Electrospray Ionization Tandem Mass Spectrometry Analysis of the

ARTICLE pubs.acs.org/ac

Electrospray Ionization Tandem Mass Spectrometry Analysis of the Reactivity of Structurally Related Bromo-methyl-benzoquinones toward Oligonucleotides Janna Anichina,†,‡ Yuli Zhao,† Steve E. Hrudey,† Andre Schreiber,‡ and Xing-Fang Li*,† †

Division of Analytical and Environmental Toxicology, Department of Laboratory Medicine and Pathology, Faculty of Medicine and Dentistry, University of Alberta, 10-102 Clinical Sciences Building, Edmonton, Alberta, Canada T6G 2G3 ‡ AB SCIEX, 71 Four Valley Drive, Concord, Ontario, Canada L4K 4V8 ABSTRACT: We report the use of electrospray ionization tandem mass spectrometry (ESI-MS/MS) as a tool for rapid screening of structurally related chemicals toward oligonucleotides using the binding of five bromobenzoquinones with single-stranded (ss) and double-stranded (ds) oligonucleotides (ODNs) as a model. We found that these compounds interact differentially with oligonucleotides depending on the extent of their bromination and methylation. Three dibromobenzoquinones, 2,6-dibromo1,4-benzoquinone (2,6-DBBQ), 2,5-dibromo-1,4-benzoquinone (2,5-DBBQ), and 2,5dimethyl-3,6-dibromo-1,4-benzoquinone (DMDBBQ), bound to ssODN to form 1:1 adducts, and the binding constant of DMDBBQ bound to ssODN was 100-fold lower than those of 2,6-DBBQ and 2,5-DBBQ to ssODN, indicating that methyl groups hindered interactions of the bromoquinones with ODNs. Collision-induced dissociation (CID) of the 1:1 and 1:2 adducts of ODN with 2,6-DBBQ and 2,5-DBBQ demonstrated neutral loss of DBBQ and charge separations. Incubation of two tetrabromobenzoquinones (TBBQ), 2,3,5,6-tetrabromo-1,4-benzoquinone and 3,4,5,6-tetrabromo-1,2-benzoquinone, with the same ODNs did not form any adducts of TBBQ with ssODN or dsODN; however, bromide
ODNs were detected. Fragmentation of the bromide
ODN adducts showed loss of the HBr molecule, supporting the presence of bromide on ODNs. High-resolution MS and MS/MS analysis of the mixtures of dinucleotides (AA, GG, CC, and TT) and TBBQ confirmed the presence of bromide on the dinucleotides, supporting the transfer of bromide to ODNs through interaction with TBBQ. This study presents evidence of differential interactions of structurally related bromo and methyl-benzoquinones with oligonucleotides and demonstrates a potential application of ESI-MS/MS analysis of chemical interactions with ODN for rapid screening of the reactivity of other structurally related environmental contaminants toward DNA.

H

alobenzoquinones are a class of disinfection byproducts (DBPs) recently discovered in drinking water treated with common disinfection methods, such as chlorination or chloramination.1 These compounds have been predicted to be highly toxic and to have lowest observed adverse effect levels (LOAEL) up to 10 000 times lower than the regulated DBPs such as chloroform.2 Disinfection of drinking water has been proven to be the most efficient measure for preventing waterborne disease by inactivation of pathogenic microorganisms. However, unintended reactions between disinfectants and natural organic matter (NOM), anthropogenic contaminants, and bromide and iodide present in raw water lead to generation of DBPs in the treated water. Epidemiological studies have found an association between the consumption of chlorinated water and an increased risk of bladder cancer.1,3,4 The regulated DBPs, such as halomethanes and haloacetic acids, are easily detectable, but the carcinogenic potency of these DBPs is insufficient to explain epidemiological observations. Other DBPs, yet to be identified, may have higher toxicity. Recently, significant attention was given to the origin of the brominated and iodinated DBPs formed upon chloramination disinfection treatment.4,5 Overall, the toxicity of the established brominated and iodinated disinfection byproducts r 2011 American Chemical Society

significantly exceeds that of the chlorinated DBPs. We have also found that 2,6-dibromo-1,4-benzoquinone exhibited a much higher affinity toward oligonucleotides (ODNs) than the chlorobenzoquinones.6 The toxicological effects of haloquinone compounds are unclear due to limited information. A previous study examined the cytotoxicity of substituted benzoquinones, including methyl, chloro, and bromo substituents with rat and human hepatocytes.7 A correlation was found between the electron-withdrawing strength of the substituents on the benzoquinone ring and the cytotoxicity of the corresponding compounds. No study has examined the effects of halo and methyl groups on a quinone ring on their ability to interact with DNA. Benzoquinones are well-documented as causing DNA damage. The mechanisms of DNA damage by quinones have been suggested to involve one-electron reduction of the quinone ring to produce an unstable semiquinone radical that auto-oxidizes by O2, initiating the cascade of formation of a wide range of reactive oxygen species that may eventually cause Received: June 27, 2011 Accepted: September 9, 2011 Published: September 09, 2011 8145

dx.doi.org/10.1021/ac201646z | Anal. Chem. 2011, 83, 8145–8151

Analytical Chemistry Scheme 1

oxidative damage to DNA.8 In vitro studies indicated that nonhalogenated p-benzoquinones are capable of interacting with oligonucleotides to form benzetheno-type of adducts, such as 3-hydroxy-3,N(4)-benzetheno-20 -deoxycytidine or 7-hydroxy1N(2)benzetheno-2;-dexyguanosine9
11 under specific experimental conditions. Human and yeast Polη were found to actively mediate the process in vitro.12 Interactions of the halogenated benzoquinones with DNA have been much less investigated. Our previous study demonstrated that dibromobenzoquinone has higher affinity binding to ODNs than dichlorobenzoquinone. In the present study, we examined the effects of bromination and methylation of benzoquinones on the reactivity of these compounds toward single-stranded (ss) and double-stranded (ds) ODNs. Therefore, we selected five structurally related bromoand methyl-benzoquinones (Scheme 1) and used electrospray ionization tandem mass spectrometry (ESI-MS/MS) to study their interactions with oligonucleotides.

’ EXPERIMENTAL SECTION Chemicals and Oligonucleotides. HPLC-grade methanol and water were used to prepare the solvent mixture. The 2,6dibromo-1,4-benzoquinone (2,6-DBBQ) was purchased from Indofine Chemical Company (Hillsborough, NJ). The 2,5-dibromo1,4-benzoquinone (2,5-DBBQ), 2,3,5,6-tetrabromo-1,4-benzoquinone (2,3,5,6-TBBQ), 3,4,5,6-tetrabromo-1,2-benzoquinone (3,4,5,6-TBBQ), and 2,5-dibromo-3,6-dimethyl-1,4-benzoquinone (DMDBBQ) were purchased from Sigma-Aldrich. Two decameric ODNs were purchased from Integrated DNA Technologies (Coralville, IA, U.S.A.). The dinucleotides were purchased from ACGT Corporation (Toronto, ON, Canada) and were used without additional preparation. The sequences of the oligonucleotides were as follows: d(50 -GCGCATGCGC-30 ) (ODN1), d(50 GCGCGCGCGC-30 ) (ODN2) (nominal MW for ODN1 and ODN2 is 3030), d(50 -GG-30 ), d(50 -TT-30 ), d(50 -CC-30 ), d(50 -AA30 ). All oligonucleotides were desalted and cartridge-purified by the suppliers. Working solutions (20 μM) of ODNs were prepared by dilution of the stock solutions with a mixed solvent of 90% water with 10% methanol. ODN1 and ODN2 were annealed to form double-stranded species in 100 mM ammonium acetate buffer prior to ESI-MS analysis. Annealing was performed by heating the solutions of ODNs at 90 °C for 10 min and slowly cooling them down to room temperature to ensure the formation of their duplexes.

ARTICLE

Instrumental Conditions. Electrospray data were acquired in the negative ion mode using two AB SCIEX (Concord, ON, Canada) instruments: a 4000 QTRAP triple-quadrupole mass spectrometer (Q1q2Q3 with Q3 being a linear ion trap) and a Triple TOF 5600 hybrid quadrupole time-of-flight (TOF) mass spectrometer. Both instruments were equipped with a “turbo ion spray” ion source. Experiments with the 4000 QTRAP were performed at an ion spray voltage of
4200 V and a range of potential differences between the orifice and the skimmer. N2 was used as a curtain and nebulizing gas. Experiments with the Triple TOF 5600 were performed at an ion spray voltage of
4000 V and potential difference between the orifice and the skimmer of
80 V. Samples were infused directly into the electrospray ionization source at a flow rate of 3 μL min
1 with the 4000 QTRAP and 10 μL min
1 with the Triple TOF 5600. Tandem mass spectrometric measurements (MS/MS) with the 4000 QTRAP instrument were performed in the product ion (as well as enhanced product ion, EPI), precursor ion, neutral loss, and multiple reaction monitoring (MRM) modes using N2 as the collision gas. The collision offset voltage [the differential potential between the quadrupole entrance lens (q0) and the collision cell quadrupole (q2)], which nominally gives the collision voltage, was adjusted between
5 and
130 V at 1 V intervals. Product ion, precursor ion, and neutral loss spectra were obtained by scanning Q3 over the mass range of m/z 10
2800. In the product ion mode, the interquadrupole lens potentials and the float potential of the resolving quadrupole Q3 were linked to the q2 potential to maintain proper transmission through Q3. In the precursor ion mode, Q1 was scanned while Q3 was kept fixed. TOF MS/MS measurements were acquired at various fixed values of the collision offset voltage. AB SCIEX PeakView software was used for the analysis of the TOF MS/MS data. The FormulaFinder function of the software was used to independently assess the elemental composition of the peaks of interest. ESI-MS determination of the equilibrium association constants of ODNs 1 and 2 with 2,6-DBBQ, 2,5-DBBQ, and DMDBBQ was conducted using the 4000 QTRAP instrument. After separate incubation of the oligonucleotides with 2,6-DBBQ, 2,5-DBBQ, and DMDBBQ for 30 min, the incubated mixtures were analyzed using ESI at an ion spray voltage of
4200 V, the differential potential between the orifice and the skimmer of
30 V, N2 curtain gas at a setting of 20 psi, and N2 nebulizer at a flow rate of 8 L min
1. The solution concentrations of the ODNs were 20 μM for the binding with 2,6-DBBQ and 2,5-DBBQ and 20 and 50 μM for binding with DMDBBQ. The concentrations of 2,6-DBBQ and 2,5-DBBQ were 5- and 10-fold excess of ODNs, respectively, and of DMDBQ 10- and 20-fold excess of ODNs. The ionic strength of solutions was 0.1 M. The pH value of the solutions used in the direct infusion mode was 6.5.

’ RESULTS AND DISCUSSION We first examined the interactions of positional isomers of dibromobenzoquinones with both single- and double-stranded longer oligonucleotides. ODN1 and ODN2 are self-complementary ODNs capable of assembling into duplexes upon annealing. After annealing, both single-stranded (ss) and double-stranded (ds) ODN1 or ODN2 are present in the solution. The solution of ssODN1 and dsODN1 was incubated separately with 2,6-DBBQ and 2,5-DBBQ under the same conditions, and the incubation mixtures were subsequently analyzed using ESI-MS. ESI-MS analyses 8146

dx.doi.org/10.1021/ac201646z |Anal. Chem. 2011, 83, 8145–8151

Analytical Chemistry

ARTICLE

Table 1. ESI-MS Determination of the Equilibrium Association Constants of ODNs 1 and 2 with 2,6-DBBQ, 2,5-DBBQ, and DMDBBQa,b 2,6-DBBQ

2,5-DBBQ

DMDBBQ

(M )

(2.8 ( 0.5)

(2.2 ( 0.5)

(0.03 ( 0.01)

(ssODN1) K2  10
4 (M
1)

(1.3 ( 0.3)

(1.1 ( 0.6)

ndc


4

(ssODN1) K1  10


4

(ssODN2) K1  10


1


1

(M )

(1.9 ( 0.4)

(1.8 ( 0.5)

(0.03 ( 0.01)

(ssODN2) K2  10
4 (M
1)

(1.1 ( 0.3)

(1.0 ( 0.5)

nd

(dsODN1) K1  10
4 (M
1) (dsODN2) K1  10
4 (M
1)

(7.9 ( 0.5) (5.1 ( 0.5)

(8.1 ( 1.0) (7.0 ( 1.1)

nd nd

The solution concentrations of the ODNs were 20 μM for binding with 2,6-DBBQ and 2,5-DBBQ and 20 and 50 μM for binding with DMDBBQ. The concentrations of 2,6-DBBQ and 2,5-DBBQ were 5and 10-fold excess of the ODNs, respectively, whereas DMDBQ was 10and 20-fold excess of the ODNs. ss and ds designate single- and doublestranded species, respectively. The ionic strength of the solutions was approximately 0.1 M. b Raw data were smoothed and backgroundsubtracted. c nd: not detected. a

Figure 1. Negative ESI mass spectra of (A) ODN2 alone and (B) ODN2 and 2,6-DBBQ mixture (ratio 1:3) in a 10:90 methanol
water solution. The declustering potential was
5 V. [ODN] = 20 μM. / indicates 1:1 complexes of 2,6-DBBQ with the single- and double-stranded ODNs of different charges, as follows: ss/2,6-DBBQ6
(m/z 548.1); ss/2,6-DBBQ5
(m/z 658.1); ss/2,6-DBBQ4
(m/z 823.0); ds/2,6-DBBQ7
(m/z 902.4). ss = single-stranded ODN; ds = double-stranded ODN. The measurements were performed with the 4000 QTRAP instrument.

Figure 2. Negative ESI-MS spectra of (A) the mixture of ODN1 (20 μM) with 2,5-DBBQ (100 μM), showing m/z 604.5 ([ODN1
5H]5
), 658.4 ([(2,5-DBBQ)ODN1
5H]5
), 711.8 ([(2,5-DBBQ)2ODN1
5H]5
); (B) the mixture of ODN1 (50 μM) incubated with DMDBBQ (500 μM), showing m/z 604.5 ([ODN1
5H]5
), 663.8 ([(DMDBBQ)ODN1
5H]5
), 756.3 ([ODN1
4H]4
), and 830.2 [(DMDBBQ)ODN1
4H]4
. The measurements were performed with the 4000 QTRAP instrument at differential potential =
5 V.

of the incubation mixtures showed that 2,6-DBBQ and 2,5-DBBQ form 1:1 adducts with both single- and double-stranded ODN1 and ODN2 at molar ratios (ligand/ODN) of 3, 5, and 10. Figure 1 shows representative negative ESI mass spectra of ODN2 alone (top) and a mixture (ratio 1:3) of ODN2 and 2,6-DBBQ in a 90% water/10% methanol solution, showing the formation of the 1:1 adducts

(indicated with /). The 1:2 (ODN2/2,6-DBBQ) adducts were also detected at a molar ratio of 10. The peak at m/z 864.5 corresponds to the dsODN2 in the charge state of 7
. The peak with the asterisk at m/z 902.5 represents the 1:1 adduct of the dsODN with 2,6-DBBQ. Similar results were obtained when 2,5DBBQ was incubated with ODN1 and ODN2 in both single- and double-stranded forms, suggesting that the positional isomers have similar ODN binding affinity. To improve the signal-to-noise ratio and the quality of the spectra, we then monitored the reaction mixture for the narrower mass ranges that cover m/z of the native ODN and the complexes of bromobenzoquinones with the oligonucleotides. Various concentrations were examined, and representative spectra are shown in Figure 2, including ESI-MS spectra obtained from (Figure 2A) the mixture of ODN1 with 2,5-DBBQ and (Figure 2B) the mixture of ODN1 with DMDBBQ. Figure 2A clearly shows the 1:1 and 1:2 complexes formed from the incubation of ODN1 (20 μM) with 2,5-DBBQ (100 μM), whereas 1:1 adducts of ODN1 with DMDBBQ were produced from higher concentrations of 50 μM ODN1 with 500 μM DMDBBQ. Figure 2 also demonstrates the effect of the methyl group added to 2,6-DBBQ on ODN binding. Under the same conditions, incubating DMDBBQ with the ODNs at 1:5 (DMDBBQ/ODN) ratio and higher did not produce the 1:1 ODN/DMDBBQ complexes until the 1:10 ratio was reached. The 1:1 complexes were detected only for the ssODN species and not for the dsODN species (Figure 2). The estimated equilibrium binding constants of the 1:1 ODN/DMDBBQ complexes are summarized in Table 1. The enhanced ESI spectra of all complexes were obtained and used to estimate the equilibrium binding constants for all the complexes detected (Table 1). The equilibrium binding constants (defined in the equations below) for the 1:1 and 1:2 complexes of dibromobenzoquinones (L) with the studied ODNs were determined according to the method presented elsewhere.13 The concentrations of the ionic species at equilibrium were calculated based on the peak areas of the species of interest. The concentrations of free ODN and the complexed ODN were calculated from the total nucleic acid concentration and peak areas using the following equations: ½ODN ¼ ½ODNtotal AðODNÞ½ðAðODNÞ þ Að1 : 1Þ þ Að1 : 2Þ
1 8147

dx.doi.org/10.1021/ac201646z |Anal. Chem. 2011, 83, 8145–8151

Analytical Chemistry

ARTICLE

½1 : 1 ¼ ½ODNtotal Að1 : 1Þ½ðAðODNÞ þ Að1 : 1Þ þ Að1 : 2Þ
1 A designates the peak area. The total concentration of the bound L is then calculated: ½Lbound ¼ ½1 : 1 þ 2½1 : 2 ½Lfree ¼ ½Ltotal
½Lbound The concentrations of all the species at equilibrium therefore allow the calculation of the equilibrium binding constants. The stepwise binding constants are defined in the following equations: K 1 ¼ ½1 : 1½ODN
1 ½Lfree
1 K 2 ¼ ½2 : 1½1 : 1
1 ½Lfree
1 Table 1 summarizes the estimated binding constants of the 1:1 and 1:2 complexes of ODNs with 2,6-DBBQ, 2,5-DBBQ, and DMDBBQ. The estimated binding constants of the two position isomers, 2,6-DBBQ and 2,5-DBBQ, with the ODNs were similar, suggesting that the positions of bromine substituents on the benzoquinone did not affect their ability to bind to ODN. On the other hand, DMDBBQ only formed adducts with the singlestranded ODNs with lower binding constants compared to the DBBQs, and did not bind with the double-stranded ODNs, supporting the premise that methyl groups restrict DMDBBQ binding with ODN. The equilibrium binding constants determined for 2,6-DBBQ and 2,5-DBBQ in this study are consistent with that of 2,6-DBBQ binding to ODNs determined in the previous study using the same method.6 The structural similarities between 2,6-DBBQ and 2,5-DBBQ together with the close values of the stability constants may suggest some similarities between the binding of 2,5-DBBQ to ODNs and that of 2,6-DBBQ. The previous study demonstrated that 2,6-DBBQ bound noncovalently to singleand double-stranded nucleic acids without the requirement for a particular nucleobase sequence or secondary structure. It was suggested that 2,6-DBBQ binding to ODNs might involve partial intercalation of the benzoquinone moiety into the ODN combined with electrostatic interactions of bromine substituents with suitable sites on the ODN.6 Significantly lower values of the equilibrium binding constants of DMDBBQ with both ODNs may be rationalized on the basis of structural differences between 2,6-DBBQ, 2,5-DBBQ, and DMDBBQ. Introduction of the methyl groups increases hydrophobicity of the molecule and limits the interaction of the DMDBBQ molecules with negatively charged ODN molecules. The methyl group may also impose extra sterical hindrances to prevent intercalation of the quinone moiety into the ODN. To confirm the noncovalent nature of 2,6-DBBQ, 2,5-DBBQ, and DMDBBQ binding to ODN, we exposed 1:1 and 1:2 complexes of the three dibromobenzoquinones to collision-induced dissociation (CID). As shown in Figure 3A, the 1:2 complexes of ODN2 with 2,5DBBQ (m/z 711.2 at the charge state 5
) were primarily dissociated through a neutral loss of the 2,5-DBBQ moiety to produce the 1:1 complex (m/z 658.2 in the charge state of 5
). Dissociation of the 1:1 adduct (Figure 3B) was found to proceed via two competing channels: charge separation to produce a deprotonated [DBBQ
H]
(m/z 266.2) and ODN2 in the charge state of 4
(m/z 756.4) and neutral loss of the 2,5-DBBQ moiety to form the ODN in the

Figure 3. MS/MS spectra of the 1:2 and 1:1 complexes of ODN2 with 2,5-DBBQ (DBBQ0 representing 2,5-DBBQ) in the charge states of 5
, showing (A) m/z 711.2 and (B) m/z 658.1, respectively. Collision voltage was
10 V. The measurements were performed with the 4000 QTRAP instrument.

Figure 4. Negative ESI mass spectra of (A) ODN2 alone (sodium cluster of single-strand 3
charge state) and (B) ODN2 incubated with a 10-fold excess of 2,3,5,6-TBBQ in a 10:90 methanol
water solution. m/z 1036.5 corresponds to a singly brominated adduct of ODN2 in charge state of 3
. The measurements were performed with the 4000 QTRAP instrument.

charge state of 5
(m/z 604.5). Equations 1 and 2 summarize the processes observed in the MS/MS experiments of the two noncovalent complexes of dibromobenzoquinones and ODN1 and ODN2. ½ð2; 5-DBBQ Þ2 ODN
nHn
f 2; 5-DBBQ þ ½2; 5-DBBQODN
nHn


ð1Þ

These CID spectra show that the dissociation of the 1:1 adduct proceeds via the two pathways of charge separation and loss of a 8148

dx.doi.org/10.1021/ac201646z |Anal. Chem. 2011, 83, 8145–8151

Analytical Chemistry neutral 2,5-DBBQ, whereas the 1:2 adducts are dissociated through a neutral loss, followed by the charge separation and a neutral loss again. Our observations of the losses of neutral or deprotonated entities of 2,5-DBBQ suggest that ODN/2,5-DBBQ adducts are bound noncovalently. Similar results were obtained for 1:1 and 1:2 complexes of 2,6-DBBQ. Our observation regarding the binding mode of these two dibromobenzoquinones is consistent with the trends in CID of noncovalent complexes of groove binders and intercalators.13
15 The 1:1 adducts of DMDBBQ with single-stranded ODN1 and ODN2 also showed two CID pathways: a loss of a neutral DMDBBQ and a charge separation that results in the formation of a deprotonated ligand. However, the onset energies required for the dissociation of DMDBBQ/ODN adducts were found to

Figure 5. MS/MS spectrum of m/z 1036.5, [ODN281Br
2H]3
, averaged for the range of the laboratory collision voltage from
30 to
1 V. m/z 1030.5 corresponds to the loss of a water molecule from the parent ion, whereas m/z 1009.1 corresponds to the loss of H81Br molecule. The measurements were performed with the 4000 QTRAP instrument.

ARTICLE

be lower than those required for the adducts of 2,6-DBBQ and 2,5-DBBQ, indicating a lower gas-phase stability of ODN/ DMDBBQ adducts. We also investigated the reactions of two tetrabrominated benzoquinones, 2,3,5,6-TBBQ and 3,4,5,6-TBBQ, with ODN1 and ODN2 under the same conditions described above. The ratios of ODN/TBBQs were varied from 1:1 to 1:100. No adducts were detected, neither with the single- nor with the double-stranded ODNs, even at a 100-fold excess of tetrabromobenzoquinones. Instead, we observed peaks in the interval m/z 1035
1037 that nominally correspond to [d(50 -GCGCGCGCGC-30 )Br
2H]3
/[d(50 -GCGCATGCGC-30 )Br
2H]3
suggesting that 2,3,5,6-TBBQ and 3,4,5,6-TBBQ can mediate Br
transfer to the ODNs. Figure 4 provides an example of ESI-MS data at a low resolution. The selected resolution, however, did not allow unambiguous identification of the two isotopes of bromine at this charge state of the ODN1 and ODN2. To confirm the Br

ODN adducts detected, we performed CID of the corresponding peaks. Figure 5 presents an MS/MS spectrum of the peak m/z 1036.5 detected from the mixture of ODN2 and 2,3,5,6-TBBQ. The m/z 1036.5 nominally corresponds to [d(50 -GCGCGCGCGC-30 )81Br
2H]3
. The MS/MS spectrum of m/z 1036.5 clearly shows nominal losses of H2O and H81Br. We then performed enhanced resolution scans and neutral loss measurements to provide additional evidence of Br
transfer to the ODNs using the 4000 QTRAP instrument. In addition, we obtained accurate mass measurements, isotopic patterns of the bromide
ODN adducts, and confirmatory MS/MS

Figure 6. (A) Full negative ESI-TOF MS mass spectrum of adenine dideoxynucleotide (AA) in a 10:90 methanol
water solution in the presence of a 10-fold excess of 2,3,5,6-TBBQ. The concentration of d(50 -AA-30 ) was 10 μM. m/z 563.1525 corresponds to a deprotonated adenine dideoxynucleotide. (B) Enhanced area of m/z 642
648 (in black). Pink peaks outline the theoretical coverage of the corresponding isotopic contributions presented in the labeled m/z ratios. (C) MS/MS of monoisotopic [d(50 -AA-30 )79Br]
(nominal mass to charge of 643) at collision voltage of
50 V with collision voltage spread of 15 V. (D) Theoretical formula for the primary dissociation product of the monoisotopic m/z 643 with the elemental composition of C20H25N10O8PBr at the charge state of
1. The fragment ion of m/z 563.1514 9 was determined with 1.4 ppm mass accuracy, corresponding to the elemental composition of the deprotonated adenine dinucleotide [d(50 -AA-30 )
H]
. 8149

dx.doi.org/10.1021/ac201646z |Anal. Chem. 2011, 83, 8145–8151

Analytical Chemistry

ARTICLE

Scheme 2

measurements. For these sets of experiments, four homogeneous dinucleotides were used instead of the other ODNs for two reasons: (1) to show whether Br transfer is specific to a particular nucleotide and (2) to be able to detect adducts within the mass range of the high-resolution mass spectrometer. We separately incubated 2,3,5,6-TBBQ (100 μM) with four homogeneous dinucleotides: d(50 -AA-30 ), d(50 -GG-30 ), d(50 TT-30 ), and d(50 -CC-30 ) (20 μM each). ESI-MS measurements of all incubated mixtures using the low-resolution mass measurements with the 4000 QTRAP instrument showed the detection of bromide with the four dinucleotides, suggesting that bromide in TBBQ was transferred to AA, GG, TT, and CC after incubation. To further confirm bromide on the dinucleotides, we used high-resolution mass measurements using the Triple TOF 5600 mass spectrometer. Figure 6A shows a representative TOF MS spectrum of the mixture of a dinucleotide, d(50 -AA-30 ) (20 μM), with 2,3,5,6-TBBQ (100 μM). We detected a transfer of Br
to the ODNs in the presence of 2,3,5,6-TBBQ and 3,4,5,6-TBBQ with all four dinucleotides with less than 2% yield. To further confirm the proposed elemental composition of the species observed in the TOF MS and TOF MS/MS spectra, we used AB SCIEX PeakView software for the data analysis. Figure 6B illustrates the measured spectrum of bromide
AA matching with the theoretical spectrum, supporting the proposed elemental composition of brominated adenine dinucleotide. The theoretical elemental composition of the overall singly negatively charged [d(50 -AA-30 )Br]
is C20H25N10O8PBr. To confirm the assignments of the peaks, we acquired triple TOF MS/MS spectra of the brominated species and the separately deprotonated dinucleotides and used the Formula Finder option of the PeakView software to simultaneously confirm elemental composition of the daughter ions observed in the MS/MS spectra of both ions. Figure 6C shows a triple TOF MS/MS spectrum of [d(5 0 -AA-3 0 )79 Br]
at m/z 643.1 which is

completely fragmented to the major daughter ions (nominal m/z 563), corresponding to the elemental composition of [d(50 AA-30 )
H]
(Figure 6D) as the result of the neutral loss of H79Br molecule. Separate experiments confirmed the ion of nominal m/z 563 resulted from the dissociation of [d(50 -AA30 )79Br]
. The onset energies required for HBr dissociation from all complexes of bromide with the four dinucleotides were found to be similar, suggesting that Br
had similar affinity binding to the four dinucleotides. The MS/MS spectrum of [d(50 -AA30 )79Br]
at m/z 643.1 (Figure 6C) corresponds to the dissociation pathways described in Scheme 2. To support that TBBQ is associated with the bromide anion transferred to the ODNs and not inorganic bromide, we analyzed TBBQ alone using declustering potential (DP) varying from DP 5 to 20 V with ESI-Q1MS scan (m/z = 50
600) and did not detect Br
or loss of HBr, indicating that the TBBQ standard does not contain inorganic Br
ion (m/z = 79/81 with isotopic ratio of 100/100) as an impurity. In addition, we incubated the ODNs with ethidium bromide which is known to have Br
in its solution and did not observe any bromide adducts of ODNs. Incubation of dibromoquinones with ODNs did not produce any bromide ODNs either. All these results support the role of tetrabromobenzoquinone interaction with ODNs in the formation of bromide ODNs. Our study demonstrated differences in the interactions of bromobenzoquinones with ODNs depending on the extent of bromination of these compounds. Increasing the degree of bromination of benzoquinones (from dibrominated to tetrabrominated) altered the reactivity of these compounds toward oligonucleotides: from the formation of the noncovalently bound assemblies for the dibrominated p-benzoquinones to Br
transfer to the ODNs for the tetrabrominated p-benzoquinones. The modification on ODN by Br
transfer indicates potency of DNA damage. This is consistent with the previous observation that increasing the 8150

dx.doi.org/10.1021/ac201646z |Anal. Chem. 2011, 83, 8145–8151

Analytical Chemistry

ARTICLE

electron-withdrawing strength of the substituents on a benzoquinone ring increased cytotoxicity to the cells. Introduction of bromine groups (that are electron-withdrawing groups) increases the electrophilicity of the benzoquinone ring. One may also expect that four bromo substituents will make the benzoquinone ring more electrophilic than two substituents. Our results also show that methylation of bromobenzoquinones hinders their interaction with ODN. This study demonstrates an application of ESI-MS/MS analysis of chemical interactions with ODNs for rapid screening of reactivity of structurally related environmental contaminants toward DNA.

’ AUTHOR INFORMATION Corresponding Author

*E-mail: [email protected].

’ ACKNOWLEDGMENT The authors acknowledge that this study is partially supported by Grants from the Natural Sciences and Engineering Research Council of Canada, the Canadian Water Network, Alberta Health and Wellness, and the Alberta Water Research Institute, Alberta Innovates. ’ REFERENCES (1) Zhao, Y.; Qin, F.; Boyd, J. M.; Anichina, J.; Li, X.-F. Anal. Chem. 2010, 82, 4599–4605. (2) Bull, R. J.; Reckhow, D. K.; Rotello, V.; Bull, O. M.; Kim, J. Use of Toxicological and Chemical Models to Prioritize DBP Research. Denver, CO, 2006, http://waterrf.org/ProjectsReports/PublicReportLibrary/ 91135.pdf, accessed 9/26/2011. (3) IPCS. Disinfectants and Disinfection By-Products; Environmental Health Criteria 216; International Program on Chemical Safety; World Health Organization: Geneva, Switzerland, 2000. (4) Hrudey, S. E.; Hrudey, E. J. Safe Drinking Water; IWA Publishing: London, 2004. (5) Richardson, S.; Fasano, F.; Ellington, J. J.; Crumley, F. G.; Buettner, K. M.; Evans, J. J.; Blount, B. C.; Silva, L. K.; Waite, T. J.; Luther, G. W.; McKague, A. B.; Miltner, R. J.; Wagner, E. D.; Plewa, M. J. Environ. Sci. Technol. 2008, 42, 8330–8338. (6) Anichina, J.; Zhao, Y.; Hrudey, S. E.; Le, X. C.; Li, X.-F. Environ. Sci. Technol. 2010, 44, 9557–9563. (7) Chan, K.; Jensen, N.; O’Brien, P. J. J. Appl. Toxicol. 2008, 28, 608–620. (8) O’Brien, P. J. Chem.
Biol. Interact. 1991, 80, 1–41. (9) Linhart, I.; Mikes, P.; Frantik, E.; Mraz, J. Chem. Res. Toxicol. 2011, 24, 383–391. (10) Cemna, A.; Singer, B. Chem. Res. Toxicol. 1995, 8, 865–874. (11) Gaskell, M.; Jukes, R.; Jones, D. J. L.; Martim, E. A.; Farmer, P. B. Chem. Res. Toxicol. 2002, 15, 1088–1095. (12) Xie, Z.; Zhang, Y.; Guliaev, A. B.; Shen, H.; Hang, B.; Singer, B.; Wang, Z. DNA Repair 2005, 4, 1399–1409. (13) Rosu, F.; Gabelica, V.; Houssier, C.; De Pauw, E. Nucleic Acids Res. 2002, 30, e82. (14) Gabelica, V.; De Pauw, E.; Rosu, F. J. Mass Spectrom. 1999, 34, 1328–1337. (15) Anichina, J.; Bohme, D. K. J. Phys. Chem. B 2009, 113, 328–335.

8151

dx.doi.org/10.1021/ac201646z |Anal. Chem. 2011, 83, 8145–8151