne Spuches (Nanodrop) at East Carolina University for graciously allowing us to use her equipment.

■

REFERENCES

(1) Friedberg, E. C. Nature 2003, 421, 436−440. (2) Schaerer, O. D. Angew. Chem., Int. Ed. 2003, 42, 2946−2974. (3) Hvastkovs, E. G.; Schenkman, J. B.; Rusling, J. F. Ann. Rev. Anal. Chem. 2012, 5, 79−105. (4) Knight, A. W.; Little, S.; Houck, K.; Dix, D.; Judson, R.; Richard, A.; McCarroll, N.; Akerman, G.; Yang, C.; Birrell, L.; Walmsley, R. M. Reg. Toxicol. Pharm. 2009, 55, 188−199. (5) Lee, M.; Kumar, R. A.; Sukumaran, S. M.; Hogg, M. G.; Clark, D. S.; Dordick, J. S. Proc. Natl. Acad. Sci. U.S.A. 2008, 105, 59−63. (6) Palecek, E.; Fojta, M.; Tomschik, M.; Wang, J. Biosens. Bioelectron. 1998, 13, 621−628. (7) Diculescu, V. C.; Paquim, A. C.; Brett, A. M. O. Sensors 2005, 5, 377−393. (8) Lucarelli, F.; Palchetti, I.; Marrazza, G.; Mascini, M. Talanta 2002, 56, 949−957. (9) Hvastkovs, E. G.; So, M.; Krishnan, S.; Bajrami, B.; Tarun, M.; Jansson, I.; Schenkman, J. B.; Rusling, J. F. Anal. Chem. 2007, 79, 1897−1906. (10) Zhou, L.; Yang, J.; Estavillo, C.; Stuart, J. D.; Schenkman, J. B.; Rusling, J. F. J. Am. Chem. Soc. 2003, 125, 1431−1436. (11) Rusling, J. F.; Hvastkovs, E. G.; Schenkman, J. B. Screening for Reactive Metabolites Using Genotoxicity Arrays and Enzyme/DNA Biocolloids. In Drug Metabolism Handbook: Concepts and Application; Nassar, A. F., Ed.; Wiley: Hoboken, NJ, 2009; pp 307−340. (12) Krishnan, S.; Hvastkovs, E. G.; Bajrami, B.; Choudhary, D.; Schenkman, J. B.; Rusling, J. F. Anal. Chem. 2008, 80, 5279−5285. (13) Drummond, T. G.; Hill, M. G.; Barton, J. K. Nat. Biotechnol. 2003, 21, 1192−1199. (14) Odenthal, K. J.; Gooding, J. J. Analyst 2007, 132, 603−610. (15) Hvastkovs, E. G.; Buttry, D. A. Analyst 2010, 135, 1817−1829. (16) Del Giallo, M. L.; Lucarelli, F.; Cosulich, E.; Pistarino, E.; Santamaria, B.; Marrazza, G.; Mascini, M. Anal. Chem. 2005, 77, 6324−6330. (17) Kelley, S. O.; Barton, J. K.; Jackson, N. M.; Hill, M. G. Bioconjugate Chem. 1997, 8, 31−37. (18) Boon, E. M.; Jackson, N. M.; Wightman, M. D.; Kelley, S. O.; Hill, M. G.; Barton, J. K. J. Phys. Chem. B 2003, 107, 11805−11812. (19) Boal, A. K.; Barton, J. K. Bioconjugate Chem. 2005, 16, 312−321. (20) Boal, A. K.; Yavin, E.; Barton, J. K. J. Inorg. Biochem. 2007, 101, 1913−1921. (21) Boal, A. K.; Genereux, J. C.; Sontz, P. A.; Gralnick, J. A.; Newman, D. K.; Barton, J. K. Proc. Natl. Acad. Sci. U.S.A. 2009, 106, 15237−15242 (S15237/1-S15237/11). (22) Wong, E. L. S.; Gooding, J. J. J. Am. Chem. Soc. 2007, 129, 8950−8951. (23) Satterwhite, J. E.; Pugh, A. M.; Danell, A. S.; Hvastkovs, E. G. Anal. Chem. 2011, 83, 3327−3335. (24) Hvastkovs, E. G.; Buttry, D. A. Langmuir 2006, 22, 10821− 10829. (25) Hvastkovs, E. G.; Buttry, D. A. Langmuir 2009, 25, 3839−3844. 1190

dx.doi.org/10.1021/ac303077h | Anal. Chem. 2013, 85, 1183−1191

Analytical Chemistry

Article

(65) Tretyakova, N.; Matter, B.; Jones, R.; Shallop, A. Biochemistry 2002, 41, 9535−9544. (66) Balzani, V.; Bandmann, H.; Ceroni, P.; Giansante, C.; Hahn, U.; Klaerner, F.; Mueller, U.; Mueller, W. M.; Verhaelen, C.; Vicinelli, V.; Voegtle, F. J. Am. Chem. Soc. 2006, 128, 637−648. (67) Chen, F. M. Biochemistry 1987, 26, 4323−4331. (68) Chowdhury, G.; Guengerich, F. P. Angew. Chem., Int. Ed. 2008, 47, 381−384.

1191

dx.doi.org/10.1021/ac303077h | Anal. Chem. 2013, 85, 1183−1191