Bioorganic & Medicinal Chemistry Letters 22 (2012) 863–867

Contents lists available at SciVerse ScienceDirect

Bioorganic & Medicinal Chemistry Letters

journal homepage: www.elsevier .com/ locate/bmcl

A DNA aptamer sensor for 8-oxo-7,8-dihydroguanine

Jyoti Roy, Payal Chirania, Sanniv Ganguly, Haidong Huang ⇑Department of Chemistry and Environmental Science, New Jersey Institute of Technology, Newark, NJ 07102, USA

a r t i c l e i n f o a b s t r a c t

Article history:Received 20 October 2011Revised 6 December 2011Accepted 8 December 2011Available online 13 December 2011

Keywords:Aptamer sensorDNA damage

0960-894X/$ - see front matter � 2011 Elsevier Ltd. Adoi:10.1016/j.bmcl.2011.12.046

⇑ Corresponding author. Tel.: +1 973 596 3576; faxE-mail address: haidong.huang@njit.edu (H. Huang

Abasic site-containing DNA duplex is a versatile structural motif that can be used for the design of purineaptamers and sensors. In this study, several modifications were introduced to the abasic site to explorepossible specific binding of free 8-oxoG, a product of DNA base excision repair. The nucleoside oppositethe abasic site was replaced by pyrrolo-dC as a reporter group. Binding of 8-oxoG quenched pyrrolo-dCfluorescence by as much as 70%. In contrast, adenine, guanine, thymine, and cytosine showed only min-imal fluorescence quenching effect. The best aptamer binds 8-oxoG with a dissociation constant of5.5 ± 0.8 lM. This sensor can be used to accurately measure 8-oxoG concentrations in the presence ofguanine.

� 2011 Elsevier Ltd. All rights reserved.

It is well accepted that one major cause of human cancer is DNAdamage induced by environmental factors. One of the most abun-dant and widely studied DNA damage products is 8-oxo-7,8-dihy-droguanine (8-oxoG), which is formed via the oxidation of guaninewhen DNA is exposed to oxidative stress.1 8-OxoG causes muta-tions by generating G?T transversions.2 Once formed, 8-oxoG inDNA can be cleaved by DNA repair enzymes and released intothe extracellular fluid. The level of free 8-oxoG may reflect theoverall level of DNA damage and can potentially serve as an indica-tor for early cancer development.3

Fluorescence methods can be used to detect free 8-oxoG or8-oxodG (8-oxo-7,8-dihydro-20-deoxyguanosine) in DNA in a fastand inexpensive way compared to various GC- and LC-based meth-ods.4 Fluorescence detection of free 8-oxoG has been reportedusing an antibody-conjugated assay.5 8-OxodG in a DNA sequencecan also be detected by fluorescence techniques using a spermine–biotin conjugate or 8-oxodG clamps.6,7 Surprisingly, not muchwork has been reported on the development of fluorescent 8-oxoGaptamer sensors. To the best of our knowledge, the only relatedexamples are the 8-oxodG DNA aptamers developed by Rink et aland Miyachi et al.8 These aptamers bind 8-oxodG with high affinity(Kd in the order of 0.1 lM). At present, no aptamer has beenreported to selectively bind a free 8-oxoG base.

We and other groups have developed a rational method todesign DNA aptamers for aromatic molecules.9,10 We have previ-ously reported a triplex DNA aptamer sensor that binds adenosinewith high binding affinity and selectivity.10 The cavity of theaptamer is created using an abasic site to replace a nucleobase ornucleoside. A fluorescent base is incorporated opposite the abasic

ll rights reserved.

: +1 973 596 3586.).

site as the reporter. We intended to apply the same strategy tothe design of an 8-oxoG aptamer sensor using the 8-oxodGclamp-containing duplex DNA as a guide.7a The fluorescence of8-oxodG clamp in a duplex is quenched when the opposite nucle-oside is 8-oxodG. If the 8-oxodG was replaced with an abasic site,the cavity would function as a binding pocket for a free 8-oxoGbase (Fig. 1a). However, when we tested such a duplex DNA, nofluorescence change was observed when 8-oxoG was added. Wespeculated that the phenyl group of the 8-oxodG clamp might haveoccupied the cavity and that the binding of 8-oxoG was not strongenough to drive the phenyl ring away from the cavity. Thus, aro-matic rings should be avoided when designing binding groups tospecifically interact with 8-oxoG via non-Watson–Crick hydrogenbonds. Here, we report a strategy of introducing additional ali-phatic groups that bind 8-oxoG into the abasic site of a duplexDNA aptamer.

8-OxoG has multiple hydrogen bonding sites in addition to theWatson-Crick hydrogen bonding sites. We envisioned that attach-ing an aliphatic side chain containing terminal functional groups toa threoninol spacer could be a convenient method to introducesuch interactions. To determine the correct side chain orientation,we synthesized two structures (DTC3 and LTC3) with the sidechains in opposite configurations (Fig. 1b). The phosphoramiditesof DTC3 and LTC3 (6a, b) were prepared from Boc-b-alanine andthreoninol (Scheme 1). We attempted to change the protectinggroup on the terminal amine from Boc to trifluoroacetate duringthe course of the synthesis. However, a mixture of inseparablematerials was obtained following the removal of Boc from 5a and5b. It was likely that cyclization occurred under the strong acidicconditions used for the Boc deprotection. Thus, the Boc protectinggroup remained until the end of the oligonucleotide synthesis. Thephosphoramidites 6a and 6b were incorporated into oligonucleo-tides 2 and 3, respectively. The Boc group was removed with

NN

NH

O

O

ONH

OO

OO

O O

O

Proposed8-oxoGbinding site

NN

NH

O

O

O

OO

O O

O

NHO O

5' 3' 5' 3'

(a)

DNA 1 X = O O5' 3'

X = O O5' 3'

HN O

NH2

DNA 3

DNA 4

X = O O5' 3'

HN O

NH2

X = O O5' 3'

HN O

DNA 2

NH2O

C3

DTC 3

LTC3

Y = O

O

O N

N

O

NH

pC

(b) 5'-TTT AGC CTA ATX CGG ATC GCT CC-3'3'-AAA TCG GAT TAY GCC TAG CGA GG-5'

DNA Duplexes 1-4

Figure 1. (a) 8-OxodG clamp opposite an abasic site. Left panel, initially proposed cavity for the binding of 8-oxoG. Right panel, possible binding of the benzene side chain of8-oxoG clamp in the cavity. (b) DNA duplexes used in this study as potential aptamer sensors for 8-oxoG.

864 J. Roy et al. / Bioorg. Med. Chem. Lett. 22 (2012) 863–867

1:10 TFA/CH2Cl2 using the protocol described by James et al beforethe resin was subjected to ammonia deprotection.11

The fluorescent reporter used in this study was pyrrolo-dC (pC),a fluorescent C analogue known to be sensitive to structuralchanges.15 pC was incorporated into the complementary DNAstrand opposite the abasic sites of DNAs 1, 2, and 3 (Fig. 1b). Thebinding of the duplexes (designated as duplexes 1, 2, and 3) with8-oxoG was examined by fluorescence titration. A concentratedDMSO solution of 8-oxoG was used for the titration due to the lim-ited aqueous solubility of 8-oxoG. An independent titration exper-iment with blank DMSO showed that DMSO alone did not have anyimpact on the fluorescence spectrum of the DNA duplex. Figure 2ashows the fluorescence titration of duplex 2 with the DMSO solu-tion of 8-oxoG excited at 290 nm. The observed maximum percent

fluorescence quenching at an emission wavelength of 446 nm was70%, which corresponded to approximately 120 lM of 8-oxoG.Further addition of 8-oxoG caused precipitation due to the lowsolubility of this base. We also observed that, when the 8-oxoGconcentration was between 50 lM and 120 lM, the quenching ef-fect became linear and would not reach saturation even if 8-oxoGwas more soluble (Fig. 2b). This apparent unsaturation effect athigh 8-oxoG concentrations can be attributed to collisionalquenching. The fluorescence intensity at 380 nm increased duringthe titration. This increase of fluorescence intensity was caused bythe increasing concentration of 8-oxoG, which emits fluorescencewhen excited at 290 nm (see Fig. S4 for 8-oxoG emission and exci-tation spectra). A Job plot at low 8-oxoG concentrations indicatedthat the binding mode between the base and duplex 2 was 1:1

Figure 2. (a) Fluorescence titration of duplex 2 with 8-oxoG. The fluorophore was excited at 290 nm. The DNA solution contained 400 nM duplex 2, 20 mM NaCl, 2 mMMgCl2, and 10 mM PIPES (pH 7.5). 8-OxoG concentrations were as follows: 0, 3.68 lM, 7.36 lM, 11.04 lM, 14.72 lM, 18.40 lM, 22.45 lM, 30.54 lM, 38.64 lM, 50.78 lM,69.18 lM, 96.78 lM, 124.38 lM, 142.78 lM. (b) Plot of F0/F versus 8-oxoG concentration. Fluorescence intensity was recorded at 446 nm. (c) Plot of F/F0 versus 8-oxoGconcentration. Fluorescence intensity was recorded at 446 nm. (d) Job plot for the titration. The total concentration of duplex 2 and 8-oxoG was 1 lM. f8-oxoG stands for themole fraction of 8-oxoG ([8-oxoG]/[8-oxoG] + [duplex 2]). Fluorescence intensity was recorded at 446 nm.

J. Roy et al. / Bioorg. Med. Chem. Lett. 22 (2012) 863–867 865

(Fig. 2d). Linear and nonlinear fitting (see Supplementary data)were used to calculate the collisional quenching constant and thedissociation constant, respectively (Fig. 2b, c).12 The collisionalquenching constant (KSV) calculated for duplex 2 was 7.83 mM�1.The dissociation constant (Kd) was 5.1 lM, which was comparableto similar aptamer systems.9,10

The fluorescence emissions of duplexes 1 and 3 were alsoquenched when 8-oxoG was added, and the dissociation constantswere 12.1 lM and 18.2 lM, respectively (Table 1). Comparison ofthe dissociation constants of the three duplexes implies that, whenthe side chain of the threoninol moiety is in the D configuration, theterminal group of the side chain is more accessible to interact with8-oxoG.

We examined the selectivity of duplex 2 for 8-oxoG over othernatural nucleobases. Minimal fluorescence quenching was ob-served for adenine, thymine, cytosine, and guanine (Fig. 3). Unlikethe titration with 8-oxoG, the quenching induced by the naturalbases reached saturation. This result indicates that no collisionalquenching occurs between the natural bases and pC. The dissocia-tion constants were not calculated for adenine, thymine, guanine,and cytosine due to the small magnitude of the observed changesin fluorescence. Instead, the quenching abilities were estimated asthe concentrations of half maximum quenching ([L]DF1/2, Table 1).Because commercial guanine has some fluorescence emission at446 nm, the fluorescence intensities were corrected for back-ground (Fig. S1). Guanine showed the most significant quenching

Table 1Constants derived from fluorescence titration experiments. KSV is the Stern–Volmerconstant for collisional quenching. [L]DF1/2 is the concentration of half maximumfluorescence quenching

Base Constant Duplex 1 Duplex 2 Duplex 3

8-OxoG Kd (lM) 12.5 ± 1.3a 5.5 ± 0.8a 16.5 ± 2.2a

8-OxoG KSV (mM�1) 7.67 ± 0.44a 7.72 ± 0.52a 7.42 ± 0.28a

Adenine [L]DF1/2 (lM) 27 34 NDc

Cytosine [L]DF1/2 (lM) NDb 22 NDc

Guanine [L]DF1/2 (lM) 8 15 NDc

Thymine [L]DF1/2 (lM) NDb 28 NDc

a Average of three independent titration experiments.b Not determined because fluorescence changes are smaller than the standard

deviation of the measurement.c Fluorescence titrations were not performed. Figure 4. Measurement of 8-oxoG concentration in the presence and absence of

guanine. The concentration of duplex 2 was 400 nM. Black square, in the presenceof 4 lM guanine. Red circle, in the absence of guanine. Fluorescence intensity wasrecorded at 446 nm.

N NNH

OO

O

O O

5'

5'3'

3'

HN NH2N

NH

HNO

OO

NH

H3N

O

Figure 5. Predicted binding mode between duplex 2 and 8-oxoG.

866 J. Roy et al. / Bioorg. Med. Chem. Lett. 22 (2012) 863–867

ability among the four natural bases. To further understand theinfluence of guanine on the measurement of 8-oxoG concentra-tions, we added 8-oxoG to duplex 2 in the presence and absenceof 4 lM guanine in a side-by-side experiment. A concentrationrange of 1–4 lM 8-oxoG was used in this study; the collisionalquenching was not significant in this range. pC fluorescencequenching was observed when 8-oxoG was added despite the pres-ence of guanine (Fig. 4). The magnitudes of the errors arising fromthe guanine interference depended on the relative abundance of8-oxoG and guanine. When the 8-oxoG concentration was low,guanine introduced negative errors into the fluorescence readings.The opposite effect was observed when the 8-oxoG concentrationwas high.

Pyrrolo-dC does not contain any additional binding groupsother than the Watson–Crick hydrogen bonding sites and cannotbe used to selectively distinguish between 8-oxoG and guanine.However, the fluorescence quenching was much more dramaticwhen 8-oxoG was added to duplex 1 compared to the addition ofguanine. A greater quenching effect induced by 8-oxodG was pre-viously observed for the fluorescent base furano-dC and the 8-ox-odG clamp, even though 8-oxodG destabilized the DNA duplexwhen incorporated opposite these fluorescent bases.7a,13 All three

Figure 3. Fluorescence spectra of duplex 2 in the presence of adenine (a), thymine (b), cytosine (c), and guanine (d). Red lines represent duplex 2 in the absence of the bases.Blue lines represent duplex 2 in the presence of 20 lM of each base.

HO OHNH2

(D)- or (R)-threoninol

R OH

Oa HO OH

HN O

R

5a (D), R = NHBoc5b (L), R = NHBoc7 (D), R = CO2Me

b, c DMTrO OHN O

R

PN(iPr)2

O

CN

6a (D), R = NHBoc6b (L), R = NHBoc8 (D), R = CO2Me

Scheme 1. Synthesis of the phosphoramidites of DTC3 and LTC3. Reagents and conditions: (a) EDC, DMAP, Et3N, DMF, 4 h. (b) Dimethoxytrityl chloride, Et3N, DMAP, CH2Cl2,12 h. (c) N,N0-diisopropyl 2-cyanoethyl phosphoramidic chloride, N,N0-diisopropylethylamine, CH2Cl2, 2 h.

J. Roy et al. / Bioorg. Med. Chem. Lett. 22 (2012) 863–867 867

examples demonstrate the greater inherent fluorescence quench-ing ability of 8-oxoG compared to guanine when quenching occursthrough the formation of hydrogen bonds. Other than the differen-tiation of the bases by their inherent quenching abilities, a selec-tive sensor is also required to selectively bind the desired base.The 8-oxodG-clamp monomer has a higher affinity for the 8-oxodGmonomer than for the dG monomer.14 Nevertheless, these bindingassays must be carried out in detergent due to the weakenedhydrogen bonds in protic solvents. This requirement for detergentis expected to be circumvented by designing DNA aptamers thatcontain larger hydrophobic binding surfaces. Our DNA aptamer(2) showed such selectivity in an aqueous environment. The mea-surement of 8-oxoG was highly accurate, although some errorscaused by the interference of guanine could be observed. The bind-ing mode of duplex 2 to 8-oxoG is predicted in Figure 5. The aminogroup may interact with the 8-oxo group of 8-oxoG via hydrogenbonding. This additional interaction could explain why the affinityand selectivity of 2 are the best among the three duplexes,although we cannot exclude the possibility that the amino groupinteracts with a phosphate group of DNA and changes the overallaptamer structure.

To further demonstrate that the amino group is important forbinding, we synthesized a succinamide side chain (4) to replacethe b-alanine. The phosphoramidite of the succinate ester (8)was synthesized (Scheme 1) and incorporated into DNA 4. Theester group was aminolyzed to form carboxylamide during theammonia deprotection following the solid phase synthesis. Titra-tion of as much as 10 lM of 8-oxoG to duplex DNA 4 did not gen-erate any fluorescence change (Fig. S2). This result indicates that acarboxylamide side chain disrupts the aptamer structure to anextent that greatly affects binding.

In summary, we have developed the first aptamer sensor for8-oxoG using a rational approach. The fluorescence change atlow 8-oxoG concentrations arose from specific binding of the baseto the fluorophore. The combined effects of the moderately selec-tive binding and the greater fluorescence quenching ability of8-oxoG resulted in a highly selective method to detect free 8-oxoG.Although the affinity and selectivity of this aptamer is not compa-rable to 8-oxoG antibody, it is smaller, easier to produce, more sta-ble, and does not require secondary reporters when used as asensor. The detection limit of sensor 2 is 100 nM, estimated fromthe standard deviation of the fluorescence signal in the absenceof 8-oxoG (ca. 2 a.u. under our experiment conditions) and theslope of the linear dependence (Fig. 4). This sensor can be used

to accurately measure 8-oxoG concentrations in the range between100 nM and 5 lM.

Acknowledgement

We thank New Jersey Institute of Technology for financialsupport.

Supplementary data

Supplementary data associated with this article can be found, inthe online version, at doi:10.1016/j.bmcl.2011.12.046.

References and notes

1. Gajewski, E.; Rao, G.; Nackerdien, Z.; Dizdaroglu, M. Biochemistry 1990, 29,7876.

2. Neeley, W. L.; Essigmann, J. M. Chem. Res. Toxicol. 2006, 19, 491.3. Shigenaga, M. K.; Gimeno, C. J.; Ames, B. N. Proc. Natl. Acad. Sci. U.S.A. 1989, 86,

9697.4. (a) Helbock, H. J.; Beckman, K. B.; Shigenaga, M. K.; Walter, P. B.; Woodall, A. A.;

Yeo, H. C.; Ames, B. N. Proc. Natl. Acad. Sci. U.S.A. 1998, 95, 288; (b) Malayappan,B.; Garret, T. J.; Segal, M.; Leeuwenburgh, C. J. Chromatogr. A 2007, 1167, 54; (c)Weimann, A.; Belling, D.; Poulsen, H. E. Nucleic. Acids. Res. 2002, 30, e7; (d)Ravanat, J.-L.; Turesky, R. J.; Gremaud, E.; Trudel, L. J.; Stadler, R. H. Chem. Res.Toxicol. 1995, 8, 1039; (e) Mangal, D.; Vudathala, D.; Park, J. H.; Lee, S. H.;Penning, T. M.; Blair, I. A. Chem. Res. Toxicol. 2009, 22, 788.

5. Park, E.-M.; Shigenaga, M. K.; Degan, P.; Korn, T. S.; Kitzler, J. W.; Wehr, C. M.;Kolachana, P.; Ames, B. N. Proc. Natl. Acad. Sci. U.S.A. 1992, 89, 3375.

6. Xue, L.; Greenberg, M. M. J. Am. Chem. Soc. 2007, 129, 7010.7. (a) Nasr, T.; Li, Z.; Nakagawa, O.; Taniguchi, Y.; Ono, S.; Sasaki, S. Bioorg. Med.

Chem. Lett. 2009, 19, 727; (b) Taniguchi, Y.; Kawaguchi, R.; Sasaki, S. J. Am.Chem. Soc. 2011, 133, 7272.

8. (a) Rink, S. M.; Shen, J.-C.; Loeb, L. A. Proc. Natl. Acad. Sci. U.S.A. 1998, 95, 11619;(b) Miyachi, Y.; Shimizu, N.; Ogino, C.; Fukuda, H.; Kondo, A. Bioorg. Med. Chem.Lett. 2009, 19, 3619.

9. (a) Coppel, Y.; Constant, J.-F.; Coulombeau, C.; Demeunynck, M.; Garcia, J.;Lhomme, J. Biochemistry 1997, 36, 4831; (b) Sankaran, N. B.; Nishizawa, S.;Seino, T.; Yoshimoto, K.; Teramae, N. Angew. Chem., Int. Ed. 2006, 45, 1563; (c)Li, M.; Sato, Y.; Nishizawa, S.; Seino, T.; Nakamura, K.; Teramae, N. J. Am. Chem.Soc. 2009, 131, 2448; (d) Thiagarajan, V.; Rajendran, A.; Satake, H.; Nishizawa,S.; Teramae, N. ChemBioChem 2010, 11, 94.

10. Patel, M.; Dutta, A.; Huang, H. Anal. Bioanal. Chem. 2011, 400, 3035.11. James, K. D.; von Kiedrowski, G.; Ellington, A. D. Nucleos. Nucleot. 1997, 16,

1821.12. Kramer, M. L.; Kratzin, H. D.; Schmidt, B.; Romer, A.; Windl, O.; Liemann, S.;

Hornemann, S.; Kretzschmar, H. J. Biol. Chem. 2001, 276, 16711.13. Greco, N. J.; Sinkeldam, R. W.; Tor, Y. Org. Lett. 2009, 11, 1115.14. Nakagawa, O.; Ono, S.; Li, Z.; Tsujimoto, A.; Sasaki, S. Angew. Chem., Int. Ed.

2007, 46, 4500.15. (a) Thompson, K. C.; Miyake, N. J. Phys. Chem. B 2005, 109, 6012; (b) Marti, A. A.;

Jockusch, S.; Li, Z.; Ju, J.; Turro, N. Nucleic Acids Res. 2006, 34, e50.