- Email: [email protected]
Classification of DNA-binding mode of antitumor and antiviral agents by the electrochemiluminescence of ruthenium complex
ANALYTICAL BIOCHEMISTRY Analytical Biochemistry 314 (2003) 30–37 www.elsevier.com/locate/yabio
Classification of DNA-binding mode of antitumor and antiviral agents by the electrochemiluminescence of ruthenium complex Tetsuo Kuwabara,a,* Tomohide Noda,a Hideki Ohtake,b,c Toshihito Ohtake,b,d Shigeru Toyama,b and Yoshihito Ikariyamab a
c
Department of Applied Chemistry and Biotechnology, Faculty of Engineering, University of Yamanashi, 4-3-11 Takeda, Kofu 400-8511, Japan b Research Institute, National Rehabilitation Center for the Disabled 4-1 Namiki, Tokorozawa 359-8555, Japan Department of Functional Materials Science, Faculty of Engineering, Saitama University, 255 Shimoohkubo, Saitama-shi, Saitama 338-8570, Japan d Japan Science and Technology Corporation, 4-1-8 Honmachi, Kawaguchi, Saitama 332-0012, Japan Received 26 June 2002
Abstract The DNA-binding mode of antitumor and antiviral agents has been evaluated by electrochemiluminescence (ECL) of tris(1,10phenanthroline)-ruthenium complex ðRuðphenÞ2þ 3 Þ in the presence of oxalate ion in pH 7.3 Tris buffer solution. An emission of RuðphenÞ2þ was observed repeatedly with a voltage above 1000 mV subjected to a potential sweep from 0 to 1250 mV. The addition 3 of kDNA into the solution containing 1 lM of RuðphenÞ2þ 3 caused the decrease in the ECL intensity, which became half at a DNA concentration of 20 lM. This is due to the binding of D-type of RuðphenÞ2þ 3 with DNA in the major groove of DNA. When the various concentrations of the drug were added to the solution containing 1 lM RuðphenÞ2þ 3 , the ECL intensity was not affected by the concentration of the drug in the absence of DNA. In the presence of DNA (10 lM), however, two ECL emission patterns were observed when the concentration of the drug was varied. The pattern that the ECL intensity increased with increasing the drug concentration was observed for cisplatin, daunomycin, and DC92-B. This may have resulted from the DNA binding of the drug 2þ with a major groove site, where RuðphenÞ2þ 3 should bind. RuðphenÞ3 nonbinding to DNA might exist in the bulk solution and exhibits ECL emission. The drug exhibiting the drug-concentration-dependent ECL is classified as a drug with a major groove binding character. The addition of drugs, such as mitomycin C and duocarmycin SA, did not cause a change in the ECL intensity even in the presence of DNA. This result indicates that these drugs bind to DNA with minor groove binding. Since similar trends were observed for actinomycin D, distamycin A, doxorubicin, and chromomycin A3; these drugs are also considered as minor groove binding agents. All these results demonstrate that the DNA-binding mode of the drug can be evaluated easily by utilizing the ECL of RuðphenÞ2þ 3 , which is used as the sensing probe. Ó 2003 Elsevier Science (USA). All rights reserved. Keywords: Electrochemiluminescence; Ruthenium Complex; Drug; Antitumor; DNA
Much effort has been devoted to the development of a variety of drugs as antitumor and antiviral agents, which can act as inhibitors for DNA replication and DNA transcription. Such inhibitive reactions proceed after binding of the drug to DNA. It is important, therefore, for understanding the reaction mechanism of * Corresponding author. Fax: +81-55-220-8548. E-mail address: [email protected] (T. Kuwabara).
the drugs as well as for designing novel drugs to clarify the binding mode of these drugs to DNA. When the drug binds to DNA, there are mainly three binding modes: (i) major groove binding, (ii) minor groove binding, and (iii) intercalation binding between base pairs. Although there are some reports on the structure of the DNA-drug adducts, which have been investigated by spectroscopic techniques such as NMR and Xray crystallography, it is important to develop a simple
0003-2697/03/$ - see front matter Ó 2003 Elsevier Science (USA). All rights reserved. doi:10.1016/S0003-2697(02)00651-6
T. Kuwabara et al. / Analytical Biochemistry 314 (2003) 30–37
31
is an [4]. Another interesting aspect of RuðphenÞ2þ 3 ability to generate electrochemiluminescence (ECL) in the presence of oxalate in aqueous solution [5]. The emission mechanism of RuðphenÞ2þ 3 is as follows. 3þ RuðphenÞ2þ 3 ! RuðphenÞ3 þ e
ð1Þ
3þ
Fig. 1. Chemical structures of tris(1,10-phenanthroline)-ruthenium complex ðRuðphenÞ2þ 3 Þ and its D-type and K-type forms.
and convenient method for evaluating the DNA-binding mode of the drug [1]. On the other hand, there are many reports on organic metal complexes that are capable of binding to DNA [2]. The organic metal complexes have some advantages, such as easy preparation and variety of design possibilities with different metal ions and ligands, for probing DNA structure and investigating the binding process as well as for facilitating individual applications. Among them, ruthenium complex is one of the most extensively investigated members of a class of the DNA-binding organic metal complexes. Tris(1,10-phenanthroline)-ruthenium complex, RuðphenÞ2þ 3 , whose structure is shown in Fig. 1, is one of the complexes that binds to DNA with the intercalation mode. Barton et al. reported that the complex bound to DNA stays in the major groove of DNA, in which one of the ligands inserts between the base pairs of the double helix [3], although the binding property of RuðphenÞ2þ 3 to DNA is under investigation
2þ
RuðphenÞ3 þ C2 O2 4 ! RuðphenÞ3 þ C2 O4
ð2Þ
C2 O 4 ! CO2 þ CO2
ð3Þ
þ RuðphenÞ2þ 3 þ CO2 ! RuðphenÞ3 þ CO2
ð4Þ
3þ
2þ
RuðphenÞ3 þ CO 2 ! RuðphenÞ3
þ CO2
ð5-1Þ
þ RuðphenÞ3þ 3 þ RuðphenÞ3 2þ
! RuðphenÞ3
2þ
þ RuðphenÞ3
RuðphenÞ2þ ! RuðphenÞ2þ 3 3 þ hm
ð5-2Þ ð6Þ
The reaction of the electrogenerated RuðphenÞ3þ 3 with the reductant produced from oxalate ion and/or þ with RuðphenÞ3 produced by chemical reduction of 2þ RuðphenÞ3 causes the ECL emission [6]. In this report, by utilizing RuðphenÞ2þ 3 as a probe, which is simultaneously capable of binding to the major groove in DNA and of generating ECL emission, the DNA-binding mode of antitumor and antiviral agents has been investigated. The principle of this system is shown in Fig. 2. RuðphenÞ2þ emits ECL through 3
2þ Fig. 2. Principle of ECL-based typing of anticancer agents. (A) ECL of RuðphenÞ2þ 3 : the ECL intensity of RuðphenÞ3 alone was used as the standard 2þ intensity. (B) ECL of RuðphenÞ3 in the presence of DNA: the luminescence decreases by steric hindrance of DNA. (C) ECL of RuðphenÞ2þ 3 in the presence of DNA and the major groove binding drug: ECL increases with increasing the drug concentration. (D) ECL of RuðphenÞ2þ 3 in the presence of DNA and the minor groove binding drug: ECL is not affected by the drug concentration. Thus the ECL intensity depends on the DNA-binding mode of each agent.
32
T. Kuwabara et al. / Analytical Biochemistry 314 (2003) 30–37
electrochemical oxidation as shown in Fig. 2A. The addition of DNA into the solution would cause a decrease in the intensity of ECL (Fig. 2B) [a,b,5]. When RuðphenÞ2þ is added to the solution containing drug 3 and DNA, however, the change in the ECL intensity should reflect the DNA-binding mode of the drug; the ECL intensity should depend on the drug concentration when the drug binds to the major groove in DNA (Fig. 2C), while the intensity should not change when the drug binds to the minor groove. This is due to the major binding character of RuðphenÞ2þ 3 in DNA (Fig. 2D) [3]. On this basis, the DNA-binding mode of antitumor and antiviral agents can be evaluated by utilizing ECL of RuðphenÞ2þ 3 as a sensing probe. Here we report the results obtained for a various kinds of antitumor and antiviral agents.
Experimental Materials RuðphenÞ2þ 3 and kDNA were obtained from Aldrich (USA) and Takara Shuzo Co., Ltd. (Japan), respectively. The drugs used in this study are as follows. Cisplatin, transplatin, actinomycin D, distamycin A, daunomycin, and chromomycin A3 were purchased from Sigma (USA). Mitomycin C and doxorubicin were obtained from Wako (Japan). Duocarmycin SA and DC92-B were a generous gift from Kyowa Hakko Kogyo Co., Ltd. (Japan). Measurements The experimental setup consisted of an electrochemical part and a photon-counting part. The ECL of RuðphenÞ2þ was generated by the potential sweep by 3 using a potentiostat C2090 and a function generator FG-02 (Toho Technical Research, Japan) with a threeelectrode system in a dark box. Platinum plates were used as the counter and the working electrodes, and an Ag-AgCl electrode was used as the reference electrode. The photon counting was performed by a photomultiplier tube R464 and a photon counter C1230 (Hamamatsu Photonics, Japan). A 50 mM Tris buffer at pH 7.3 was used as an electrolyte solution, which involves 15 mM oxalate ion and 50 mM sodium chloride. RuðphenÞ2þ 3 and kDNA were used as a probe molecule and a model of double-helical DNA, respectively. The solution was prepared as follows. In order to prepare the solution containing various concentrations of the drug from 0.01 to 100 lM, drug was added to the buffer solution containing kDNA, and then RuðphenÞ2þ 3 was added. The concentrations of RuðphenÞ2þ 3 and kDNA were fixed at 1 and 10 lM, respectively. After incubation at 40 °C for 8 h,
the ECL of RuðphenÞ2þ 3 was monitored by the photoncounting system. The potential sweep was performed from 0 to 1250 mV with a speed of 100 mV/s. The control experiment was performed under similar conditions in the absence of DNA. The ECL of RuðphenÞ2þ 3 alone was used as the standard ECL intensity.
Results and discussion ECL of Ru complex in the absence and the presence of DNA When the solution containing RuðphenÞ2þ 3 was subjected to a potential sweep from 0 to 1250 mV, the ECL emission was observed. RuðphenÞ2þ emitted above 3 1000 mV sweep voltage with the maximum intensity at 1250 mV repeatedly [7]. The stable ECL was obtained after the third sweep. The mechanism of ECL of RuðphenÞ2þ 3 has been reported by Carter and Bard [5]. They reported that ECL involves the reaction of the electrogenerated RuðphenÞ3þ 3 with the strong reductant produced as intermediate in the oxidation of oxalate ion [Eq. (5-1)] and with chemically produced RuðphenÞþ 3 species [Eq. (5-2)]. Fig. 3 shows the ECL intensity at a given concentration of RuðphenÞ2þ 3 . The increase in the concentration 2þ of RuðphenÞ3 caused an increase in the intensity of ECL. However, no change in the ECL intensity was observed above ca. 6 lM RuðphenÞ2þ 3 , suggesting the existence of a rate-determining step of the electron transfer reaction between RuðphenÞ2þ 3 and the surface of the working electrode. Upon the addition of kDNA into the solution containing 1 lM RuðphenÞ2þ 3 , the ECL intensity decreased, as shown in Fig. 4. This suggests that the addition of DNA caused the decrease in the concentration of free RuðphenÞ2þ in the bulk solution, associated with the 3 binding of RuðphenÞ2þ with DNA [5]. RuðphenÞ2þ 3 3 bound in DNA emits no luminescence because of the
Fig. 3. Change in ECL intensity as a function of RuðphenÞ2þ con3 centration.
T. Kuwabara et al. / Analytical Biochemistry 314 (2003) 30–37
Fig. 4. Change in the relative ECL intensity of 1 lM of RuðphenÞ2þ 3 as a function of kDNA concentration.
steric hindrance of DNA. The ECL intensity became half at a concentration of ca. 20 lM DNA. Further addition of DNA leads to a negligible change in the ECL intensity. These results demonstrate that, even in the presence of excess DNA, a half amount of RuðphenÞ2þ 3 remains in the bulk solution to emit. This fact is well coincident with the result reported by Barton et al., in which D-type RuðphenÞ2þ 3 prefers to bind to DNA with the major groove binding mode [3]. The K-type RuðphenÞ2þ 3 having no binding ability to DNA remains in the bulk solution and exhibits the ECL emission. It is likely that there is one RuðphenÞ2þ 3 molecule in four helices of DNA under our conditions because the formation of one helix of DNA needs 10 bp. Binding mode of nonanthracycline-type drugs Fig. 5 shows the chemical structures of the drugs used in this study except for the anthracycline drugs.
33
Cisplatin and transplatin, both stereoisomers that differ in geometric arrangement of two attached ammonia molecules and chloride ions as the ligands, are known to bind to duplex DNA [8]. The structure of cisplatin-DNA adduct is of interest because cisplatin is an anticancer agent used to treat a variety of tumors, while trans-isomer is devoid of biological activity [9]. Takahara et al. [10] reported the X-ray crystal structure of platinated oligonucleotides, in which cisplatin binds to guanine residues in DNA. Fig. 6 shows the relative ECL intensity of RuðphenÞ2þ 3 as a function of cisplatin concentration when cisplatin was used as the drug. It was observed, in the absence of DNA, that the ECL intensity of RuðphenÞ2þ 3 even in the presence of excess cisplatin was similar to that of RuðphenÞ2þ 3 alone, which is used as the standard intensity. Thus the ECL intensity was not affected by the presence of cisplatin. This fact indicates that there is no interaction between RuðphenÞ2þ and cisplatin. In the presence of DNA, 3 however, the ECL intensity was affected by the presence of cisplatin. When the solution contained 0.01 lM cisplatin, the intensity was 50% as compared to the intensity of DNA-free solution or to the standard intensity. This suggests that almost all D-type RuðphenÞ2þ 3 binds to DNA under this condition. However, with increasing concentrations of cisplatin, the ECL intensity increased and almost recovered at 100 lM cisplatin. This result demonstrates that the concentration of free RuðphenÞ2þ 3 in the bulk solution increased with increasing concentration of cisplatin, namely, that the binding of RuðphenÞ2þ 3 to DNA is inhibited by cisplatin. At 100 lM cisplatin, almost no RuðphenÞ2þ 3 can bind to DNA, resulting in an ECL intensity similar to that of the DNAfree solution. Cisplatin may bind to the major groove of DNA, causing inhibition of RuðphenÞ2þ 3 binding to this
Fig. 5. Chemical structure of nonanthracycline drugs.
34
T. Kuwabara et al. / Analytical Biochemistry 314 (2003) 30–37
site. This is well coincident with the result obtained by X-ray measurement [10]. However, the dependency of the ECL intensity on the transplatin concentration was different from that on cisplatin. The intensity increased with increasing concentrations of transplatin even in the absence of DNA. The intensity became 3 times higher than the standard intensity of RuðphenÞ2þ when it 3 contained 100 lM transplatin. Since transplatin itself shows no ECL emission, the strong ECL observed resulted from an unexpected reason. Transplatin seems to promote the ECL emission of RuðphenÞ2þ 3 . Therefore the DNA-binding mode of transplatin as well as its binding ability could not be evaluated. Mitomycin C [11] and duocarmycin SA [12] have relatively simple structures and have an ability to alkylate the 2-amino group of guanines and N3 adenine in DNA, respectively. Fig. 7 shows the ECL intensity of RuðphenÞ2þ 3 as a function of the concentration of mitomycin C and duocarmycin SA in the absence and the presence of DNA. The ECL intensities of DNA-free solutions were similar to that of the standard solution
and were not affected by the drug concentrations, suggesting no interaction between these drugs and DNA. In the presence of DNA, their ECL patterns were different from those of cisplatin. The ECL intensities for both drugs were not influenced by the concentrations of mitomycin C and duocarmycin SA in the range from 0.01 to 100 lM of the drugs. No changes in the ECL intensities as a function of the drug concentration suggest that neither drug prevents RuðphenÞ2þ 3 from binding to DNA. It can be concluded, therefore, that these drugs bind to DNA with the minor groove, which is a different site than that of RuðphenÞ2þ 3 . The minor groove binding character of these drugs was coincident with the results reported previously [11,12]. The difference between these drugs was observed in the relative intensity of the ECL, whose intensities were observed ca. 70 and 77% for mitomycin C and duocarmycin SA, respectively. This indicates that the concentration of RuðphenÞ2þ 3 in the bulk solution is higher when the solution contained duocarmycin SA rather than mitomycin C. This may result from the following reasons. When the drug binds to the minor groove in DNA, (i) the bound drug may widen the groove width around the drug, resulting in a decrease of the major groove width and/or (ii) the extra moiety of the drug, which could not bind to the minor groove, may partly bind to the major groove, resulting in a decrease of the binding space of the complex in the major groove. This may occur remarkably in the case of duocarmycin SA rather than mitomycin C. Similar trends were observed in the case of actinomycin D [13] and distamycin A [14], which are the cyclic peptide and naturally occurring polyamine antibiotic, respectively (Fig. 8). Although a little decrease in the ECL intensity of RuðphenÞ2þ was observed with 3 increasing concentrations of actinomycin D in the absence of DNA, the ECL intensity in the presence of DNA was not significantly affected by the concentration of this agent. This result suggests that actinomycin D
Fig. 7. Effect of mitomycin C (square) and duocarmycin SA (circle) on the relative ECL intensity of RuðphenÞ2þ 3 in the absence (white) and the presence (black) of kDNA.
Fig. 8. Effect of actinomycin D (square) and distamycin A (circle) on the relative ECL intensity of RuðphenÞ2þ 3 in the absence (white) and the presence (black) of kDNA.
Fig. 6. Effect of cisplatin on the relative ECL intensity of RuðphenÞ2þ 3 in the absence (white square) and the presence (black square) of kDNA.
T. Kuwabara et al. / Analytical Biochemistry 314 (2003) 30–37
35
binds to DNA with the minor groove binding mode, coincident with the result obtained by X-ray or NMR analysis [13]. While for distamycin A, the ECL intensity was similar to that of the standard solution in the absence of DNA and was not affected by the drug concentration. In the presence of DNA, the 65% ECL intensity was observed at 0.01 lM concentration of distamycin A and the intensity did not change even in the presence of 100 lM drug. These results indicate that distamycin A binds to the minor groove of DNA, which is well coincident with the report by Lian et al. [13]. DNA-binding mode of anthracyclic drugs Fig. 9 shows the chemical structures of anthracyclic drugs. Among them, daunomycin [15] and doxorubicin [16] have a similar structure bearing a glucosamine unit with little difference in the structure in the substitution introduced at C9 position in ring A, the hydroxyl and ester groups for daunomycin and methyl and alcohol groups for doxorubicin, respectively. However, the remarkable difference in the ECL-emitting pattern of RuðphenÞ2þ 3 was observed between these drugs. Fig. 10 shows the effect of daunomycin and doxorubicin on the ECL intensitiy in the absence and the presence of DNA. In the absence of DNA, the ECL intensities were not affected by the concentration of both drugs. In the presence of DNA, however, the increase in the ECL intensity was observed with increasing concentrations of daunomycin, while no increase in the intensity was observed when doxorubicin concentration was increased. These results indicate that daunomycin and doxorubicin bind to DNA with the major and minor groove, respectively. It is interesting that such a small difference in their structures causes the change in the DNA-binding mode of the drug. It was
Fig. 10. Effect of daunomycin (square) and doxorubicin (circle) on the relative ECL intensity of RuðphenÞ2þ 3 in the absence (white) and the presence (black) of kDNA.
observed, furthermore, that the ECL intensity was almost recovered when the solution contained 1 lM daunomycin. Since, in the case of cisplatin, which also binds to the major groove of DNA, the ECL intensity was recovered at 100 lM, it seems that daunomycin binds to DNA with a stronger binding ability than cisplatin. As compared to the standard intensity, doxorubicin showed the relatively high ECL intensity among the minor groove binding drugs, indicating that doxorubicin binds not only to the minor groove but also to the major one. The major site seems to be occupied by a part of doxorubicin, although it stays mainly in the minor groove. Fig. 11 shows the results obtained when chromomycin A3 [17] and DC92-B [18] were used as the antitumor reagent. The ECL intensity of RuðphenÞ2þ was not 3 affected by the concentration of chromomycin A3 in the absence and the presence of DNA. This fact indicates that there are no interactions between RuðphenÞ2þ 3 and chromomycin A3 and that chromomycin A3 binds to
Fig. 9. Chemical structure of anthracycline drugs.
36
T. Kuwabara et al. / Analytical Biochemistry 314 (2003) 30–37
[3]
Fig. 11. Effect of chromomycin A3 (square) and DC92-B (circle) on the relative ECL intensity of RuðphenÞ2þ 3 in the absence (white) and the presence (black) of kDNA [4]
DNA with the minor groove binding mode. On the other hand, in order to evaluate the binding mode of DC92-B it is necessary to correct the result obtained because the ECL intensity is affected by DC92-B in the absence of DNA. The corrected result shows that ECL intensity increased with increasing concentrations of the drug, indicating that DC92-B binds to DNA with the major groove. These results obtained for chromomycin A3 and DC92-B are coincident with the results reported [17,18]. All these results demonstrate that the DNA-binding mode of the antitumor and antiviral agents can be evaluated by utilizing the DNA-binding and the ECLgenerating abilities of RuðphenÞ2þ 3 .
[5]
[6]
[7]
Acknowledgments We are grateful to Kyowa Hakko Kogyo Co., Ltd., for providing us with duocarmycin SA and DC92-B. We thank Dr. Akira Asai and Dr. Hirofumi Nakano in Kyowa Hakko Kogyo Co., Ltd., for valuable discussions. We also thank Professor Masato Nanasawa in Yamanashi University and Professor Takeaki Iida in Saitama University for kind discussions. This work was supported by Kato Memorial Bioscience Fund. Finally we hope that one of the authors, Y.I., who died in September 2000 on the way to achieving this work, will enjoy reading this paper in heaven.
[8]
[9]
[10]
[11]
References [12] [1] Y. Ikariyama, Y. Taguchi, T. Suzawa, E. Kobatake, M. Aizawa, Electrochemiluminescent sensing for the characterization of DNA-interacting antitumor and antiviral agents, Sensor Acta B 13–14 (1993) 725–727. [2] (a) R.H. Terbrueggen, T.W. Johann, J.K. Barton, Functionalized rhodium intercalators for DNA recognition, Inorg. Chem. 37 (1998) 6874–6883;
(b) J.K. Barton, A.L. Raphael, Site-specific cleavage of lefthanded DNA in pBR322 by K-tris(diphenylphenanthroline)cobalt(III), Proc. Natl. Acad. Sci. USA 82 (1985) 6460; (c) J.K. Barton, J.J. Dannenberg, A.L. Raphael, Enantiometric selectivity in binding tris(phenanthroline)zinc(II) to DNA, J. Am. Chem. Soc. 104 (1982) 4967–4969. (a) R.E. Holmlin, E.D.A. Stemp, J.K. Barton, RuðphenÞ2 dppz2þ luminescence: dependence on DNA sequences and groove-binding, Inorg. Chem. 37 (1998) 29–34; (b) C.M. Dupureur, J.K. Barton, Structural studies of K- and D½RuðphenÞ2 dppz2þ Bound to dðGTCGACÞ2 : characterization of enantioselective intercalation, Inorg. Chem. 36 (1997) 33–43; (c) C.M. Dupureur, J.K. Barton, Use of selective deutation and 1 H NMR in demonstrating major groove binding of D-½RuðphenÞ2 dppz2þ to dðGTCGACÞ2 , J. Am. Chem. Soc. 116 (1994) 10286–10287; (d) J.K. Barton, A.T. Danishefsky, J.M. Goldberg, Tris(phenanthroline)ruthenium(II): stereoselectivity in binding to DNA, J. Am. Chem. Soc. 106 (1984) 2172–2176. (a) E. Tuite, P. Lincoln, B. Norden, Photophysical evidence that D- and K-½RuðphenÞ2 ðdppzÞ2þ intercalate DNA from the minor groove, J. Am. Chem. Soc. 119 (1997) 239240; (b) I. Haq, P. Lincoln, D. Suh, B. Norden, B.Z. Chowdhry, J.B. Chires, Interaction of D- and K-½RuðphenÞ2 DPPZ2þ with DNA: a calorimetric and equilibrium binding study, J. Am. Chem. Soc. 117 (1995) 4788–4796; (c) C. Hiort, P. Lincoln, B. Norden, DNA binding of K- and D-½RuðphenÞ2 DPPZ2þ , J. Am. Chem. Soc. 115 (1993) 3448–3454. M.T. Carter, A.J. Bard, Electrochemical investigations of the metal chelates with DNA. 3. Electrogenerated chemiluminscent investigation of the inteaction of tris(1,10-phenanthroline)ruthenium(II) with DNA, Bioconj. Chem. 1 (1990) 257–263. For ECL of metal complexes of this type, see (a) I. Rubinstein, A.J. Bard, Electrogenerated chemiluminscence. 37. Aqueous Ecl systems based on Ruð2; 20 -bipyridineÞ2þ and oxalate or organic 3 acids, J. Am. Chem. Soc. 103 (1981) 512–516. When a constant potential was applied to the solution containing ruthenium complex, strong ECL was observed immediately after the potential application. However, it was rapidly attenuated to a very weak level due to the rapid oxidation of Ru complex at the electrode surface, while other reactions cannot catch up with the electrode reaction. E.E. Trimmer, J.M. Essigmann, in: D.P. Ballou (Ed.), Essays in Biochemistry, vol. 34, Princeton University Press, New Jersey, 1999, pp. 191–211. E. Bernal-Mendez, M. Boudvillain, F. Gonzalei-Vilchez, M. Leng, Chemical versatility of transplatin monofunctional adducts within multiple site-specifically platinated DNA, Biochemistry 36 (1997) 7281–7287. P.M. Takahara, C.A. Frederick, S.J. Lippard, Crystal structure of the anticancer drug cisplatine bound to duplex DNA, J. Am. Chem. Soc. 118 (1996) 12309–12321. (a) G.S. Kumar, S.M. Musser, J. Cumming, M. Tomasz, 2,7Diaminomitosene, a monofunctional mitomycin C derivative, alkylates DNA in the major groove. Structure and base-sequence specificity of the DNA adduct and mechanism of the alkylation, J. Am. Chem. Soc. 118 (1996) 9209–9217; (b) M. Tomasz, R. Lipmann, Alkylation reactions of mitomycin C at acid pH, J. Am. Chem. Soc. 101 (1979) 6063–6067. (a) A. Asai, S. Nagamura, H. Saito, A novel property of duocarmycin and its analogues for covalent reaction with DNA, J. Am. Chem. Soc. 116 (1994) 4171–4177; (b) A. Asai, K. Yano, T. Mizukami, H. Nakano, Characterization of duocarmycin-DNA adduct-recognizing protein in cancer cells, Cancer Res. 59 (1999) 5417–5420; (c) D.L. Boger, D.S. Johnson, CC-1065 and the duocarmycins: unraveling the keys to a new class naturally derived DNA
T. Kuwabara et al. / Analytical Biochemistry 314 (2003) 30–37 alkylating agent, Proc. Natl. Acad. Sci. USA 92 (1995) 3642– 3649. [13] (a) C. Lian, H. Robinson, A.H.-J. Wang, Structure of actinomycin D bound with ðGAAGCTTCÞ2 and ðGATGCTTCÞ2 and its binding to the ðCAGÞn :ðCTGÞn triplet sequence as determined by NMR analysis, J. Am. Chem. Soc. 118 (1996) 8791– 8801; (b) H. Robinson, Y.-G. Gao, X.-l. Yang, R. Sanishvili, A. Joachimiak, A.H.J. Wang, Crystallographic analysis of a novel complex of actinomycin D bound to the DNA decamer CGATCGATCG, Biochemistry 40 (2001) 5587–5592. [14] (a) R. Baliga, D.M. Crothers, The kinetic basis for sequence discrimination by distamycin A, J. Am. Chem. Soc. 122 (2000) 11751–11752; (b) J. Lah, G. Vesnaver, Binding of distamycin A and netropsin to the 12mer DNA duplexes containing mixed AT-GC sequences with at most five or three successive AT base pairs, Biochemistry 39 (2000) 9317–9326.
37
[15] C.A. Frederick, L.D. Williams, G. Ughetto, G.A. van der Marel, J.H. van Boom, A. Rich, A.H.-J. Wang, Structural comparison of anticancer drug-DNA complexes: adriamycin and daunomycin, Biochemistry 29 (1990) 2538–2549. [16] G.E. Kellogg, J.N. Scarsdale, F.A. Fornari Jr., Identification and hydropathic characterization of structural features affecting sequence specificity for doxorubicin intercalation into DNA doublestranded polynucleotides, Nucleic Acids Res. 20 (1998) 4721–4732. [17] (a) X. Gao, D.J. Patel, Chromomycin dimer-DNA oligomer complexes. Sequence selectivity and divalent cation specificity, Biochem. 28 (1990) 10940–10956; (b) X. Gao, P. Miura, D.J. Patel, Structure refinement of the chromomycin dimer-DNA oligomer complex in solution, Mol. Biol. 223 (1992) 259–279. [18] D. Sun, J. Hansen, J.J. Clement, L.H. Hurley, Structure of altromycin B(N7-Guanine)-DNA adduct. A proposed prototypic DNA adduct structure for the pluramycin antitumor antibiotics, Biochemistry 32 (1993) 8068–8074.
Classification of DNA-binding mode of antitumor and antiviral agents by the electrochemiluminescence of ruthenium complex Tetsuo Kuwabara,a,* Tomohide Noda,a Hideki Ohtake,b,c Toshihito Ohtake,b,d Shigeru Toyama,b and Yoshihito Ikariyamab a
c
Department of Applied Chemistry and Biotechnology, Faculty of Engineering, University of Yamanashi, 4-3-11 Takeda, Kofu 400-8511, Japan b Research Institute, National Rehabilitation Center for the Disabled 4-1 Namiki, Tokorozawa 359-8555, Japan Department of Functional Materials Science, Faculty of Engineering, Saitama University, 255 Shimoohkubo, Saitama-shi, Saitama 338-8570, Japan d Japan Science and Technology Corporation, 4-1-8 Honmachi, Kawaguchi, Saitama 332-0012, Japan Received 26 June 2002
Abstract The DNA-binding mode of antitumor and antiviral agents has been evaluated by electrochemiluminescence (ECL) of tris(1,10phenanthroline)-ruthenium complex ðRuðphenÞ2þ 3 Þ in the presence of oxalate ion in pH 7.3 Tris buffer solution. An emission of RuðphenÞ2þ was observed repeatedly with a voltage above 1000 mV subjected to a potential sweep from 0 to 1250 mV. The addition 3 of kDNA into the solution containing 1 lM of RuðphenÞ2þ 3 caused the decrease in the ECL intensity, which became half at a DNA concentration of 20 lM. This is due to the binding of D-type of RuðphenÞ2þ 3 with DNA in the major groove of DNA. When the various concentrations of the drug were added to the solution containing 1 lM RuðphenÞ2þ 3 , the ECL intensity was not affected by the concentration of the drug in the absence of DNA. In the presence of DNA (10 lM), however, two ECL emission patterns were observed when the concentration of the drug was varied. The pattern that the ECL intensity increased with increasing the drug concentration was observed for cisplatin, daunomycin, and DC92-B. This may have resulted from the DNA binding of the drug 2þ with a major groove site, where RuðphenÞ2þ 3 should bind. RuðphenÞ3 nonbinding to DNA might exist in the bulk solution and exhibits ECL emission. The drug exhibiting the drug-concentration-dependent ECL is classified as a drug with a major groove binding character. The addition of drugs, such as mitomycin C and duocarmycin SA, did not cause a change in the ECL intensity even in the presence of DNA. This result indicates that these drugs bind to DNA with minor groove binding. Since similar trends were observed for actinomycin D, distamycin A, doxorubicin, and chromomycin A3; these drugs are also considered as minor groove binding agents. All these results demonstrate that the DNA-binding mode of the drug can be evaluated easily by utilizing the ECL of RuðphenÞ2þ 3 , which is used as the sensing probe. Ó 2003 Elsevier Science (USA). All rights reserved. Keywords: Electrochemiluminescence; Ruthenium Complex; Drug; Antitumor; DNA
Much effort has been devoted to the development of a variety of drugs as antitumor and antiviral agents, which can act as inhibitors for DNA replication and DNA transcription. Such inhibitive reactions proceed after binding of the drug to DNA. It is important, therefore, for understanding the reaction mechanism of * Corresponding author. Fax: +81-55-220-8548. E-mail address: [email protected] (T. Kuwabara).
the drugs as well as for designing novel drugs to clarify the binding mode of these drugs to DNA. When the drug binds to DNA, there are mainly three binding modes: (i) major groove binding, (ii) minor groove binding, and (iii) intercalation binding between base pairs. Although there are some reports on the structure of the DNA-drug adducts, which have been investigated by spectroscopic techniques such as NMR and Xray crystallography, it is important to develop a simple
0003-2697/03/$ - see front matter Ó 2003 Elsevier Science (USA). All rights reserved. doi:10.1016/S0003-2697(02)00651-6
T. Kuwabara et al. / Analytical Biochemistry 314 (2003) 30–37
31
is an [4]. Another interesting aspect of RuðphenÞ2þ 3 ability to generate electrochemiluminescence (ECL) in the presence of oxalate in aqueous solution [5]. The emission mechanism of RuðphenÞ2þ 3 is as follows. 3þ RuðphenÞ2þ 3 ! RuðphenÞ3 þ e
ð1Þ
3þ
Fig. 1. Chemical structures of tris(1,10-phenanthroline)-ruthenium complex ðRuðphenÞ2þ 3 Þ and its D-type and K-type forms.
and convenient method for evaluating the DNA-binding mode of the drug [1]. On the other hand, there are many reports on organic metal complexes that are capable of binding to DNA [2]. The organic metal complexes have some advantages, such as easy preparation and variety of design possibilities with different metal ions and ligands, for probing DNA structure and investigating the binding process as well as for facilitating individual applications. Among them, ruthenium complex is one of the most extensively investigated members of a class of the DNA-binding organic metal complexes. Tris(1,10-phenanthroline)-ruthenium complex, RuðphenÞ2þ 3 , whose structure is shown in Fig. 1, is one of the complexes that binds to DNA with the intercalation mode. Barton et al. reported that the complex bound to DNA stays in the major groove of DNA, in which one of the ligands inserts between the base pairs of the double helix [3], although the binding property of RuðphenÞ2þ 3 to DNA is under investigation
2þ
RuðphenÞ3 þ C2 O2 4 ! RuðphenÞ3 þ C2 O4
ð2Þ
C2 O 4 ! CO2 þ CO2
ð3Þ
þ RuðphenÞ2þ 3 þ CO2 ! RuðphenÞ3 þ CO2
ð4Þ
3þ
2þ
RuðphenÞ3 þ CO 2 ! RuðphenÞ3
þ CO2
ð5-1Þ
þ RuðphenÞ3þ 3 þ RuðphenÞ3 2þ
! RuðphenÞ3
2þ
þ RuðphenÞ3
RuðphenÞ2þ ! RuðphenÞ2þ 3 3 þ hm
ð5-2Þ ð6Þ
The reaction of the electrogenerated RuðphenÞ3þ 3 with the reductant produced from oxalate ion and/or þ with RuðphenÞ3 produced by chemical reduction of 2þ RuðphenÞ3 causes the ECL emission [6]. In this report, by utilizing RuðphenÞ2þ 3 as a probe, which is simultaneously capable of binding to the major groove in DNA and of generating ECL emission, the DNA-binding mode of antitumor and antiviral agents has been investigated. The principle of this system is shown in Fig. 2. RuðphenÞ2þ emits ECL through 3
2þ Fig. 2. Principle of ECL-based typing of anticancer agents. (A) ECL of RuðphenÞ2þ 3 : the ECL intensity of RuðphenÞ3 alone was used as the standard 2þ intensity. (B) ECL of RuðphenÞ3 in the presence of DNA: the luminescence decreases by steric hindrance of DNA. (C) ECL of RuðphenÞ2þ 3 in the presence of DNA and the major groove binding drug: ECL increases with increasing the drug concentration. (D) ECL of RuðphenÞ2þ 3 in the presence of DNA and the minor groove binding drug: ECL is not affected by the drug concentration. Thus the ECL intensity depends on the DNA-binding mode of each agent.
32
T. Kuwabara et al. / Analytical Biochemistry 314 (2003) 30–37
electrochemical oxidation as shown in Fig. 2A. The addition of DNA into the solution would cause a decrease in the intensity of ECL (Fig. 2B) [a,b,5]. When RuðphenÞ2þ is added to the solution containing drug 3 and DNA, however, the change in the ECL intensity should reflect the DNA-binding mode of the drug; the ECL intensity should depend on the drug concentration when the drug binds to the major groove in DNA (Fig. 2C), while the intensity should not change when the drug binds to the minor groove. This is due to the major binding character of RuðphenÞ2þ 3 in DNA (Fig. 2D) [3]. On this basis, the DNA-binding mode of antitumor and antiviral agents can be evaluated by utilizing ECL of RuðphenÞ2þ 3 as a sensing probe. Here we report the results obtained for a various kinds of antitumor and antiviral agents.
Experimental Materials RuðphenÞ2þ 3 and kDNA were obtained from Aldrich (USA) and Takara Shuzo Co., Ltd. (Japan), respectively. The drugs used in this study are as follows. Cisplatin, transplatin, actinomycin D, distamycin A, daunomycin, and chromomycin A3 were purchased from Sigma (USA). Mitomycin C and doxorubicin were obtained from Wako (Japan). Duocarmycin SA and DC92-B were a generous gift from Kyowa Hakko Kogyo Co., Ltd. (Japan). Measurements The experimental setup consisted of an electrochemical part and a photon-counting part. The ECL of RuðphenÞ2þ was generated by the potential sweep by 3 using a potentiostat C2090 and a function generator FG-02 (Toho Technical Research, Japan) with a threeelectrode system in a dark box. Platinum plates were used as the counter and the working electrodes, and an Ag-AgCl electrode was used as the reference electrode. The photon counting was performed by a photomultiplier tube R464 and a photon counter C1230 (Hamamatsu Photonics, Japan). A 50 mM Tris buffer at pH 7.3 was used as an electrolyte solution, which involves 15 mM oxalate ion and 50 mM sodium chloride. RuðphenÞ2þ 3 and kDNA were used as a probe molecule and a model of double-helical DNA, respectively. The solution was prepared as follows. In order to prepare the solution containing various concentrations of the drug from 0.01 to 100 lM, drug was added to the buffer solution containing kDNA, and then RuðphenÞ2þ 3 was added. The concentrations of RuðphenÞ2þ 3 and kDNA were fixed at 1 and 10 lM, respectively. After incubation at 40 °C for 8 h,
the ECL of RuðphenÞ2þ 3 was monitored by the photoncounting system. The potential sweep was performed from 0 to 1250 mV with a speed of 100 mV/s. The control experiment was performed under similar conditions in the absence of DNA. The ECL of RuðphenÞ2þ 3 alone was used as the standard ECL intensity.
Results and discussion ECL of Ru complex in the absence and the presence of DNA When the solution containing RuðphenÞ2þ 3 was subjected to a potential sweep from 0 to 1250 mV, the ECL emission was observed. RuðphenÞ2þ emitted above 3 1000 mV sweep voltage with the maximum intensity at 1250 mV repeatedly [7]. The stable ECL was obtained after the third sweep. The mechanism of ECL of RuðphenÞ2þ 3 has been reported by Carter and Bard [5]. They reported that ECL involves the reaction of the electrogenerated RuðphenÞ3þ 3 with the strong reductant produced as intermediate in the oxidation of oxalate ion [Eq. (5-1)] and with chemically produced RuðphenÞþ 3 species [Eq. (5-2)]. Fig. 3 shows the ECL intensity at a given concentration of RuðphenÞ2þ 3 . The increase in the concentration 2þ of RuðphenÞ3 caused an increase in the intensity of ECL. However, no change in the ECL intensity was observed above ca. 6 lM RuðphenÞ2þ 3 , suggesting the existence of a rate-determining step of the electron transfer reaction between RuðphenÞ2þ 3 and the surface of the working electrode. Upon the addition of kDNA into the solution containing 1 lM RuðphenÞ2þ 3 , the ECL intensity decreased, as shown in Fig. 4. This suggests that the addition of DNA caused the decrease in the concentration of free RuðphenÞ2þ in the bulk solution, associated with the 3 binding of RuðphenÞ2þ with DNA [5]. RuðphenÞ2þ 3 3 bound in DNA emits no luminescence because of the
Fig. 3. Change in ECL intensity as a function of RuðphenÞ2þ con3 centration.
T. Kuwabara et al. / Analytical Biochemistry 314 (2003) 30–37
Fig. 4. Change in the relative ECL intensity of 1 lM of RuðphenÞ2þ 3 as a function of kDNA concentration.
steric hindrance of DNA. The ECL intensity became half at a concentration of ca. 20 lM DNA. Further addition of DNA leads to a negligible change in the ECL intensity. These results demonstrate that, even in the presence of excess DNA, a half amount of RuðphenÞ2þ 3 remains in the bulk solution to emit. This fact is well coincident with the result reported by Barton et al., in which D-type RuðphenÞ2þ 3 prefers to bind to DNA with the major groove binding mode [3]. The K-type RuðphenÞ2þ 3 having no binding ability to DNA remains in the bulk solution and exhibits the ECL emission. It is likely that there is one RuðphenÞ2þ 3 molecule in four helices of DNA under our conditions because the formation of one helix of DNA needs 10 bp. Binding mode of nonanthracycline-type drugs Fig. 5 shows the chemical structures of the drugs used in this study except for the anthracycline drugs.
33
Cisplatin and transplatin, both stereoisomers that differ in geometric arrangement of two attached ammonia molecules and chloride ions as the ligands, are known to bind to duplex DNA [8]. The structure of cisplatin-DNA adduct is of interest because cisplatin is an anticancer agent used to treat a variety of tumors, while trans-isomer is devoid of biological activity [9]. Takahara et al. [10] reported the X-ray crystal structure of platinated oligonucleotides, in which cisplatin binds to guanine residues in DNA. Fig. 6 shows the relative ECL intensity of RuðphenÞ2þ 3 as a function of cisplatin concentration when cisplatin was used as the drug. It was observed, in the absence of DNA, that the ECL intensity of RuðphenÞ2þ 3 even in the presence of excess cisplatin was similar to that of RuðphenÞ2þ 3 alone, which is used as the standard intensity. Thus the ECL intensity was not affected by the presence of cisplatin. This fact indicates that there is no interaction between RuðphenÞ2þ and cisplatin. In the presence of DNA, 3 however, the ECL intensity was affected by the presence of cisplatin. When the solution contained 0.01 lM cisplatin, the intensity was 50% as compared to the intensity of DNA-free solution or to the standard intensity. This suggests that almost all D-type RuðphenÞ2þ 3 binds to DNA under this condition. However, with increasing concentrations of cisplatin, the ECL intensity increased and almost recovered at 100 lM cisplatin. This result demonstrates that the concentration of free RuðphenÞ2þ 3 in the bulk solution increased with increasing concentration of cisplatin, namely, that the binding of RuðphenÞ2þ 3 to DNA is inhibited by cisplatin. At 100 lM cisplatin, almost no RuðphenÞ2þ 3 can bind to DNA, resulting in an ECL intensity similar to that of the DNAfree solution. Cisplatin may bind to the major groove of DNA, causing inhibition of RuðphenÞ2þ 3 binding to this
Fig. 5. Chemical structure of nonanthracycline drugs.
34
T. Kuwabara et al. / Analytical Biochemistry 314 (2003) 30–37
site. This is well coincident with the result obtained by X-ray measurement [10]. However, the dependency of the ECL intensity on the transplatin concentration was different from that on cisplatin. The intensity increased with increasing concentrations of transplatin even in the absence of DNA. The intensity became 3 times higher than the standard intensity of RuðphenÞ2þ when it 3 contained 100 lM transplatin. Since transplatin itself shows no ECL emission, the strong ECL observed resulted from an unexpected reason. Transplatin seems to promote the ECL emission of RuðphenÞ2þ 3 . Therefore the DNA-binding mode of transplatin as well as its binding ability could not be evaluated. Mitomycin C [11] and duocarmycin SA [12] have relatively simple structures and have an ability to alkylate the 2-amino group of guanines and N3 adenine in DNA, respectively. Fig. 7 shows the ECL intensity of RuðphenÞ2þ 3 as a function of the concentration of mitomycin C and duocarmycin SA in the absence and the presence of DNA. The ECL intensities of DNA-free solutions were similar to that of the standard solution
and were not affected by the drug concentrations, suggesting no interaction between these drugs and DNA. In the presence of DNA, their ECL patterns were different from those of cisplatin. The ECL intensities for both drugs were not influenced by the concentrations of mitomycin C and duocarmycin SA in the range from 0.01 to 100 lM of the drugs. No changes in the ECL intensities as a function of the drug concentration suggest that neither drug prevents RuðphenÞ2þ 3 from binding to DNA. It can be concluded, therefore, that these drugs bind to DNA with the minor groove, which is a different site than that of RuðphenÞ2þ 3 . The minor groove binding character of these drugs was coincident with the results reported previously [11,12]. The difference between these drugs was observed in the relative intensity of the ECL, whose intensities were observed ca. 70 and 77% for mitomycin C and duocarmycin SA, respectively. This indicates that the concentration of RuðphenÞ2þ 3 in the bulk solution is higher when the solution contained duocarmycin SA rather than mitomycin C. This may result from the following reasons. When the drug binds to the minor groove in DNA, (i) the bound drug may widen the groove width around the drug, resulting in a decrease of the major groove width and/or (ii) the extra moiety of the drug, which could not bind to the minor groove, may partly bind to the major groove, resulting in a decrease of the binding space of the complex in the major groove. This may occur remarkably in the case of duocarmycin SA rather than mitomycin C. Similar trends were observed in the case of actinomycin D [13] and distamycin A [14], which are the cyclic peptide and naturally occurring polyamine antibiotic, respectively (Fig. 8). Although a little decrease in the ECL intensity of RuðphenÞ2þ was observed with 3 increasing concentrations of actinomycin D in the absence of DNA, the ECL intensity in the presence of DNA was not significantly affected by the concentration of this agent. This result suggests that actinomycin D
Fig. 7. Effect of mitomycin C (square) and duocarmycin SA (circle) on the relative ECL intensity of RuðphenÞ2þ 3 in the absence (white) and the presence (black) of kDNA.
Fig. 8. Effect of actinomycin D (square) and distamycin A (circle) on the relative ECL intensity of RuðphenÞ2þ 3 in the absence (white) and the presence (black) of kDNA.
Fig. 6. Effect of cisplatin on the relative ECL intensity of RuðphenÞ2þ 3 in the absence (white square) and the presence (black square) of kDNA.
T. Kuwabara et al. / Analytical Biochemistry 314 (2003) 30–37
35
binds to DNA with the minor groove binding mode, coincident with the result obtained by X-ray or NMR analysis [13]. While for distamycin A, the ECL intensity was similar to that of the standard solution in the absence of DNA and was not affected by the drug concentration. In the presence of DNA, the 65% ECL intensity was observed at 0.01 lM concentration of distamycin A and the intensity did not change even in the presence of 100 lM drug. These results indicate that distamycin A binds to the minor groove of DNA, which is well coincident with the report by Lian et al. [13]. DNA-binding mode of anthracyclic drugs Fig. 9 shows the chemical structures of anthracyclic drugs. Among them, daunomycin [15] and doxorubicin [16] have a similar structure bearing a glucosamine unit with little difference in the structure in the substitution introduced at C9 position in ring A, the hydroxyl and ester groups for daunomycin and methyl and alcohol groups for doxorubicin, respectively. However, the remarkable difference in the ECL-emitting pattern of RuðphenÞ2þ 3 was observed between these drugs. Fig. 10 shows the effect of daunomycin and doxorubicin on the ECL intensitiy in the absence and the presence of DNA. In the absence of DNA, the ECL intensities were not affected by the concentration of both drugs. In the presence of DNA, however, the increase in the ECL intensity was observed with increasing concentrations of daunomycin, while no increase in the intensity was observed when doxorubicin concentration was increased. These results indicate that daunomycin and doxorubicin bind to DNA with the major and minor groove, respectively. It is interesting that such a small difference in their structures causes the change in the DNA-binding mode of the drug. It was
Fig. 10. Effect of daunomycin (square) and doxorubicin (circle) on the relative ECL intensity of RuðphenÞ2þ 3 in the absence (white) and the presence (black) of kDNA.
observed, furthermore, that the ECL intensity was almost recovered when the solution contained 1 lM daunomycin. Since, in the case of cisplatin, which also binds to the major groove of DNA, the ECL intensity was recovered at 100 lM, it seems that daunomycin binds to DNA with a stronger binding ability than cisplatin. As compared to the standard intensity, doxorubicin showed the relatively high ECL intensity among the minor groove binding drugs, indicating that doxorubicin binds not only to the minor groove but also to the major one. The major site seems to be occupied by a part of doxorubicin, although it stays mainly in the minor groove. Fig. 11 shows the results obtained when chromomycin A3 [17] and DC92-B [18] were used as the antitumor reagent. The ECL intensity of RuðphenÞ2þ was not 3 affected by the concentration of chromomycin A3 in the absence and the presence of DNA. This fact indicates that there are no interactions between RuðphenÞ2þ 3 and chromomycin A3 and that chromomycin A3 binds to
Fig. 9. Chemical structure of anthracycline drugs.
36
T. Kuwabara et al. / Analytical Biochemistry 314 (2003) 30–37
[3]
Fig. 11. Effect of chromomycin A3 (square) and DC92-B (circle) on the relative ECL intensity of RuðphenÞ2þ 3 in the absence (white) and the presence (black) of kDNA [4]
DNA with the minor groove binding mode. On the other hand, in order to evaluate the binding mode of DC92-B it is necessary to correct the result obtained because the ECL intensity is affected by DC92-B in the absence of DNA. The corrected result shows that ECL intensity increased with increasing concentrations of the drug, indicating that DC92-B binds to DNA with the major groove. These results obtained for chromomycin A3 and DC92-B are coincident with the results reported [17,18]. All these results demonstrate that the DNA-binding mode of the antitumor and antiviral agents can be evaluated by utilizing the DNA-binding and the ECLgenerating abilities of RuðphenÞ2þ 3 .
[5]
[6]
[7]
Acknowledgments We are grateful to Kyowa Hakko Kogyo Co., Ltd., for providing us with duocarmycin SA and DC92-B. We thank Dr. Akira Asai and Dr. Hirofumi Nakano in Kyowa Hakko Kogyo Co., Ltd., for valuable discussions. We also thank Professor Masato Nanasawa in Yamanashi University and Professor Takeaki Iida in Saitama University for kind discussions. This work was supported by Kato Memorial Bioscience Fund. Finally we hope that one of the authors, Y.I., who died in September 2000 on the way to achieving this work, will enjoy reading this paper in heaven.
[8]
[9]
[10]
[11]
References [12] [1] Y. Ikariyama, Y. Taguchi, T. Suzawa, E. Kobatake, M. Aizawa, Electrochemiluminescent sensing for the characterization of DNA-interacting antitumor and antiviral agents, Sensor Acta B 13–14 (1993) 725–727. [2] (a) R.H. Terbrueggen, T.W. Johann, J.K. Barton, Functionalized rhodium intercalators for DNA recognition, Inorg. Chem. 37 (1998) 6874–6883;
(b) J.K. Barton, A.L. Raphael, Site-specific cleavage of lefthanded DNA in pBR322 by K-tris(diphenylphenanthroline)cobalt(III), Proc. Natl. Acad. Sci. USA 82 (1985) 6460; (c) J.K. Barton, J.J. Dannenberg, A.L. Raphael, Enantiometric selectivity in binding tris(phenanthroline)zinc(II) to DNA, J. Am. Chem. Soc. 104 (1982) 4967–4969. (a) R.E. Holmlin, E.D.A. Stemp, J.K. Barton, RuðphenÞ2 dppz2þ luminescence: dependence on DNA sequences and groove-binding, Inorg. Chem. 37 (1998) 29–34; (b) C.M. Dupureur, J.K. Barton, Structural studies of K- and D½RuðphenÞ2 dppz2þ Bound to dðGTCGACÞ2 : characterization of enantioselective intercalation, Inorg. Chem. 36 (1997) 33–43; (c) C.M. Dupureur, J.K. Barton, Use of selective deutation and 1 H NMR in demonstrating major groove binding of D-½RuðphenÞ2 dppz2þ to dðGTCGACÞ2 , J. Am. Chem. Soc. 116 (1994) 10286–10287; (d) J.K. Barton, A.T. Danishefsky, J.M. Goldberg, Tris(phenanthroline)ruthenium(II): stereoselectivity in binding to DNA, J. Am. Chem. Soc. 106 (1984) 2172–2176. (a) E. Tuite, P. Lincoln, B. Norden, Photophysical evidence that D- and K-½RuðphenÞ2 ðdppzÞ2þ intercalate DNA from the minor groove, J. Am. Chem. Soc. 119 (1997) 239240; (b) I. Haq, P. Lincoln, D. Suh, B. Norden, B.Z. Chowdhry, J.B. Chires, Interaction of D- and K-½RuðphenÞ2 DPPZ2þ with DNA: a calorimetric and equilibrium binding study, J. Am. Chem. Soc. 117 (1995) 4788–4796; (c) C. Hiort, P. Lincoln, B. Norden, DNA binding of K- and D-½RuðphenÞ2 DPPZ2þ , J. Am. Chem. Soc. 115 (1993) 3448–3454. M.T. Carter, A.J. Bard, Electrochemical investigations of the metal chelates with DNA. 3. Electrogenerated chemiluminscent investigation of the inteaction of tris(1,10-phenanthroline)ruthenium(II) with DNA, Bioconj. Chem. 1 (1990) 257–263. For ECL of metal complexes of this type, see (a) I. Rubinstein, A.J. Bard, Electrogenerated chemiluminscence. 37. Aqueous Ecl systems based on Ruð2; 20 -bipyridineÞ2þ and oxalate or organic 3 acids, J. Am. Chem. Soc. 103 (1981) 512–516. When a constant potential was applied to the solution containing ruthenium complex, strong ECL was observed immediately after the potential application. However, it was rapidly attenuated to a very weak level due to the rapid oxidation of Ru complex at the electrode surface, while other reactions cannot catch up with the electrode reaction. E.E. Trimmer, J.M. Essigmann, in: D.P. Ballou (Ed.), Essays in Biochemistry, vol. 34, Princeton University Press, New Jersey, 1999, pp. 191–211. E. Bernal-Mendez, M. Boudvillain, F. Gonzalei-Vilchez, M. Leng, Chemical versatility of transplatin monofunctional adducts within multiple site-specifically platinated DNA, Biochemistry 36 (1997) 7281–7287. P.M. Takahara, C.A. Frederick, S.J. Lippard, Crystal structure of the anticancer drug cisplatine bound to duplex DNA, J. Am. Chem. Soc. 118 (1996) 12309–12321. (a) G.S. Kumar, S.M. Musser, J. Cumming, M. Tomasz, 2,7Diaminomitosene, a monofunctional mitomycin C derivative, alkylates DNA in the major groove. Structure and base-sequence specificity of the DNA adduct and mechanism of the alkylation, J. Am. Chem. Soc. 118 (1996) 9209–9217; (b) M. Tomasz, R. Lipmann, Alkylation reactions of mitomycin C at acid pH, J. Am. Chem. Soc. 101 (1979) 6063–6067. (a) A. Asai, S. Nagamura, H. Saito, A novel property of duocarmycin and its analogues for covalent reaction with DNA, J. Am. Chem. Soc. 116 (1994) 4171–4177; (b) A. Asai, K. Yano, T. Mizukami, H. Nakano, Characterization of duocarmycin-DNA adduct-recognizing protein in cancer cells, Cancer Res. 59 (1999) 5417–5420; (c) D.L. Boger, D.S. Johnson, CC-1065 and the duocarmycins: unraveling the keys to a new class naturally derived DNA
T. Kuwabara et al. / Analytical Biochemistry 314 (2003) 30–37 alkylating agent, Proc. Natl. Acad. Sci. USA 92 (1995) 3642– 3649. [13] (a) C. Lian, H. Robinson, A.H.-J. Wang, Structure of actinomycin D bound with ðGAAGCTTCÞ2 and ðGATGCTTCÞ2 and its binding to the ðCAGÞn :ðCTGÞn triplet sequence as determined by NMR analysis, J. Am. Chem. Soc. 118 (1996) 8791– 8801; (b) H. Robinson, Y.-G. Gao, X.-l. Yang, R. Sanishvili, A. Joachimiak, A.H.J. Wang, Crystallographic analysis of a novel complex of actinomycin D bound to the DNA decamer CGATCGATCG, Biochemistry 40 (2001) 5587–5592. [14] (a) R. Baliga, D.M. Crothers, The kinetic basis for sequence discrimination by distamycin A, J. Am. Chem. Soc. 122 (2000) 11751–11752; (b) J. Lah, G. Vesnaver, Binding of distamycin A and netropsin to the 12mer DNA duplexes containing mixed AT-GC sequences with at most five or three successive AT base pairs, Biochemistry 39 (2000) 9317–9326.
37
[15] C.A. Frederick, L.D. Williams, G. Ughetto, G.A. van der Marel, J.H. van Boom, A. Rich, A.H.-J. Wang, Structural comparison of anticancer drug-DNA complexes: adriamycin and daunomycin, Biochemistry 29 (1990) 2538–2549. [16] G.E. Kellogg, J.N. Scarsdale, F.A. Fornari Jr., Identification and hydropathic characterization of structural features affecting sequence specificity for doxorubicin intercalation into DNA doublestranded polynucleotides, Nucleic Acids Res. 20 (1998) 4721–4732. [17] (a) X. Gao, D.J. Patel, Chromomycin dimer-DNA oligomer complexes. Sequence selectivity and divalent cation specificity, Biochem. 28 (1990) 10940–10956; (b) X. Gao, P. Miura, D.J. Patel, Structure refinement of the chromomycin dimer-DNA oligomer complex in solution, Mol. Biol. 223 (1992) 259–279. [18] D. Sun, J. Hansen, J.J. Clement, L.H. Hurley, Structure of altromycin B(N7-Guanine)-DNA adduct. A proposed prototypic DNA adduct structure for the pluramycin antitumor antibiotics, Biochemistry 32 (1993) 8068–8074.