DNA modification promoted by water-soluble nickel(II) salen complexes: A switch to DNA alkylation

DNA Modification Promoted by Water-Soluble Nickel(II) Salen Complexes : A Switch to DNA Alkylation James G. Muller, Sari J . Paikoff, Steven E. Rokita, and Cynthia J . Burrows JGM, SJP, SER, CJB . Department of Chemistry, State University of New York at Stony Brook, Stony Brook, New York, U.S.

ABSTRACT Reaction of a 17-base hairpin-forming oligonucleotide with [N,N'-bis(salicylaldehyde)-meso-1,2bis(4-trimethylaminophenyl)ethylenediiminolnickel(II) perchlorate, 2, and KHSO 5 produced two types of high molecular weight products, an alkaline-labile species and a nonalkaline-labile species, which co-migrated on gel electrophoresis . Upon treatment with piperidine, the base-labile derivative led to strand scission products only at accessible guanine residues that were not part of a Watson-Crick duplex. The formation of higher molecular weight species is proposed to occur via a highly reactive ligand-centered radical acting as a DNA alkylating agent .

INTRODUCTION Transition metal complexes capable of interacting with nucleic acids bear potential applications in chemotherapy in characterizing nucleic acid structure [1-3], and in elucidating the mechanism of metal ion toxicity and carcinogenicity [4]. Of particular interest to chemical toxicity are the nickel-promoted reactions leading to DNA strand breaks, depurination, and DNA-protein and DNA interstrand cross-links . In these cases, nickel may ligate to bioligands such as amino acids, peptides and proteins, and in the presence of an oxidant, produce free radicals, either oxygen-, carbon-, or sulfur-centered, that may damage t+ nucleic acids [4b] . The oxidative damage of guanine residues in DNA by Ni and H 2 O2 is enhanced, for example, by addition of histidine to form a 2 :1 complex between histidine and nickel [5]. However, much remains to be learned

Address reprint requests to : Dr. Cynthia J . Burrows, Department of Chemistry, State University of New York at Stony Brook, Stony Brook, NY 11794-3400, U .S. Journal of Inorganic Biochemistry, 54, 199-206

© 1994 Elsevier Science Inc ., 655 Avenue of the Americas, NY, NY 10010

199 0162-0134/94/$7.00

200

J. G. Muller et al.

about the chemical mechanisms by which nickel complexes modify DNA bases under oxidative conditions . Recent studies in our laboratories revealed that the square planar nickel(II) complex [Ni(Me2[14]py-dieneN4 ](C104 )2 , 1, in the presence of KHSO S , causes reaction of accessible guanine residues in a wide variety of DNA structures leading to strand scission upon treatment with piperidine [6-9]. While a series of nickel(II) complexes containing tetraazamacrocycles promoted this guaninespecific modification of DNA, reactivity studies using related tetradentate N202 ligands, such as N,N'-bis(salicylaldehyde)ethylenediimine (salen) were not possible because of water insolubility [8] . Although a brief account of the cleavage of plasmid DNA by Ni(salen) and an oxidant, in an acetonitrile/water mixture, has been reported [10], no details of this chemistry are known . The examination of the DNA modification capabilities of water-soluble transition metal salen complexes has been recently explored, wherein the specific cleavage of duplex DNA at A :T regions by Mn(salen) and an oxidant has been observed [111 . Like their manganese analogues, the study of Ni(salen) complexes is of interest because of their demonstrated ability to act as catalysts for olefin epoxidation [12]. In order to evaluate the DNA modification capabilities of Ni(salen) complexes, under physiologically relevant conditions, we have synthesized the new water-soluble complexes [N,N'-bis(salicylaldehyde)-meso-1,2bis(4-trimethylaminophenyl)ethylenediimino}nickel(II) perchlorate, 2, and [N,N'bis(acetylacetone)-meso-1,2-bis (4-trimethylaminophenyl)ethylenediimino]nickel (II) perchlorate, 3 (see Scheme 1) :

(CH3)3N+ I Ni i

I

(CH3)3N+

I

2

3

Scheme 1 . Although salen ligands bearing anionic substituents were known positive charge on the complex is desirable to enhance water solu affinity.

an overall nd DNA

EXPERIMENTAL Materials The oligonucleotide, d(AGTCTATGGGI IAGACT), was obtained and purified as previously described [8] . All aqueous solutions used purified water (Nanopure, Sybron/Barnsted) and reagents of the highest commercial quality . All other chemicals were of reagent grade and used without further purification . N,N'bis(salicylaldehyde)-mesa-1,2-bis(4-dimethylaminophenyl)ethylenediimine [14),

DNA MODIFICATION PROMOTED BY NICKEL COMPLEXES

201

meso-1,2-bis(4-trimethylaminophenyl)ethylenediamine [14] and [Ni(Me 2 [14]pydieneN 4 ](C104 )2 , 1, [15] were synthesized following published procedures . Preparation of Complexes Caution! While we have used perchlorate as a counterion with a number of nickel complexes without incident, perchlorate salts of metal complexes with organic ligands are potentially explosive . Care should be exercised when using a spatula or stirring rod to mechanically agitate any solid perchlorate . These complexes, as well as any other perchlorate salt, should only be handled in small quantities . [N,N'-bis(salicylaldehyde)-meso-1,2-bis(4-trimethylaminophenyl)ethylenediimino]nickel(II) perchlorate (2) . A solution consisting of 0 .100 g of N,N'bis(salicylaldehyde)-meso-1,2-bis(4-dimethylaminophenyl)ethylenediimine (0.20 mmol) and 0 .054 g of nickel(II) acetate tetrahydrate (0 .22 mmol) in 20 mL of n-propanol was refluxed for 2 hr and then cooled to room temperature . The resulting red precipitate was collected by vacuum filtration and washed with cold water and n-propanol . Yield : 0.100 g, 90% . A sample of this solid (0 .050 g) was dissolved in 8 mL of a 1 :1 mixture of chloroform and acetonitrile, followed by the addition of 1 mL of iodomethane . After stirring at room temperature for 24 hr, an orange solid was collected by vacuum filtration and washed with chloroform and diethyl ether . Yield : 0.066 g, 88% . This solid (0 .050 g) was dissolved in 1 .5 mL of a 1 :1 mixture of acetonitrile and water and passed down an ion-exchange column (Dowex 1X8-100) in the chloride form . The red eluant was concentrated to dryness by rotary evaporation, dissolved in 2 mL of water, and a concentrated aqueous sodium perchlorate solution was added until precipitation of an orange solid was observed . The solid was collected by vacuum filtration and washed with cold water. Yield: 0.030 g, 71% . 'H NMR ((CD 3 )2 SO) : S 6.46-7.81 (m, 16 H, ArH), 7.59 (s, 2 H, CH = N), 5 .25 (s, 2 H, CH-N), 3 .50 (s, 18 H, CH 3 ) . 13 C NMR ((CD 3 )2 SO): 8 164.8, 164 .0, 147 .5, 138 .4, 135 .0, 133 .7, 130.8, 120.8, 120 .6, 120 .4, 115 .2, 75.0, 57.0. Anal . Calcd. for C34H3sN4O10C 1 2Ni(2H2O) : C, 49 .30; H, 5 .11 . Found: C, 49 .40; H, 4.99. [N,N'-bis(acetylacetone)-meso 1,2-bis(4-trimethylaminophenyl)ethylenediimino]nickel(II) perchlorate (3) . A solution containing 0 .067 g of acetylacetone (0.67 mmol), 0.083 g of nickel(II) acetate tetrahydrate (0 .33 mmol), and 0 .100 g of meso-1,2-bis(4-trimethylaminophenyl)ethylenediamine (0 .33 mmol) in 2 mL of methanol was heated to reflux under N 2 for 18 hr and then cooled to room temperature . A brown solid was collected by vacuum filtration and washed with water and diethyl ether (0.100 g, 57%) . A sample of this solid (0.050 g) was dissolved in 8 mL of a 1 :1 mixture of chloroform and acetonitrile, followed by the addition of 1 mL of iodomethane . The solution was stirred at room temperature for 24 hr. The resulting brown solution was concentrated to dryness by rotary evaporation and then dissolved in a minimum amount of dichloromethane . The brown solid obtained after the addition of diethyl ether to this solution and the crude product was collected by vacuum filtration and washed with diethyl ether . Yield: 0.071 g, 92% . Then, 0 .060 g of this solid was dissolved in 2.0 mL of a 1 :1 mixture of acetonitrile and water and passed down an ion-exchange column (Dowex 1X8-100) in the chloride form . The brown eluant was concentrated and dissolved in 2 mL of water, and a concentrated aqueous sodium perchlorate solution was added until precipitation of a tan solid was

202

J. G. Muller et al.

observed . The tan solid was collected by vacuum filtration and washed with cold water. Yield: 0.051 g, 91% . ' H NMR ((CD 3 ) 2 S0): S 7 .69 (m, 8 H, ArH), 4 .99 (s, 2 H, CH-N or N = C-CH), 4 .93 (s, 2 H, CH-N or N = C-CH) 3 .46 (s, 18 H, ArNCH3 ), 1 .73 (s, 3H, N=C-CH 3 or O-C-CH 3 ), 1 .49 (s, 3 H, N=C-CH 3 or O-C-CH 3 ) . 13C NMR ((CD 3 ) 2 S0) : 8 177 .8, 167 .8, 146 .2, 142 .8, 129.4, 120.1, 101 .5, 71 .3, 56 .9, 24 .2, 22.4. Anal. Calcd . for C30 H 42 N40 10 CI 2 Ni(H 20): C, 47.02; H, 5 .79. Found: C, 46.60; H, 5.90. DNA Studies DNA experiments were conducted as previously described [8] . Reaction mixtures (100 µL) contained final concentrations of 3 AM unlabeled oligonucleotide, 2 nCi labeled oligonucleotide, 3 AM for the desired nickel(II) complex, 10-60 µM KHSO 5 , 100 mM NaCI, and 10 mM potassium phosphate (pH 7.0). RESULTS AND DISCUSSION Synthesis of new complexes took advantage of a preparation of 1,2-diarylethylenediamines reported by Vogtle and Goldschmitt [14] in which the salen ligand 5 resulted from the [3,3] sigmatropic rearrangement of diimine 4 . Hydrolysis of 5 provided mesa-1,2-bis(4-dimethylaminophenyl)ethylenediamine [14] required in the synthesis of 3, while 5 could be used directly in the synthesis of 2 by simple complexation of nickel(II) and N-alkylation with methyl iodide (See Scheme 2) :

0 4

(CH3)3N+

5

(1) Ni(OAc)2 (2) CH3I

4

(CH3)3N+ 2

Scheme 2 . Surprisingly, experiments with oligonucleotides revealed that complex 2, in the presence of KHSO 5 , induced the formation of DNA adducts of substantially increased molecular weight rather than simple base oxidation products . Upon treatment with piperidine, partial conversion of the higher molecular weight species to guanine-specific strand scission products occurred . Such formation of higher molecular weight DNA products with complex 2 sharply contrasts with the chemistry of complex 1 and its analogues [8] which showed only DNA damage leading to fragmentation of the strand . The current investigation demonstrates how the reactivity of nickel complexes towards DNA can be modulated by design and control of the ligand environment .

DNA MODIFICATION PROMOTED BY NICKEL COMPLEXES

203

For DNA modification studies, the hairpin-forming oligonucleotide d(AGTCTATGGGTTAGACT) was chosen as the substrate because it contains both unpaired guanine residues in the loop region, G 8 _ 10 , as well as typical WatsonCrick base-paired guanine residues G 2 and G 14 . Thus, it can probe both reactivity and selectivity for G conformations [7, 8] . The reaction of the oligonucleotide (3 µM) with complex 2 (3 µM) and KHSO 5 1 (10 µM) and subsequent analysis by denaturing gel electrophoresis indicated the formation of a product exhibiting a higher molecular weight and reduced electrophoretic mobility relative to the starting oligonucleotide (see Fig . 1, lane 2). The molecular weight was estimated to be approximately that of the oligonucleotide plus the organic ligand. Some of this higher molecular weight material was alkaline labile ; when it was isolated from the gel (Fig . 1, lane 3) and treated with 0 .2 M piperidine for 30 min . at 90°C, partial strand scission at the G sites in the hairpin loop was observed (Fig. 1, lane 4). Isolation of the oligonucleotide species showing normal electrophoretic mobility (Fig . 1, lane 5) and treatment with piperidine (Fig . 1, lane 6) indicated that essentially no products leading to strand scission were formed . The formation of a higher molecular weight species was not observed in reactions employing complex 1 [6-9] where only modification leading to strand scission after piperidine treatment was observed (Fig . 1, lane 7). Conversely, DNA modification under the conditions described for complex 2 did not form the oxidized and alkaline-labile products observed with complex 1 . The data suggest that the higher molecular weight species formed in reactions with complex 2 is a covalent adduct in which complex 2 alkylates DNA bases at various sites, at least one of which (on G) leads to alkaline ability . Formation of a covalent bond between the oligonucleotide and the salen complex is supported by the observation that the higher molecular weight species is resistant to both extensive dialysis in the presence of EDTA and to the denaturing conditions of gel electrophoresis . The alkylation may be related to the ability of complex 2, upon oxidation with KHSO 5 , to form ligand radicals. Although the oxidation of nickel(II) tetraazamacrocyclic complexes such as complex 1 generally produces metal-centered oxidation yielding a nickel(III) species [16], the oxidation of Ni(salen) in weakly donating solvents leads to ligand oxidation [17-19] . From electrochemical experiments conducted on the unsubstituted Ni(salen), Goldsby has suggested that the initial oxidation of nickel(II) to nickel(III) is followed by a rapid intramolecular electron transfer yielding a ligand radical species [17]. In addition, carbon-carbon radical coupling involving the phenol portion of the salen ligand has been invoked to explain the electrochemical polymerization of Ni(salen) [18]. Based upon the electrochemical data of phenol-containing nickel complexes, it is proposed that the reactivity of complex 2 towards DNA is a result of the formation of phenoxy radicals upon oxidation and that these radicals act as alkylating (or arylating) agents of DNA . The known ability of phenoxy radicals such as 6-hydroxybenzo[a]pyrene radicals [20] and tyrosine radicals [21] to covalently bind to DNA supports this postulate .

1 While the peracids KHSO 5 and magnesium monoperoxyphthalate (MMPP) were found to be effective as oxidants, neither K 2S208 nor H 20 2 displayed reactivity towards the modification of DNA.



204

J. G. Muller et al.

complex 1 complex 2 complex 3 + KHSO5 isolated adduct isolated oligo piperidine 1

+

+

+

+

+ + +

+

+

+

+

+

+ +

+ +

2

3

4

5

6

+

G

7 8

9

5 1* T-- A I I C-- G I I A-- T I I G-- C I I A-- T I I T --A T 1

T C A

T I

G~ G ~G

FIGURE 1. Autoradiogram of denaturing polyacrylamide gel (20%) used to identify the modification of hairpin DNA by complexes 1-3 . The 17-base oligonucleotide was labeled at the 5' terminus with 32 P. DNA experiments were conducted as previously described [8] . Reaction mixtures (100 µL) contained final concentrations of 3 fiM unlabeled oligonucleotide, 2 nCi labeled oligonucleotide, 3 AM for the desired nickel(II) complex, 10 µM KHSO S (complexes 2 and 3) or 60 µM KHSO 5 (complex 1), 100 mM NaCl, and 10 mM potassium phosphate (pH 7.0). Samples were then dialyzed and purified from gels equivalent to those employed in this analysis (20% polyacrylamide, 7 M urea) . Those treated with alkaline conditions were also subjected to 0 .2 M piperidine for 30 min at 90°C. Lane 1 : control lane with base treatment . Lane 2 : complex 2 . Lane 3 : isolated higher molecular species from the reaction with complex 2 . Lane 4: same as lane 3 with base treatment . Lane 5 : isolated lower molecular species from the reaction with complex 2. Lane 6 : same as lane 5 with base treatment . Lane 7 : complex 1 with base treatment . Lane 8 : complex 3 with base treatment . Lane 9: Maxam-Gilbert G lane [23] .

In order to further evaluate the importance of the phenol moiety in the observed DNA modification, the reactivity of a nickel(II) complex containing a tetradentate N 2 0 2 ligand but lacking the phenol groups (3) was examined . Treatment of the hairpin oligonucleotide with complex 3 and KHSO 5 resulted in no detectable conversion of the oligonucleotide to a higher molecular weight

DNA MODIFICATION PROMOTED BY NICKEL COMPLEXES

205

species, and no strand scission products were observed after the treatment with piperidine (Fig . 1, lane 8), Therefore, the difference in the reactivity between complexes 2 and 3 can be attributed to the formation of a phenol-centered reactive intermediate, likely a phenoxy radical . Even though alkylation of DNA at guanine is common for both radical [22] and electrophilic [23] alkylating agents these reagents generally modify both base-paired and unpaired guanine residues. The observed specificity for accessible guanine residues has been demonstrated for complex 1 (Fig . 1, lane 7) and related nickel(II) complexes containing tetraazamacrocyclic ligands [7-9]. For 1, the specificity is believed to result from binding of the metal complex to sterically unhindered G residues, possibly through the N7 of guanine, and delivery of an oxidant to guanine . The similar specificity of complex 2 may be due to a similar steric accessibility of the metal complex while acting as an alkylating agent for guanine N7 or C8 . Alkylation at these sites is known to create base-labile lesions [23, 24] . The site of alkylation in the nonbase labile component of the reaction has not been determined. While G alkylations occurring at N2 and 06 are known to be resistant to base treatment [25, 26], the possible modification at other nucleobases cannot be ruled out. Further identification of the site(s) of modification responsible for the formation of the base resistant higher molecular weight product and the structures of these adducts is currently under investigation . Interestingly, even though Cu(salen) complexes can undergo electrochemical polymerization reactions analogous to Ni(salen) [19], the copper(II) analog of complex 2 was inactive towards DNA modification . This inactivity may be due to the fact that Cu(II) and Cu(III) complexes generally exist as 4-coordinate, square planar complexes and have lesser tendencies to accept additional axial ligands [27]. This further suggests that a key interaction takes place between the metal center and DNA . These experiments indicate that the mode of reaction of nickel complexes with DNA under oxidative conditions can be dramatically altered from G-oxidation to alkylation by use of a ligand whose nicke](III) complex undergoes rapid ligand-to-metal electron transfer. Such covalent attachment of a metal complex to DNA or RNA provides another entry into the design of DNA-targeted pharmaceuticals and new probes of nucleic acid structure . Support of this work through a grant from the National Science Foundation (to CJB), a grant from the American Cancer Society (to SER and CJB), and a Postdoctoral Fellowship (to JGM) from the American Cancer Society is gratefully acknowledged .

REFERENCES 1 . J. Stubbe and J. W. Kozarich, Chem . Rev . 87, 1107 (1987) . 2. T. D. Tullius, Ed., in Metal-DNA Chemistry ; ACS Symposium Series 402 ; American Chemical Society, Washington, DC, 1989. 3 . A. M. Pyle and J . K. Barton, Progress in Inorganic Chemistry : Bioinorganic Chemistry, S. J. Lippard, Ed., Wiley: New York, 1990, Vol. 38, pp. 413-474. 4. (a) C. B. Klein, K. Frenkel, and M . Costa, Chem . Res . Toxicol . 4, 592 (1991) ; (b) K. S. Kasprzak, Chem . Res. Toxicol. 4, 604 (1991) . 5 . A. K. Datta, M . Misra, S . L. North, and K. S . Kasprzak, Carcinogenesis 13, 283 (1992) .

206

J. G. Muller et al.

6. X. Chen, S . E. Rokita, and C . J . Burrows, J. Am . Chem . Soc . 113, 5884 (1991) . 7. X. Chen, C . J . Burrows, and S . E . Rokita, J. Am . Chem . Soc . 114, 322 (1992) . 8 . J. G . Muller, X . Chen, A . C. Dadiz, S. E. Rokita, and C . J . Burrows, J . Am . Chem . Soc . 114, 6407 (1992) . 9 . J. G. Muller, X . Chen, A . C. Dadiz, S . E . Rokita, and C. J . Burrows, Pure Appl. Chem . 65, 545 (1993) . 10. J. R . Morrow and K . A. Kolasa, Inorg . Chim . Acta 195, 245 (1992) . 11 . D . J . Gravert and J . H . Griffin, J. Org . Chem . 58, 820 (1993) . 12 . H . Yoon, T. R. Wagler, K . J . O'Connor, and C . J . Burrows, J. Am . Chem . Soc. 112, 4568 (1990) . 13 . A . K. Mukherjee and P . Ray, J. Indian Chem . Soc. 32, 632 (1955) . 14 . F. Vogtle and E . Goldschmitt, Chem . Ber . 109, 1 (1976) . 15 . J. L. Karn and D . H . Bush, Inorg. Chem . 8, 1149 (1969) . 16 . F. V. Lovecchio, E . S . Gore, and D . H . Busch, J . Am . Chem . Soc . 96, 3109 (1974) . 17 . K. A . Goldsby, J . Coord . Chem . 19, 83 (1988) . 18 . K. A . Goldsby, J . K. Blaho, and L . A . Hoferkamp, Polyhedron 8, 113 (1989) . 19 . L. A. Hoferkamp and K. A . Goldshy, Chem . Mater . 1, 348 (1989) . 20 . R . D . Irons and T . Sawahata, in Bioactiration of Foreign Compounds, M. W. Anders, Ed., Academic Press, Inc ., Florida, 1985, pp . 259-281 . 21 . E . Gajewski and M . Dizdaroglu, Biochemistry 29, 977 (1990) . 22 . (a) L . C. C. Leite and O . Augusto, Arch . Biochem . Biophys . 270, 560 (1989) ; (b) O . Augusto, E . L. Cavalier, E . G . Rogan, N . V. S . RamaKrishna, and C . Kolar, J . Biol . Chem . 265, 22093 (1990) . 23 . A . M . Maxam and W . Gilbert, Methods Enzymol . 65, 499 (1980) . 24 . M . Bichara and R . P. P. Fuchs, J . Mol . Biol . 183, 341 (1985). 25 . E . Sage and W . A. Haseltine, J. Biol . Chem . 259, 11098 (1984) . 26 . M.-L. Chung, H . Kiyosawa, E . Ohtsuka, S . Nishimura, and H . Kasai, Biochem . Biophys . Res. Commun . 188, 1 (1992) . 27 . N . N. Greenwood and A . Earnshaw, Chemistry of the Elements, Pergamon Press, New York, 1984, p . 1379 . Received April 22, 1993; accepted June 24, 1993