A New Cross-Link for an Old Cross-Linking Drug: The NitrogenMustard Anticancer Agent Mechlorethamine Generates Cross-LinksDerived from Abasic Sites in Addition to the Expected Drug-BridgedCross-LinksMaryam Imani Nejad,† Kevin M. Johnson,† Nathan E. Price,† and Kent S. Gates*,†,‡

†Department of Chemistry, University of Missouri, 125 Chemistry Building, Columbia, Missouri 65211, United States‡Department of Biochemistry, University of Missouri, 125 Chemistry Building, Columbia, Missouri 65211, United States

*S Supporting Information

ABSTRACT: Nitrogen mustard anticancer drugs generatehighly reactive aziridinium ions that alkylate DNA. Mono-adducts arising from reaction with position N7 of guanineresidues are the major DNA adducts generated by theseagents. Interstrand cross-links in which the drug bridgesposition N7 of two guanine residues are formed in low yieldsrelative to those of the monoadducts but are generally thoughtto be central to medicinal activity. The N7-alkylguanineresidues generated by nitrogen mustards are depurinated toyield abasic (Ap) sites in duplex DNA. Here, we show that Ap sites generated by the nitrogen mustard mechlorethamine lead tointerstrand cross-links of a type not previously associated with this drug. Gel electrophoretic data were consistent with earlyevolution of the expected drug-bridged cross-links, followed by the appearance of Ap-derived cross-links. The evidence is furtherconsistent with a reaction pathway involving alkylation of a guanine residue in a 5′-GT sequence, followed by depurination togenerate the Ap site, and cross-link formation via reaction of the Ap aldehyde residue with the opposing adenine residue at thissite [Price, N. E., Johnson, K. M., Wang, J., Fekry, M. I., Wang, Y., and Gates, K. S. (2014) J. Am. Chem. Soc. 136, 3483−3490].The monofunctional DNA-alkylating agents 2-chloro-N,N-diethylethanamine 5, (2-chloroethyl)ethylsulfide 6, and naturalproduct leinamycin similarly were found to induce the formation of Ap-derived cross-links in duplex DNA. This work providesthe first characterization of Ap-derived cross-links at sequences in which a cytosine residue is located directly opposing the Apsite. Cross-linking processes of this type could be relevant in medicine and biology because Ap sites with directly opposingcytosine residues occur frequently in genomic DNA via spontaneous or enzymatic depurination of guanine and N7-alkylguanineresidues.

Nitrogen mustards such as mechlorethamine (HN2) werethe first synthetic anticancer drugs1,2 and remain in

widespread clinical use.3−6 These bifunctional agents generateaziridinium ions that react with DNA at a variety of locations,including N7-guanine, N3-adenine, N3-cytidine, and thephosphodiester linkages of the backbone (Scheme 1).4,7−21

Monoadducts (1 and 2) at guanine residues are the major DNAalkylation products formed by these drugs (Scheme 1).8,10,11

Interstrand cross-links (3) are generated in much lower yields(1−10% of total adducts) but are generally believed to be thecritical lesions responsible for medicinal activity of the nitrogenmustards.4,22,23 Cross-link formation by nitrogen mustards canoccur via reactions with two guanine residues in 5′-GNCsequences [3 (Scheme 1)];14,24−28 however, there is alsoevidence of G-G cross-link formation at 5′-GC sequences aswell as G-A and A-A cross-linking at as-yet-undefinedsequences.12,26,29

The alkylation of guanine and adenine residues by nitrogenmustards induces the formation of abasic (Ap) sites in genomicDNA (Scheme 1).9,10,30,31 We recently showed that Ap sites

can forge DNA−DNA interstrand cross-links in somesequences via reaction of the Ap aldehyde residue with theexocyclic amino groups of nucleobases such as adenine andguanine on the opposing strand of the DNA duplex (Scheme2).32−40 The cross-linking reactions considered here, involving“true” Ap sites, are distinct from those involving oxidized abasicsites.41,42 Here, we show that Ap sites generated by the nitrogenmustard mechlorethamine (HN2) give rise to interstrand cross-links of a type not previously associated with this drug. Wepresent gel electrophoretic data that are consistent with an earlyevolution of the expected drug-bridged cross-links 3 followedby the appearance of Ap-derived cross-links. The evidence isconsistent with a reaction pathway involving alkylation of aguanine residue in a 5′-GT sequence, followed by depurinationto generate the Ap site, and cross-link formation via reaction ofthe Ap aldehyde residue with the opposing adenine residue at

Received: October 21, 2016Revised: November 22, 2016Published: November 29, 2016

Article

pubs.acs.org/biochemistry

© 2016 American Chemical Society 7033 DOI: 10.1021/acs.biochem.6b01080Biochemistry 2016, 55, 7033−7041

this site (Scheme 2).35 We further showed that the monofunc-tional DNA-alkylating agents 2-chloro-N,N-diethylethanamine5, (2-chloroethyl)ethylsulfide 6, and natural product leinamycin(LNM) similarly induce Ap-derived cross-links in duplex DNAvia alkylation and depurination at 5′-GT sequences.

These results provide the first characterization of Ap-derivedcross-link formation at sequences in which a cytosine residuedirectly opposes the Ap site. The ability of such sites togenerate interstrand cross-links was by no means certain,because the nature of the directly opposing base can exertsignificant effects on the structures of Ap-containing duplexes.43

Cross-linking reactions of this type could be important inbiology and medicine because Ap sites with directly opposing

cytosine residues occur readily in genomic DNA viaspontaneous depurination of guanine and alkylguanine residues.The processes described here expand the list of mechanisms bywhich nitrogen mustards and other DNA-alkylating drugs cangenerate cytotoxic interstrand cross-links.

■ EXPERIMENTAL SECTIONMaterials and General Procedures. Reagents were

purchased from the following suppliers and were of the highestpurity available: oligonucleotides from Integrated DNATechnologies (Coralville, IA), uracil DNA glycosylase (UDG)and T4 DNA polynucleotide kinase (T4 PNK) from NewEngland Biolabs (Ipswich, MA), [γ-32P]ATP (6000 Ci/mmol)from PerkinElmer, 19:1 acrylamide/bis-acrylamide (40%solution/electrophoresis) from Fisher Scientific (Waltham,MA), and mechlorethamine hydrochloride and alkylatingagents from Sigma-Aldrich (St. Louis, MO). LNM was a giftfrom Kyowa Hakko Kogyo, Ltd. C-18 Sep-Pak cartridges werepurchased from Waters (Milford, MA), and BS Poly Prepcolumns were obtained from Bio-Rad (Hercules, CA).Quantification of radioactivity in polyacrylamide gels wasconducted using a Personal Molecular Imager (Bio-Rad) withQuantity One (version 4.6.5).

Representative Procedure for Cross-Link FormationTime Courses by HN2, 5, 6, and LNM. Single-stranded 2′-deoxyoligonucleotides were 5′-labeled using standard proce-dures.44 Labeled DNA was annealed44 with its complementarystrand to give the duplexes shown in Figure 1. In a typical cross-

linking reaction, HN2 was introduced into the reaction mixtureas a stock solution in DMF, to give a mixture containing HN2(1 mM) and labeled DNA in HEPES buffer (50 mM, pH 7)containing 100 mM NaCl and 10% (v/v) DMF that wasincubated at 37 °C for 96 h unless otherwise specified. TheDNA was ethanol-precipitated from the reaction mixture,44

resuspended in formamide loading buffer,44 and loaded onto a20% denaturing polyacrylamide gel, and the gel was electro-phoresed for 5 h at 1600 V. The amount of radiolabeled DNA

Scheme 1

Scheme 2

Figure 1. DNA sequences used in these studies. Ap-containingduplexes were generated by the action of UDG on the correspondingdU-containing duplex. Cross-link locations are indicated with a redconnection.

Biochemistry Article

DOI: 10.1021/acs.biochem.6b01080Biochemistry 2016, 55, 7033−7041

7034

in each band on the gel was measured by phosphorimageranalysis. The time course for the formation of the late-formingcross-link was determined by incubating a solution containinglabeled DNA (approximately 100000 cpm) and HEPES buffer(50 mM, pH 7) containing NaCl (100 mM) and HN2 (1 mM)at 37 °C. At specified time points, aliquots (3 μL) wereremoved and formamide loading dye was added followed byfreezing at −20 °C, and gel electrophoretic analysis as describedabove. For cross-link formation by LNM, all conditions wereidentical, except LNM was introduced as a stock solution inacetonitrile, the final concentration of LNM was either 100 or500 μM, and β-mercaptoethanol (0.5 or 2.5 mM) was added toinitiate the DNA alkylation reaction.45

Representative Procedure for Preparation of Du-plexes Containing Authentic Ap-Derived Cross-Links. Asingle-stranded, uracil-containing 2′-deoxyoligonucleotide was5′-labeled using standard procedures,44 annealed with itscomplementary strand, and treated with the enzyme UDG(50 units/mL, final concentration) to generate the Ap site. TheUDG enzyme was removed by phenol/chloroform extractionand the DNA ethanol precipitated. The Ap-containing duplexeswere incubated in a buffer composed of HEPES (50 mM, pH7) containing NaCl (100 mM) at 37 °C for 120 h unlessotherwise specified. The DNA was ethanol precipitated,resuspended in formamide loading buffer, and loaded onto a20% denaturing polyacrylamide gel that was electrophoresedfor 5 h at 1600 V. The amount of radiolabeled DNA in eachband on the gel was measured by phosphorimager analysis. Thetime course for the formation of the dA-Ap cross-link wasdetermined by incubating a solution containing labeled DNA(approximately 100000 cpm) and HEPES buffer (50 mM, pH7) containing NaCl (100 mM) at 37 °C. At specified timepoints, aliquots (3 μL) were removed, formamide loading dyewas added, and the sample was frozen at −20 °C and analyzedby gel electrophoresis as described above.Sequence Specific DNA Alkylation by HN2, 2-

(Chloroethyl)ethylsulfide (6), and LNM. Typical alkylationreaction mixtures contained HN2 (1 mM) and a 32P-labeledDNA duplex (Figure 1) in a buffer composed of HEPES (50mM, pH 7), NaCl (100 mM), and DMF [10% (v/v)] andincubated at 37 °C for 2 h. The DNA was ethanol precipitated,redissolved in aqueous piperidine (50 μL of a 1 M solution),and incubated at 90 °C for 25 min (Maxam−Gilbertworkup).46 The solution was frozen on dry ice, lyophilizedfor 40 min in a SpeedVac Concentrator at 37 °C, redissolved in20 μL of water, and evaporated again. The dried DNAfragments were dissolved in formamide loading buffer, loadedonto a 20% polyacrylamide denaturing gel, and electrophoresedat 1400 V for 5 h. The amount of radioactivity in the resolvedDNA fragments was quantitatively analyzed with a phosphor-imager. For alkylation by LNM, all conditions were identicalexcept that LNM was introduced as a stock solution ofacetonitrile (instead of DMF) and 50 μM LNM and 500 μM β-mercaptoethanol were employed. For alkylation by the sulfurmustard 6, a concentration of 500 μM was used.Hydroxyl Radical Footprinting of a Duplex C-

Containing Authentic Ap-Derived Cross-Link. We fol-lowed literature protocols for the footprinting of cross-linkduplex C.35,37,42,47 In this experiment, the strand opposing theAp-containing oligonucleotide was 5′-labeled using the stand-ard procedure.44 Labeled DNA was annealed with the uracil-containing complementary strand and the duplex treated withUDG to generate the abasic site as described above. The Ap-

containing double-stranded DNA (∼400000 cpm) wasincubated in HEPES buffer (50 mM, pH 7) containing NaCl(100 mM) at 37 °C for 120 h. The DNA was ethanolprecipitated, suspended in formamide loading buffer, andresolved on a 2 mm thick 20% denaturing polyacrylamide gel.The late-forming cross-linked duplex band was visualized usingX-ray film, the band cut out of the gel, and the gel slice crushed,and the gel pieces were vortex-mixed in elution buffer (200 mMNaCl and 1 mM EDTA) at room temperature for at least 1 h.The mixture was filtered through a Poly-Prep column toremove gel fragments, and the residue was ethanol precipitated,redissolved in water, and mixed with 2× oxidation buffer [10μL of a solution composed of 20 mM sodium phosphate (pH7.2), 20 mM NaCl, 2 mM sodium ascorbate, and 1 mM H2O2].To this mixture was added a solution of iron-EDTA [2 μL of 70mM EDTA and 70 mM (NH4)2Fe(SO4)2·6H2O] to start thereaction, and the mixture was vortexed briefly and incubated atroom temperature for 5 min before addition of a thiourea stopsolution (10 μL of a 100 mM solution in water). Hydroxylradical footprinting reactions, Maxam−Gilbert G reactions, andMaxam−Gilbert A+G reactions were performed on the labeledduplex to generate marker lanes.46 The resulting DNAfragments were analyzed using gel electrophoresis as describedabove.

■ RESULTS AND DISCUSSIONTreatment of Duplex DNA with HN2 Leads to

Generation of Distinct Early-Forming and Late-FormingCross-Links. It is well established that treatment of duplexDNA with HN2 can lead to the generation of interstrand cross-links that appear as slow-migrating bands on denaturingpolyacrylamide gels (located above the full-length single-stranded DNA in the denaturing gels presented here).14,26−28

Indeed, we found that treatment of duplex A (Figure 1) withHN2 (1 mM) in HEPES buffer (50 mM, pH 7) containingNaCl (100 mM) and DMF [10% (v/v)] at 37 °C, followed byelectrophoretic analysis of the 32P-labeled DNA fragments on adenaturing 20% polyacrylamide gel, revealed several distinctslowly migrating bands (Figure 2). The formation of multiplecross-linked species has previously been observed in duplexescontaining both 5′-GNC and 5′-GC sites treated withHN2.14,26,28 Most intriguingly, we found that the two majorcross-link bands generated by treatment of duplex A with HN2displayed very different formation time courses and gelmobilities (Figure 2). A major “early-forming” cross-link bandwas immediately evident. The yield of this band peaked at theearliest time point (2 h, 5% yield), and then the intensitydecreased over the remainder of the experiment (Figure 2B).This behavior was consistent with that expected for the typicaldrug-linked G-G cross-link in two key regards: (i) InterstrandDNA cross-link formation by HN2 is rapid,30,48−50 and (ii)depurination of HN2 adducts, occurring with a half-life of ∼9 hat 37 °C, was expected to cause spontaneous “unhooking” ofthe cross-link (with a corresponding disappearance of the cross-link band).10,30,31

We also observed a “late-forming” cross-link band thatbecame visible after 12 h and predominant after 48 h [4% yield(Figure 2)]. It seemed unlikely that this could be a drug-linkedcross-link because the chloroethyl groups of HN2 in both DNAmonoadducts [1 (Scheme 1)] and free HN2 hydrolyze rapidly(t1/2 = 1−30 min).29,48−51 On the other hand, in light of ourrecent studies,32−40 it was reasonable to consider that a buildupof Ap sites resulting from depurination of HN2 adducts (1−3)

Biochemistry Article

DOI: 10.1021/acs.biochem.6b01080Biochemistry 2016, 55, 7033−7041

7035

might permit the generation of Ap-derived cross-links. In thefollowing sections, we provide evidence that the late-formingcross-link band seen in Figure 2A is a dA-Ap cross-link resultingfrom alkylation-induced generation of an Ap site at position 2of duplex A (Figure 1).A DNA Duplex Containing an Authentic dA-Ap Cross-

Link at Position 2 Co-Migrates with the Late-FormingCross-Link Band Generated by Treatment of Duplex Awith HN2. The formation of Ap-derived cross-links has beenobserved in two different sequence motifs: 5′-CAp/5′-AG and5′-ApT/5′-AA (the cross-linked base is underlined; see Figure1 for images of the cross-link sequence motifs).35,37,38 In thedG-Ap cross-link, the cross-linking guanine residue is offset onebase to the 5′-side of the Ap site, while in the dA-Ap cross-link,the cross-linking adenine residue is offset one base to the 3′-side of the Ap site.35,37,38 In all of the Ap-derived cross-linksdescribed to date, an adenine residue has been located directlyopposing the Ap site and the dA-Ap cross-linking motifgenerates cross-link yields substantially greater than that of thedG-Ap motif.35,37,38

Recognition that alkylation-induced Ap sites have thepotential to generate interstrand DNA−DNA cross-links,combined with the knowledge that guanine residues are theprimary alkylation sites for HN2, drew our attention to 5′-GTsequences in duplex A as potential progenitors of HN2-induceddA-Ap cross-links. Specifically, alkylation at these sites, followedby depurination of the resulting alkylguanine residue, couldallow dA-Ap cross-link formation at the resulting 5′-ApT site.35

In duplex A, there are four such sites involving the guanineresidues at positions 1, 2, 6, and 8 (Figure 1). As an initial testof whether the late-forming cross-link band seen in Figure 2

could be an Ap-derived cross-link, we simultaneously exploredtwo questions: (i) Does generation of an Ap site at position 1,2, 6, or 8 in duplex A lead to dA-Ap cross-link formation, and(ii) if so, does the gel mobility of any of the resulting cross-linked duplexes match with that of the late-forming cross-linkband produced following treatment of duplex A with HN2?Toward this end, we conducted a series of experiments in

which an authentic Ap site was introduced at position 1, 2, 6, or8 in duplex A. The 5′-32P-labeled Ap-containing duplexes B−Ewere prepared as described in our previous work32−40 by theaction of uracil DNA glycosylase (UDG) on the corresponding2′-deoxyuridine-containing oligodeoxynucleotide duplexes.52,53

The formation of slowly migrating, Ap-derived cross-link bandswas easily detected in each of the Ap-containing duplexes B−E(Figure 3). The equilibrium yields of the cross-link generated in

these duplexes varied substantially, with values of 5, 30, 14, and2%, for duplexes B−E, respectively (Figure S1). Interestingly,cross-linked duplexes B−E displayed distinct gel mobilities(Figure 3). It has previously been observed that the location ofcross-links, with respect to the end of the DNA duplex, candramatically alter mobility in denaturing gels.54−56 In general,our observations follow the trends reported in two earlierstudies involving drug-derived cross-links, in which duplexescontaining cross-links near their ends were shown to migratefaster in denaturing gels than those containing cross-links neartheir center.54,55 More important in the context of the currentstudies is the fact that cross-linked duplexes B, D, and E derivedfrom Ap sites at positions 1, 6, and 8, respectively, did not co-migrate with the major late-forming cross-link band producedby treatment of duplex A with HN2. On the other hand, thecross-link derived from the Ap site at position 2 (duplex C) didco-migrate with the major late-forming band from the HN2reaction (Figure 3).We carefully characterized cross-link formation in the

authentic Ap-containing duplex C. We found that this cross-link was generated in an equilibrium yield of 30%, with anapparent formation half-time of 24 h in HEPES buffer (50 mM,

Figure 2. Treatment of duplex A with HN2 leads to the generation ofdistinct early-forming and late-forming cross-links. Duplex A wasincubated with HN2 (1 mM) in HEPES buffer (50 mM, pH 7)containing 100 mM NaCl and 10% (v/v) DMF at 37 °C. Aliquotswere removed at 0, 2, 4, 8, 24, 48, 72, and 96 h and stored frozen priorto gel electrophoretic analysis (lanes 1−8, respectively). The 32P-labeled 2′-deoxyoligonucleotides were resolved on a sequencing gel,and the radioactivity in each band was quantitatively measured byphosphorimager analysis. The bottom band corresponds to the full-length labeled 2′-deoxyoligonucleotides, and the slowly migratingupper bands correspond to cross-linked DNA. Figure 3. Late-forming cross-link generated by treatment of duplex A

with HN2 co-migrates with authentic dA-Ap cross-linked duplex C.The lower bands correspond to the 32P-labeled full-length labeled 2′-deoxyoligonucleotides and the upper bands cross-linked DNA. Lanes 1and 2 show duplex A incubated with HN2 (1 mM) in HEPES buffer(50 mM, pH 7) containing 100 mM NaCl and 10% (v/v) DMF at 37°C for 48 and 96 h, respectively, prior to sequencing gel analysis. Lanes3−6 show authentic dA-Ap cross-links in duplexes C, B, E, and D,respectively. The 32P-labeled 2′-deoxyoligonucleotides were resolvedon a 20% polyacrylamide denaturing gel, and the radioactivity in eachband was quantitatively measured by phosphorimager analysis.

Biochemistry Article

DOI: 10.1021/acs.biochem.6b01080Biochemistry 2016, 55, 7033−7041

7036

pH 7) containing NaCl (100 mM) at 37 °C (Figure S1). Iron-EDTA footprinting47 of the cross-linked DNA providedevidence that the Ap site was cross-linked to the adenineresidue at position 9 (Figure 1 and Figure S2).To further probe the involvement of the adenine residue at

position 9 in the generation of the late-forming dA-Ap cross-link induced by treatment of duplex A with HN2, weinvestigated the properties of duplex F, lacking this criticalresidue (Figure 1). We found that treatment of duplex F withHN2 generated the early-forming cross-link, but not the late-forming cross-link band (Figure S3). This result is consistentwith the assignments of the early-forming cross-link band as adrug-linked cross-link and the late-forming cross-link band as adA-Ap cross-link involving residue A9 of duplex A.Evidence That the Late-Forming Cross-Link Involves

Generation of Ap Sites in Duplex A by H2N. Resultsdescribed above provided evidence that a dA-Ap cross-link wasforged between an Ap site at position 2 and the adenine residueat position 9 in duplex A. HN2-induced formation of therequisite Ap site at position 2 must be preceded by alkylation ofthe corresponding guanine residue at this position. To explorethis issue, we treated the 32P-labeled duplex A with HN2 (1mM) in HEPES buffer (50 mM, pH 7) containing NaCl (100mM) for 2 h at 37 °C, followed by piperidine workup to inducestrand cleavage at the alkylated guanine residues. Electro-phoretic analysis of the 32P-labeled DNA fragments on adenaturing 20% polyacrylamide gel showed that guanineresidues at positions 2−4 and 6−8 were all alkylated by HN2[alkylation at position 1 could not be accurately measuredbecause the band was not resolved from full-length DNA, andposition 5 could not be measured because the band migratedoff the end of the gel (Figure 4 and Figure S4)]. Nonetheless,the results confirmed that position 2 was alkylated in substantialyield by HN2.To provide additional evidence that the late-forming cross-

link band generated by treatment of duplex A with HN2 wasderived from an Ap site in the duplex, we examined the effectsexerted upon the cross-linking reaction by two differenttreatments that modify Ap sites in duplex DNA. Specifically,we examined the effects of apurinic endonuclease (APE), anenzyme that cleaves on the 5′-side of Ap sites,57 andmethoxyamine (MX), a reagent that forms a stable oximederivative at Ap sites and inhibits the formation of Ap-derivedcross-links.35,58 In separate experiments, these reagents wereadded after 24 h to reaction mixtures containing HN2 andduplex A. These Ap-modifying reagents inhibited generation ofthe late-forming cross-link band, consistent with the notion thatHN2-induced generation of Ap sites was central to theproduction of the late-forming band (Figure 5). As expected,MX prevented both cross-linking and spontaneous strandcleavage by capping the Ap aldehyde residues as an inert oxime(lane 2, Figure 5).35 APE inhibited cross-link formation whilegenerating the anticipated strand cleavage at Ap sites resultingfrom spontaneous depurination of HN2−guanine adducts atG2−G4 (lane 3, Figure 5). The multiple bands generated byAPE may reflect both cleavage at Ap sites generated bydepurination of alkylated adenine residues and the result of theenzyme’s 3′-exonuclease activity on the initial cleavageproducts. Previous work by Osborne et al. showed that N3-alkyladenine residues are spontaneously released from DNAtreated with HN2, presumably with concomitant generation ofabasic sites.29

Generation of the Abasic-Derived Cross-Link byMonoalkylating Agents 2-Chloro-N,N-diethylethan-amine (5), 2-(Chloroethyl)ethylsulfide (6), and Leinamy-cin (7). In the reaction pathway proposed here, formation ofthe dA-Ap cross-link following treatment of duplex A with HN2is dependent upon the alkylating properties of HN2, but notdependent upon the cross-linking properties of the drug.Accordingly, the monofunctional nitrogen mustard 2-chloro-N,N-diethylethanamine 5 and other monofunctional DNA-alkylating agents that target guanine residues should becompetent to induce the late-forming cross-link band in duplexA. With this in mind, we examined the ability of two well-known21,59 monofunctional DNA-alkylating agents, 2-chloro-N,N-diethylethanamine (5) and (2-chloroethyl)ethylsulfide(6), to induce formation of the dA-Ap cross-link band induplex A. N7-Alkylguanine residues are major DNA adductsgenerated by both of these agents (Figure S4).21,59,60 We foundthat treatment of duplex A with either 5 (1 mM) or 6 (1 mM)in HEPES buffer (50 mM, pH 7) containing NaCl (100 mM)for 96 h at 37 °C did indeed produce the cross-linked duplex Cin yields comparable to that achieved by bifunctional alkylatorHN2 (Figure 6 and Figure S5).We also examined the ability of the Streptomyces-derived

anticancer natural product leinamycin (LNM) to generate thelate-forming dA-Ap cross-link in duplex A. LNM seems well-

Figure 4. Alkylation of guanine residues in duplex A by HN2. 32P-labeled oligonucleotide duplex A was incubated with HN2 in HEPESbuffer (50 mM, pH 7) containing 100 mM NaCl and 10% (v/v) DMFat 37 °C for 2 h followed by Maxam−Gilbert workup and separation ofthe labeled fragments on a 20% denaturing polyacrylamide gel.Labeled DNA was visualized by phosphorimager analysis. Lane 1shows an untreated duplex A. Lane 2 shows a Maxam−Gilbert Greaction on the labeled strand of duplex A. Lane 3 shows an A+Greaction on the labeled strand of duplex A. Lane 4 shows DNA treatedwith HN2 (1 mM) followed by Maxam−Gilbert workup. The 32P-labeled 2′-deoxyoligonucleotides were resolved on a sequencing geland visualized by phosphorimager analysis.

Biochemistry Article

DOI: 10.1021/acs.biochem.6b01080Biochemistry 2016, 55, 7033−7041

7037

suited for the generation of Ap-derived cross-links in DNA forthree reasons. (i) The reaction of LNM with thiols generates aDNA-binding episulfonium ion 8 that selectively and efficientlyalkylates position N7 of guanine residues in duplex DNA

(Scheme 3).45,61 (ii) LNM selectively alkylates 5′-GTsequences that support the formation of dA-Ap cross-

links.60,62 (iii) LNM−guanine adducts undergo unusuallyrapid depurination to yield abasic sites.63 We found thattreatment of duplex A with LNM (100 μM) and 2-mercaptoethanol (500 μM) generated a band that co-migratedwith cross-linked duplex C (Figure S6). Interestingly, treatmentof duplex A with higher concentrations of LNM generated acomplex mixture of slow-migrating bands that did notnecessarily co-migrate with authentic cross-linked duplexesB−E (Figure S7). The identity of these species remains to bedetermined.

■ CONCLUSIONSInterstrand DNA cross-links are the critical cytotoxic lesionsgenerated by a variety of anticancer drugs, including nitrogenmustards, busulfan, mitomycin C, and cisplatin.64 Nitrogenmustards have the potential to generate a variety of cross-linkswherein the drug connects G-G, G-A, and A-A resi-dues.12,26,27,29 Our work provides evidence of the generationof a new type of interstrand cross-link not previously associatedwith nitrogen mustards, involving alkylation of guanine residuesat 5′-GT sites, depurination of the resulting alkylguanineresidue, and formation of a dA-Ap cross-link. In ourexperiments, Ap-derived cross-links were produced in yieldscomparable to those of the drug-linked cross-links. The resultsindicate that care must be exercised when assigning the identityof slow-migrating cross-link bands in gel electrophoreticanalyses of mustard-treated DNA; not all slow-migratingbands are necessarily drug-linked cross-links. Following treat-ment of duplex DNA with HN2, the gel electrophoretic datawere consistent with early evolution of the expected drug-bridged cross-links, followed by the appearance of Ap-derivedcross-links. It seems clear that HN2-induced formation of Ap-derived cross-links does not occur equally well at all 5′-GT sitesin duplex A. The major site of cross-link formation involvesalkylation and Ap generation at G2 and cross-linking with A9(Figure 1). The observed yields of Ap-derived cross-link at any

Figure 5. Treatments that modify Ap sites inhibit generation of thelate-forming cross-link in duplex A. Evidence that the late-formingcross-link is derived from an Ap site in duplex A. Lane 1 shows duplexA incubated with HN2 (1 mM) in HEPES buffer (50 mM, pH 7)containing 100 mM NaCl and 10% (v/v) DMF at 37 °C for 96 h.Lane 2 shows duplex A incubated with HN2 (1 mM) in HEPES buffer(50 mM, pH 7) containing 100 mM NaCl at 37 °C, and after 24 h,CH3ONH2 (MX) was added to a final concentration of 2 mM and themixture incubated for an additional 72 h. Lane 3 shows duplex Aincubated with HN2 (1 mM) in HEPES buffer (50 mM, pH 7)containing 100 mM NaCl and 10% (v/v) DMF at 37 °C, and after 24h, APE was added and the mixtures were incubated for an additional72 h. The 32P-labeled 2′-deoxyoligonucleotides were resolved on asequencing gel and visualized by phosphorimager analysis.

Figure 6. Treatment of duplex A with the monofunctional nitrogenmustard 5 generates the late-forming dA-Ap cross-link. Duplex A wasincubated with compound 5 (1 mM) in HEPES buffer (50 mM, pH 7)containing 100 mM NaCl and 10% (v/v) DMF at 37 °C, and at 0, 2, 4,8, 12, 24, 48, 72, and 96 h, aliquots were removed from the reactionmixture and frozen at −20 °C prior to gel electrophoretic analysis(lanes 1−9, respectively). Lane 10 shows the authentic dA-Apcontaining duplex C. Labeled DNA in the gel was quantitativelydetected by phosphorimaging analysis. The bottom band in the gelimage is the full-length labeled 2′-deoxyoligonucleotide, and the slow-moving upper band corresponds to cross-linked DNA.

Scheme 3

Biochemistry Article

DOI: 10.1021/acs.biochem.6b01080Biochemistry 2016, 55, 7033−7041

7038

given 5′-GT site must reflect an interplay of HN2 alkylationefficiency, depurination of the alkylguanine residue, and theinherent potential for Ap-derived cross-link formation at thatsequence in the duplex.Generation of the dA-Ap cross-link by HN2 relies upon the

DNA alkylating properties of the drug and not its bifunctionalcross-linking properties. Accordingly, we found that formationof the dA-Ap cross-link was similarly induced by three differentmonofunctional alkylating agents that target guanine residues induplex DNA, 2-chloro-N,N-diethylethanamine (5) and (2-chloroethyl)ethylsulfide (6), and LNM.Previously characterized Ap-derived cross-links have been

located at sequences in which an adenine residue directlyopposes the Ap site.32−40 In biological systems, Ap sites withopposing adenine residues arise via the enzymatic removal ofmisincorporated 2′-deoxyuridine by the base excision repairenzyme UDG.65 This may be a major source of Ap sites ineukaryotic cells.65 The loss of damaged thymine residues alsocan give rise to Ap sites with a directly opposing adenineresidue.66 The work described here explored the properties ofAp sites arising from another important cellular process, thedepurination of guanine and alkylguanine residues. Theseevents give rise to Ap sites opposed by a cytosine residue. Atthe outset of this work, it was uncertain whether Ap sitesopposed by a cytosine residue could engage in interstrandcross-linking reactions. This uncertainty was due to the fact thatthe nature of the base directly opposing an Ap site can exertsignificant effects on the structure of an Ap-containingduplex.43,67,68 For example, the Ap residue can adopt anextrahelical conformation when opposed by a pyrimidineresidue.43,67,68 Nonetheless, the results reported here showfor the first time that Ap sites with directly opposing cytosineresidues are competent to form dA-Ap cross-links. Indeed,formation of slow-migrating cross-link bands was easilydetected in each of the Ap-containing duplexes B−E, withthe equilibrium cross-link yields ranging from 2 to 30%. Giventhat Ap sites with directly opposing cytosine residues occurfrequently in genomic DNA via spontaneous or enzymaticdepurination of guanine and N7-alkylguanine resi-dues,9,10,63,69−71 the results provide an additional impetus toexamine the occurrence and consequences of Ap-derived cross-links in cellular DNA.

■ ASSOCIATED CONTENT

*S Supporting InformationThe Supporting Information is available free of charge on theACS Publications website at DOI: 10.1021/acs.bio-chem.6b01080.

Time courses for the formation of cross-linked duplexesB−E; iron-EDTA footprinting of duplex C; generation ofa late-forming cross-link in duplex F; sequence specificityof guanine alkylation by HN2, 6, and LNM; generationof the late-forming cross-link following treatment ofduplex A with 6 or LNM (PDF)

■ AUTHOR INFORMATION

Corresponding Author*E-mail: gatesk@missouri.edu. Phone: (573) 882-6763. Fax:(573) 882-2754.

ORCIDKent S. Gates: 0000-0002-4218-7411

FundingWe thank the National Institute of Environmental HealthSciences of the National Institutes of Health for support of thiswork (ES 021007).NotesThe authors declare no competing financial interest.

■ ABBREVIATIONSAp, DNA abasic site; dA, 2′-deoxyadenosine; dG, 2′-deoxyguanosine; UDG, uracil DNA glycosylase; T4 PNK, T4DNA polynucleotide kinase; HEPES, 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid; EDTA, ethylenediaminetetra-acetic acid; Tris, tris(hydroxymethyl)aminomethane; HN2,mechlorethamine; LNM, leinamycin; MX, methoxyamine;APE, apurinic endonuclease; nt, nucleotide.

■ REFERENCES(1) De Vita, V. T. J., and Chu, E. (2008) A history of cancerchemotherapy. Cancer Res. 68, 8643−8653.(2) Goodman, L. S., Wintrobe, M. M., Dameshek, W., Goodman, M.J., Gilman, A., and McClennan, M. T. (1946) Nitrogen mustardtherapy: use of methyl-bis(beta-chloroethyl)amine hydrochloride andtris(beta-chloroethyl)amine hydrochloride for Hodgkin’s disease,lymphosarcoma, leukemia, and certain allied and miscellaneousdisorders. JAMA 132, 126−132.(3) Cheson, B. D., and Rummel, M. J. (2009) Bendamustine: rebirthof an old drug. J. Clin. Oncol. 27, 1492−1501.(4) Povirk, L. F., and Shuker, D. E. (1994) DNA damage andmutagenesis induced by nitrogen mustards. Mutat. Res., Rev. Genet.Toxicol. 318, 205−226.(5) Physicians’ desk reference, 65th ed. (2011) Thomsom PDR,Montvale, NJ.(6) WHO Model List of Essential Medicines, 19th List (2015) WorldHealth Organization, Geneva.(7) Cullis, P. M., Green, R. E., and Malone, M. E. (1995) Mechanismand reactivity of chlorambucil and chlorambucil-spermidine conjugate.J. Chem. Soc., Perkin Trans. 2 2, 1503−1511.(8) Mohamed, D., Mowaka, S., Thomale, J., and Linscheid, M. W.(2009) Chlorambucil-adducts in DNA analyzed at the oligonucleotidelevel using HPLC-ESI MS. Chem. Res. Toxicol. 22, 1435−1446.(9) Gates, K. S. (2009) An overview of chemical processes thatdamage cellular DNA: spontaneous hydrolysis, alkylation, andreactions with radicals. Chem. Res. Toxicol. 22, 1747−1760.(10) Gates, K. S., Nooner, T., and Dutta, S. (2004) Biologicallyrelevant chemical reactions of N7-alkyl-2′-deoxyguanosine adducts inDNA. Chem. Res. Toxicol. 17, 839−856.(11) Haapala, E., Hakala, K., Jokipelto, E., Vilpo, J., and Hovinen, J.(2001) N,N-Bis(2-chloroethyl)-p-aminophenylbutyric acid (chloram-bucil) with 2′-deoxyguanosine. Chem. Res. Toxicol. 14, 988−995.(12) Balcome, S., Park, S., Quirk Dorr, D. R., Hafner, L., Phillips, L.,and Tretyakova, N. (2004) Adenine-containing DNA-DNA cross-linksof antitumor nitrogen mustards. Chem. Res. Toxicol. 17, 950−962.(13) Dutta, S., Abe, H., Aoyagi, S., Kibayashi, C., and Gates, K. S.(2005) DNA damage by fasicularin. J. Am. Chem. Soc. 127, 15004−15005.(14) Millard, J. T., Raucher, S., and Hopkins, P. B. (1990)Mechlorethamine cross-links deoxyguanosine residues at 5′-GNCsequences in duplex DNA fragments. J. Am. Chem. Soc. 112, 2459−2460.(15) Maccubbin, A. F., Caballes, L., Riordan, J. M., Huang, D. H., andGurtoo, H. L. (1991) A cyclophosphamide/DNA phosphodiesteradduct formed in vitro and in vivo. Cancer Res. 51, 886−892.(16) Kohn, K. W., Hartley, J. A., and Mattes, W. B. (1987)Mechanisms of DNA sequence selective alkylation of guanine-N7positions by nitrogen mustards. Nucleic Acids Res. 15, 10531−10549.(17) Florea-Wang, D., Haapala, E., Mattinen, J., Hakala, K., Vilpo, J.,and Hovinen, J. (2004) Reactions of N,N-Bis(2-chloroethyl)-p-

Biochemistry Article

DOI: 10.1021/acs.biochem.6b01080Biochemistry 2016, 55, 7033−7041

7039

aminophenylbutyric Acid (Chlorambucil) with 2′-Deoxycytidine, 2′-Deoxy-5-methylcytidine, and Thymidine. Chem. Res. Toxicol. 17, 383−391.(18) Florea-Wang, D., Haapala, E., Mattinen, J., Hakala, K., Vilpo, J.,and Hovinen, J. (2003) Reactions of N,N-Bis(2-chloroethyl)-p-aminophenylbutyric Acid (Chlorambucil) with 2′-Deoxyadenosine.Chem. Res. Toxicol. 16, 403−408.(19) Osborne, M. R., and Lawley, P. D. (1993) Alkylation of DNA bymelphalan with special reference to adenine derivatives and adenine-guanine cross-linking. Chem.-Biol. Interact. 89, 49−60.(20) Bank, B. B. (1992) Studies of chlorambucil-DNA adducts.Biochem. Pharmacol. 44, 571−575.(21) Price, C. C., Gaucher, M., Koneru, P., Shibakawa, R., Sowa, J. R.,and Yamaguchi, M. (1968) Relative reactivities for monofunctionalnitrogen mustard alkylation of nucleic acid components. Biochim.Biophys. Acta, Nucleic Acids Protein Synth. 166, 327−359.(22) Wijen, J. P. H., Nivard, M. J. M., and Vogel, E. W. (2000) The invivo genetic activity profile of the monofunctional nitrogen mustard 2-chloroethylamine differs drastically from its bifunctional counterpartmechlorethamine. Carcinogenesis 21, 1859−1867.(23) Yaghi, B. M., Turner, P. M., Denny, W. A., Turner, P. R.,O’Connor, C. J., and Ferguson, L. R. (1998) Comparative mutationalspectra of the nitrogen mustard chlorambucil and its half-mustardanalogue in Chinese hamster AS52 cells. Mutat. Res., Fundam. Mol.Mech. Mutagen. 401, 153−164.(24) Dong, Q., Barsky, D., Colvin, M. E., Melius, C. F., Ludeman, L.M., Moravek, J. F., Colvin, M., Bigner, D. D., Modrich, P., andFriedman, H. S. (1995) A structural basis for a phosphoramidemustard-induced DNA interstrand cross-link at 5′-d(GAC). Proc. Natl.Acad. Sci. U. S. A. 92, 12170−12174.(25) Ojwang, J. O., Grueneberg, D. A., and Loechler, E. L. (1989)Synthesis of a duplex oligonucleotide containing a nitrogen mustardinterstrand DNA-DNA cross-link. Cancer Res. 49, 6529−6537.(26) Rink, S. M., Solomon, M. S., Taylor, M. J., Rajur, S. B.,McLaughlin, L. W., and Hopkins, P. B. (1993) Covalent structure of anitrogen mustard-induced DNA interstrand cross-link: an N7-to-N7linkage of deoxyguanosine residues at the duplex sequence 5′-d(GNC). J. Am. Chem. Soc. 115, 2551−2557.(27) Hopkins, P. B., Millard, J. T., Woo, J., Weidner, M. F., Kirchner,J. J., Sigurdsson, S. T., and Raucher, S. (1991) Sequence preferences ofDNA interstrand cross-linking agents: Importance of minimal DNAstructural reorganization in the cross-linking reactions of mechloreth-amine, cisplatin, and mitomycin C. Tetrahedron 47, 2475−2489.(28) Sawyer, G. A., Frederick, E. D., and Millard, J. T. (2004)Flanking sequences modulate diepoxide and mustard cross-linkingefficiencies at the 5′-GNC site. Chem. Res. Toxicol. 17, 1057−1063.(29) Osborne, M. R., Wilman, D. E. V., and Lawley, P. D. (1995)Alkylation of DNA by the nitrogen mustard bis(2-chloroethyl)-methylamine. Chem. Res. Toxicol. 8, 316−320.(30) Bauer, G. B., and Povirk, L. F. (1997) Specificity and kinetics ofinterstrand and intrastrand bifunctional alkylation by nitrogenmustards at a G-G-C sequence. Nucleic Acids Res. 25, 1211−1218.(31) Masta, A., Gray, P. J., and Phillips, D. R. (1994) Molecular basisof nitrogen mustard effects on transcription processes: role ofdepurination. Nucleic Acids Res. 22, 3880−3886.(32) Catalano, M. J., Liu, S., Andersen, N., Yang, Z., Johnson, K. M.,Price, N. A., Wang, Y., and Gates, K. S. (2015) Chemical structure andproperties of the interstrand cross-link formed by the reaction ofguanine residues with abasic sites in duplex DNA. J. Am. Chem. Soc.137, 3933−3945.(33) Catalano, M. J., Price, N. E., and Gates, K. S. (2016) Effectivemolarity in a nucleic acid controlled reaction. Bioorg. Med. Chem. Lett.26, 2627−2630.(34) Price, N. E., Catalano, M. J., Liu, S., Wang, Y., and Gates, K. S.(2015) Chemical and structural characterization of interstrand cross-links formed between abasic sites and adenine residue in duplex DNA.Nucleic Acids Res. 43, 3434−3441.(35) Price, N. E., Johnson, K. M., Wang, J., Fekry, M., Wang, Y., andGates, K. S. (2014) Interstrand DNA−DNA Cross-Link Formation

Between Adenine Residues and Abasic Sites in Duplex DNA. J. Am.Chem. Soc. 136, 3483−3490.(36) Yang, Z., Johnson, K. M., Price, N. E., and Gates, K. S. (2015)Characterization of interstrand DNA-DNA cross-links derived fromabasic sites using bacteriophage φ29 DNA polymerase. Biochemistry54, 4259−4266.(37) Johnson, K. M., Price, N. E., Wang, J., Fekry, M. I., Dutta, S.,Seiner, D. R., Wang, Y., and Gates, K. S. (2013) On the Formation andProperties of Interstrand DNA-DNA Cross-links Forged by Reactionof an Abasic Site With the Opposing Guanine Residue of 5′-CApSequences in Duplex DNA. J. Am. Chem. Soc. 135, 1015−1025.(38) Dutta, S., Chowdhury, G., and Gates, K. S. (2007) Interstrandcrosslinks generated by abasic sites in duplex DNA. J. Am. Chem. Soc.129, 1852−1853.(39) Gamboa Varela, J., and Gates, K. S. (2015) A Simple, High-YieldSynthesis of DNA Duplexes Containing a Covalent, Thermally-Reversible Interstrand Cross-link At a Defined Location. Angew. Chem.,Int. Ed. 54, 7666−7669.(40) Gamboa Varela, J., and Gates, K. S. (2016) Simple, high-yieldsyntheses of DNA duplexes containing interstrand DNA-DNA cross-links between an N4-aminocytidine residue and an abasic site. Curr.Protoc. Nucleic Acid Chem. 65, 5.16.1−5.16.15.(41) Sczepanski, J. T., Jacobs, A. C., and Greenberg, M. M. (2008)Self-promoted DNA interstrand cross-link formation by an abasic site.J. Am. Chem. Soc. 130, 9646−9647.(42) Sczepanski, J. T., Jacobs, A. C., Majumdar, A., and Greenberg,M. M. (2009) Scope and mechanism of interstrand crosslink formationby the C4′-oxidized abasic site. J. Am. Chem. Soc. 131, 11132−11139.(43) Chen, J., Dupradeau, F.-Y., Case, D. A., Turner, C. J., andStubbe, J. (2007) DNA oligonucleotides with A, T, G and C oppositean abasic site: structure and dynamics. Nucleic Acids Res. 36, 253−262.(44) Sambrook, J., Fritsch, E. F., and Maniatis, T. (1989) MolecularCloning: A Laboratory Manual, Cold Spring Harbor Laboratory Press,Plainview, NY.(45) Gates, K. S. (2000) Mechanisms of DNA damage by leinamycin.Chem. Res. Toxicol. 13, 953−956.(46) Maxam, A. M., and Gilbert, W. (1980) Sequencing end-labeledDNA with base-specific chemical cleavages. Methods Enzymol. 65,499−560.(47) Luce, R. A., and Hopkins, P. B. (2001) Chemical cross-linking ofdrugs to DNA. Methods Enzymol. 340, 396−412.(48) Hartley, J. A., Berardini, M. D., and Souhami, R. L. (1991) Anagarose gel method for the determination of DNA interstrand cross-linking applicable to the measurement of the rate of total and ″secondarm″ cross-linking reactions. Anal. Biochem. 193, 131−134.(49) Price, C. C. (1958) Fundamental mechanisms of alkylation. Ann.N. Y. Acad. Sci. 68, 663−668.(50) Rutman, R. J., Chun, E. H., and Jones, J. (1969) Observationson the mechanism of the alkylation reaction between nitrogen mustardand DNA. Biochim. Biophys. Acta, Nucleic Acids Protein Synth. 174,663−673.(51) Ulfjohn, A., Kramer, S. P., Dorfman, H., Witten, B., Williamson,C., Sass, S., Miller, J. I., and Seligman, A. M. (1964) Duration of themyelotoxic effects of circulating alkylating agents. Cancer Res. 24,1659−1665.(52) Varshney, U., and van de Sande, J. H. (1991) Specificities andkinetics of uracil excision from uracil-containing DNA oligomers byEscherichia coli uracil DNA glycosylase. Biochemistry 30, 4055−4061.(53) Lindahl, T., Ljunquist, S., Siegert, W., Nyberg, B., and Sperens,B. (1977) DNA N-glycosidases: properties of uracil-DNA glycosidasefrom Escherichia coli. J. Biol. Chem. 252, 3286−3294.(54) Romero, R. M., Rojsittisak, P., and Haworth, I. S. (2013)Electrophoretic mobility of duplex DNA cross-linked by mechloreth-amine at a cytosine-cytosine mismatch pair. Electrophoresis 34, 917−924.(55) Millard, J. T., Weidner, M. F., Kirchner, J. J., Ribeiro, S., andHopkins, P. B. (1991) Sequence preferences of DNA interstrandcross-linking agents: quantitation of interstrand crosslink locations in

Biochemistry Article

DOI: 10.1021/acs.biochem.6b01080Biochemistry 2016, 55, 7033−7041

7040

DNA duplex fragments containing multiple crosslinkable sites. NucleicAcids Res. 19, 1885−1891.(56) Kumaresan, K. R., Ramaswamy, M., and Yeung, A. T. (1992)Structure of the DNA interstrand cross-link of 4,5′,8-trimethylpsor-alen. Biochemistry 31, 6774−6783.(57) Wilson, D. M., III, and Barsky, D. (2001) The major humanabasic endonuclease: formation, consequences and repair of abasicsites in DNA. Mutat. Res., DNA Repair 485, 283−307.(58) Rosa, S., Fortini, P., Karran, P., Bignami, M., and Dogliotti, E.(1991) Processing in vitro of an abasic site reacted with methoxy-amine: a new assay for the detection of abasic sites formed in vivo.Nucleic Acids Res. 19, 5569−5574.(59) Habraken, Y., and Ludlum, D. B. (1989) Release of chloroethylethyl sulfide-modified bases by bacterial 3-methyladenine-DNAglycosylases I and II. Carcinogenesis 10, 489−492.(60) Zang, H., and Gates, K. S. (2003) Sequence specificity of DNAalkylation by the antitumor natural product leinamycin. Chem. Res.Toxicol. 16, 1539−1546.(61) Asai, A., Hara, M., Kakita, S., Kanda, Y., Yoshida, M., Saito, H.,and Saitoh, Y. (1996) Thiol-mediated DNA alkylation by the novelantitumor antibiotic leinamycin. J. Am. Chem. Soc. 118, 6802−6803.(62) Fekry, M., Szekely, J., Dutta, S., Breydo, L., Zang, H., and Gates,K. S. (2011) Noncovalent DNA Binding Drives DNA Alkylation byLeinamycin. Evidence That the Z,E-5-(Thiazol-4-yl)-penta-2,4-dien-one Moiety of the Natural Product Serves As An Atypical DNAIntercalator. J. Am. Chem. Soc. 133, 17641−17651.(63) Nooner, T., Dutta, S., and Gates, K. S. (2004) Chemicalproperties of the leinamycin 2′-deoxyguanosine adduct. Chem. Res.Toxicol. 17, 942−949.(64) Schar̈er, O. D. (2005) DNA interstrand crosslinks: natural anddrug-induced DNA adducts that induce unique cellular responses.ChemBioChem 6, 27−32.(65) Guillet, M., and Boiteux, S. (2003) Origin of endogenous DNAabasic sites in Saccaromyces cerevisiae. Mol. Cell. Biol. 23, 8386−8394.(66) Jian, Y., Lin, G., Chomicz, L., and Li, L. (2015) Reactivity ofdamaged pyrimidines: formation of a Schiff base intermediate at theglycosidic bond of saturated dihydrouridine. J. Am. Chem. Soc. 137,3318−3329.(67) Hoehn, S. T., Turner, C. J., and Stubbe, J. (2001) Solutionstructure of an oligonucleotide containing an abasic site: evidence foran unusual deoxyribose conformation. Nucleic Acids Res. 29, 3413−3423.(68) Lhomme, J., Constant, J.-F., and Demeunynck, M. (1999)Abasic DNA structure, reactivity, and recognition. Biopolymers 52, 65−83.(69) Lindahl, T., and Nyberg, B. (1972) Rate of depurination ofnative deoxyribonucleic acid. Biochemistry 11, 3610−3618.(70) Chakravarti, D., Mailander, P. C., Li, K.-M., Higginbotham, S.,Zhang, H. L., Gross, M. L., Meza, J. L., Cavalieri, E. L., and Rogan, E.G. (2001) Evidence that a burst of DNA depurination in SENCARmouse skin induces error-prone repair and forms mutations in the H-ras gene. Oncogene 20, 7945−7953.(71) Rinne, M. L., He, Y., Pachkowski, B. F., Nakamura, J., andKelley, M. R. (2005) N-methylpurine DNA glycosylase overexpressionincreases alkylation sensitivity by rapidly removing non-toxic 7-methylguanine adducts. Nucleic Acids Res. 33, 2859−2867.

Biochemistry Article

DOI: 10.1021/acs.biochem.6b01080Biochemistry 2016, 55, 7033−7041

7041