PCCP View Article Online
Published on 01 August 2014. Downloaded by University of Newcastle on 12/09/2014 14:05:04.
PAPER
Cite this: Phys. Chem. Chem. Phys., 2014, 16, 19290
View Journal | View Issue
Interaction of electrons with cisplatin and the subsequent effect on DNA damage: a density functional theory study† Hsing-Yin Chen,* Hui-Fen Chen, Chai-Lin Kao, Po-Yu Yang and Sodio C. N. Hsu Cisplatin, Pt(NH3)2Cl2, is a leading chemotherapeutic agent that has been widely used for various cancers. Recent experiments show that combining cisplatin and electron sources can dramatically enhance DNA damage and the cell-killing rate and, therefore, is a promising way to overcome the side effects and the resistance of cisplatin. However, the molecular mechanisms underlying this phenomenon are not clear yet. By using density functional theory calculations, we confirm that cisplatin can efficiently capture the prehydrated electrons and then undergo dissociation. The first electron attachment triggers a spontaneous departure of the chloride ion, forming a T-shaped [Pt(NH3)2Cl] neutral radical, whereas the second electron attachment leads to a spontaneous departure of ammine, forming a linear [Pt(NH3)Cl] anion. We further recognize that the one-electron reduced product [Pt(NH3)2Cl] is extremely harmful to DNA. It can abstract hydrogen atoms from the C–H bonds of the ribose moiety and the methyl group of thymine, which in turn leads to DNA strand breaks and cross-link
Received 26th May 2014, Accepted 24th July 2014
lesions. The activation energies of these hydrogen abstraction reactions are relatively small compared to the hydrolysis of cisplatin, a prerequisite step in the normal mechanism of action of cisplatin. These
DOI: 10.1039/c4cp02306d
results rationalize the improved cytotoxicity of cisplatin by supplying electrons. Although the biological effects of the two-electron reduced product [Pt(NH3)Cl] are not clear at this stage, our calculations
www.rsc.org/pccp
indicate that it might be protonated by the surrounding water.
1. Introduction Cisplatin, Pt(NH3)2Cl2, is one of the most widely used anticancer drugs nowadays. It has been successfully applied to the treatment of various cancers such as testicular, ovarian, head/neck, and small-cell lung cancers.1–3 It is generally accepted that the target of cisplatin is DNA: it prefers binding to the N7 sites of two consecutive guanines, leading to intrastrand cross links and the subsequent cell apoptosis.1–10 Prior to binding with DNA, cisplatin has to undergo a hydrolysis reaction in which one or two chlorides are replaced by water molecules. Experimental measurements demonstrate that the hydrolysis of cisplatin is a slow process with an activation energy of 23–24 kcal mol1,11–13 consistent with the theoretical predictions.14–19 The bioactivity of cisplatin might also involve many other reactions such as Department of Medicinal and Applied Chemistry, Kaohsiung Medical University, Kaohsiung 807, Taiwan. E-mail: [email protected]; Fax: +886 7 3125339; Tel: +886 7 3121101 ext. 2807 † Electronic supplementary information (ESI) available: Calculations of 17 explicit water molecules plus the SMD bulk hydration model, B3LYP results, relative stability of sugar radicals, the temperature effect, the cavity effect, bond energies of nucleobases, Cartesian coordinates of optimized structures. See DOI: 10.1039/ c4cp02306d
19290 | Phys. Chem. Chem. Phys., 2014, 16, 19290--19297
tautomerization of nucleobases10,20 and interactions with nonDNA targets.21 Despite the success of cisplatin in the cancer therapy, it has two main disadvantages: severe side effects and intrinsic/ acquired resistance. To overcome these drawbacks, enormous efforts and resources have been invested in developing new Pt-based anticancer drugs in the past 35 years. To date, two new platinum drugs, carboplatin and oxaliplatin, have been approved by the US Food and Drug Administration (FDA).1–3 Compared with cisplatin, carboplatin has the advantage of lower toxicity and is thus safer, whereas oxaliplatin displays a broader spectrum of antitumour activity.3 In addition to developing new drugs, an alternative way is to use an old drug but modify its mechanism of action by certain chemical or physical approaches. Recently, a series of experiments have demonstrated that this strategy works.22–29 Zheng et al.22 and Xiao et al.25 examined the influences of cisplatin on the sensitization of DNA toward secondary low-energy electrons (LEEs). They observed that the yields of DNA strand breaks and glycosidic bond cleavages induced by LEEs were enhanced in the solid films of cisplatin-DNA adducts in comparison with the solid films consisting of DNA alone.22,23,25 A similar radiosensitization effect of cisplatin in the aqueous solution was
This journal is © the Owner Societies 2014
View Article Online
Published on 01 August 2014. Downloaded by University of Newcastle on 12/09/2014 14:05:04.
Paper
PCCP
reported by Rezaee et al. as well;24 this work demonstrates that the formation of cisplatin-DNA adducts boosts DNA strand breaks induced by both hydrated electrons and hydroxyl radicals, which are the two major species produced in the radiolysis of water. The hydroxyl radical plays a pivotal role in DNA damage and the related mechanisms have been extensively studied. It mainly undergoes two types of reactions as the first step toward DNA strand breaks: addition to C–C and C–N double bonds in nucleobases and hydrogen abstraction from the sugar moiety.30–32 In contrast, the effect of LEEs on DNA damage has long been neglected until 2000.33 Recently, there has been experimental evidence showing that LEEs might be more efficient than hydroxyl radicals in causing DNA strand breaks.34 Lu and coworkers investigated the interactions of cisplatin with electrons by time-resolved femtosecond laser spectroscopy. They discovered that cisplatin is highly reactive toward the dissociative electron transfer reaction with weakly bound electrons that may come from two-photon absorption of water,26 excitation of photosensitizers,27 or reducing agents.29 More importantly, they further showed that the combination of cisplatin and electron sources can significantly improve the yield of DNA strand breaks and the rate of cell killing, even for the cisplatin-resistant cell lines.27,29 Since no incubation time of cisplatin with DNA was needed to detect DNA strand breaks in Lu’s experiments,27,28 the enhancement of DNA damage must be a consequence of the interactions of electrons with free cisplatin rather than with cisplatin-DNA adduct. Here, we report a density functional theory study which aims to provide the molecular level understanding of Lu’s experimental findings. The dissociative behaviors of cisplatin upon one- and two-electron attachments were investigated. In addition, a number of possible reactions that cause DNA damage by the dissociative products were studied.
2. Computational methods All calculations were accomplished by the Gaussian 09 program.35 Hybrid meta functional M0636 was employed in the present study. This functional is recommended for dealing with the chemical reactions where both organic and transition-metal bonds are formed or broken.37 In addition, the capability of M06 functional to provide satisfactory description for noncovalent interactions has been verified as well.37 The 6-31+G(d) or 6-31++G(d,p) basis sets were used for the main group elements. The Pt atom was described by the Stuttgart–Dresden pseudopotential (SDD) and
Fig. 1
the corresponding basis set augmented by a set of polarization (af = 0.98) and diffuse functions (as = 0.0075, ap = 0.013, ad = 0.025).38 The combinations of the augmented SDD with the 6-31+G(d) and with the 6-31++G(d,p) are denoted by BS-1 and BS-2, respectively. The SMD continuum solvation model39 was employed to simulate the bulk hydration environment (e = 78.3553). Geometry optimizations and vibrational frequency calculations were carried out using the SMD solvation model. An ultrafine grid was used for numerical integrations. Free energy corrections were performed under the standard conditions of 298.15 K and 1 atm. Intrinsic reaction coordinate (IRC) calculations40 were performed for the transition state structures to confirm that they connect the corresponding reactants and products. It has been pointed out by Wertz41 and Abraham42 that all molecules lose the same fraction (B0.5) of their entropies when they are transferred from gas phase into aqueous solution. This entropy loss results from the fact that the free volume of a molecule in solution is reduced, and its translational, rotational, and probably parts of vibrational motions are subjected to the restriction imposed by the solvent cage. Continuum solvation models cannot provide a correct description of this effect; as a result, the entropic penalties and, thus, the activation and reaction free energies for multimolecular reactions in solutions are usually overestimated.43–46 To take into account the effect of restrained motions of solutes in solution, the free energies were evaluated by G = H 0.5TS. The empirical Wertz correction has been successfully applied to reproduce experimental measurements in many studies.17,47–49
3. Results and discussion 3.1
Dissociation of cisplatin by electron attachment
To explore the dissociative behavior of cisplatin in the aqueous environment, we employed a microhydrated structure of cisplatin including 15 explicit water molecules in the first hydration layer; this microhydrated model has been previously used to study the hydrolysis of cisplatin.19 We further improved the hydration model by immersing this supermolecule of cisplatin plus 15H2O in the bulk hydration environment described by the SMD continuum solvation model. The optimized structures of cisplatin and its one- and twoelectron adducts in the aqueous environment are shown in Fig. 1 and the relevant bond distances are presented in Table 1. For the neutral cisplatin, the optimized distances of Pt–Cl and
SMD/M06/BS-1 optimized structures of cisplatin and its one- and two-electron adducts.
This journal is © the Owner Societies 2014
Phys. Chem. Chem. Phys., 2014, 16, 19290--19297 | 19291
View Article Online
PCCP
Paper
Table 1 Bond distances (Å) of cisplatin and its one- and two-electron adducts optimized in the presence of 15 explicit water molecules plus a SMD bulk hydration environment
Pt(NH3)2Cl2 Pt–ClL a Pt–Cl Pt–NL a Pt–N
2.39 2.38 2.04 2.06
(2.39)b (2.37)b (2.04)b (2.05)b
[Pt(NH3)2Cl]
[Pt(NH3)Cl]
3.82 2.41 2.26 2.10
4.11 2.37 3.59 2.08
Published on 01 August 2014. Downloaded by University of Newcastle on 12/09/2014 14:05:04.
a
The subscript L denotes the leaving groups. b The values in parentheses are gas-phase optimized distances in the presence of 15 explicit water molecules (ref. 19).
Pt–N are 2.38, 2.39 and 2.04, 2.06 Å respectively. These values are very close to the previous gas-phase results incorporating 15 explicit H2O.19 Geometry optimizations revealed that when an extra electron was added to the neutral cisplatin one of the chlorides spontaneously left (the distance of Pt–ClL is 3.82 Å), resulting in the formation of a T-shaped [Pt(NH3)2Cl] neutral radical. It is worth noting that the distance between the Pt center and the NH3 opposite to the vacant site (i.e., Pt–NL) is substantially elongated to 2.26 Å. When the second electron was added to the [Pt(NH3)2Cl] , this ammonia immediately departed (the distance of Pt–NL is 3.59 Å), giving a linear [Pt(NH3)Cl] complex anion with the bond lengths of Pt–Cl and Pt–N being 2.37 and 2.08 Å, respectively. Notice that the leaving ammonia interacts with the Pt(0) center by a noncovalent NL–H Pt interaction (the distance of H Pt is 2.56 Å). If there exists a substantial barrier in the dissociative processes, geometry optimizations starting from Pt(NH3)2Cl2 + e and [Pt(NH3)2Cl] + e should, respectively, converge to the local minima corresponding to [Pt(NH3)2Cl2] and [Pt(NH3)2Cl], rather than directly converging to the dissociative products [Pt(NH3)2Cl] + Cl and [Pt(NH3)Cl] + NH3. However, this is not the truth, implying that the barriers, if exist, are small, and the cisplatin readily dissociates upon electron attachment. This result is consistent with the experimental observation that the transient absorption of cisplatin decays within a few tens of picoseconds after interacting with electrons.26 We also notice that the water molecules interact with the Pt center by their O–H bonds rather than by oxygen atoms. These O–H Pt interactions (blue dot line in Fig. 1) are characterized by the bond lengths considerably shorter than the sum of the van der Waals radii of hydrogen (1.20 Å) and platinum (1.75 Å), varying from 2.29 to 2.67 Å. In addition, the O–H Pt interactions display directionality with the bonding angles being in the range of 1571–1691. The number and strength (judging by bond length) of the O–H Pt interaction increase with the reduction state. This observation gives us a hint that the electron-rich Pt(0) center of [Pt(NH3)Cl] could be able to activate the polar bonds such as the O–H bond of water. In fact, the O–H Pt interaction has been recognized in many crystal structures and computational analyses have shown that the dispersion force is responsible for the formation of such an interaction.50,51 To assess the electron-accepting capability of cisplatin, the vertical electron affinities (EAv) of Pt(NH3)2Cl2 and [Pt(NH3)2Cl]
19292 | Phys. Chem. Chem. Phys., 2014, 16, 19290--19297
were calculated. The first and the second EAv values of cisplatin were evaluated to be 1.895 and 2.400 eV, respectively. Comparing these values with the electron binding energies of the prehydrated electron (1.0–1.5 eV)52 and the fully hydrated electron (3.3 eV)53 suggests that cisplatin can only be reduced by the former. It is remarkable that the EAv value of cisplatin is larger than that of the guanine–cytosine base pair (1.433 eV).54 Since cytosine is the most electron-affinic site in DNA, this result indicates that cisplatin has a higher ability to capture electrons than DNA. The hydration number for the first coordination shell of cisplatin might be higher as recently pointed out by a classical molecular dynamics simulations study.55 To test the influence of the number of explicit water molecules, we further augmented the solvation model by including two more water molecules around the ammine ligands (i.e., 17H2O + SMD model). In addition, in order to check the DFT functional dependence we also calculated the interaction of electrons with cisplatin by the B3LYP method.56,57 All these calculations revealed that while the optimized geometries are somewhat different, the main results and conclusions regarding the dissociation behavior, the characteristic interactions between water and platinum, and the electron affinity are not changed (Fig. S1 and S2 and Tables S1 and S2 in ESI†). 3.2
Reactivity of one-electron reduced product [Pt(NH3)2Cl]
3.2.1 DNA strand break by direct attack on the C3 0 –O bond. The next question is how these dissociative products damage DNA? According to the experimental results, the major forms of DNA damage induced by the simultaneous treatment of cisplatin and electrons are strand breaks, while the cross links are minor forms.27 We found that the [Pt(NH3)2Cl] can cause DNA strand break via direct oxidative addition to the C3 0 – O bond (red path in Fig. 2). According to the results of Wertzcorrected free energies, initially a stable reactant adduct (RA), which lies 2.3 kcal mol1 below the separated reactants (SR), can be formed via the hydrogen bonding between NH3 of [Pt(NH3)2Cl] and the O atom of the phosphate group. Then the Pt(I) center attacks the O atom connecting to the C3 0 (TS1), leading to a homolytic cleavage of the C30 –O bond and releasing the separated products of [Pt(NH3)2(HPO4)Cl] and C30 -sugarradical (SP1). Although this reaction is only slightly endergonic by 3.6 kcal mol1, it has a large energetic barrier of 33.6 kcal mol1. Without the Wertz correction, RA was predicted to be unstable (3.6 kcal mol1 higher than SR). If RA is not a stable complex the energetic barrier of the oxidative addition of [Pt(NH3)2Cl] to the C3 0 –O bond should be evaluated by the free-energy difference between the TS1 and the SR (rather than RA), and the resultant value becomes 38.2 kcal mol1, substantially larger than the prediction including Wertz corrections. However, the Wertz correction has a minor effect on the evaluation of the reaction free energy. This is because the molecular number of SP1 and SR is equal and, thus, the effect of Wertz corrections is largely cancelled out. Both results, no matter with or without Wertz corrections, suggest that the reaction of direct oxidative addition to the C–O
This journal is © the Owner Societies 2014
View Article Online
Published on 01 August 2014. Downloaded by University of Newcastle on 12/09/2014 14:05:04.
Paper
PCCP
Fig. 2 SMD/M06/BS-2 free energy profile (in kcal mol1) for C3 0 –O bond cleavage induced by [Pt(NH3)2Cl] . The values in parentheses are results without Wertz corrections.
bond is too slow to account for the observed high efficiency of cisplatin plus electrons in causing DNA strand breaks. 3.2.2 DNA strand break by H4 0 -abstraction. An alternative pathway leading to DNA strand break is the hydrogen abstraction from the ribose moiety (blue path in Fig. 2). Starting from the same reactant adduct, the Pt(I) center could attack and abstract the H40 atom (TS2), resulting in the formation of C40 -nucleotide-radical (1). The Wertz-corrected activation free energy of the H40 -abstraction is 16 kcal mol1 (20.3 kcal mol1 without Wertz corrections). The C30 – O bond of the radical 1 is significantly weakened and immediately undergoes a heterolytic cleavage to release [HPt(NH3)2Cl]HPO42 and sugar radical cation (2); the Wertz-corrected dissociation free energy of this bond was estimated to be only 14.2 kcal mol1 (7.1 kcal mol1 without Wertz corrections), considerably lower than 80–90 kcal mol1 for the typical C–O bonds. The unusual fragility of the C30 –O bond in radical 1 can be rationalized by a high stability of the resulting sugar radical cation due to the presence of three resonance forms (cf. chemical structures of 2 in Fig. 2).58 The [HPt(NH3)2Cl]HPO42 and the sugar radical cation might recombine to form intermediate 3 through the Pt C30 interaction. A rapid hydride transfer from Pt to C30 (TS3) subsequently takes place, producing the separated products of [Pt(NH3)2(HPO4)Cl] and C40 sugar-radical (SP2); the barrier of this step is only 8 kcal mol1. The overall reaction is almost thermodynamically neutral. The activation energies along this reaction pathway do not exceed 16 kcal mol1 (or 20 kcal mol1 without Wertz corrections) and might be responsible for the high yield of DNA strand breaks caused by cisplatin plus electrons. We would like to emphasize that the possibilities of hydrogen abstractions from other sites of sugar cannot be excluded.
This journal is © the Owner Societies 2014
The resultant sugar radicals have been shown to be highly reactive and will undergo various reactions that eventually lead to strand breaks.58 The reason we selected the 4 0 -hydrogen atom as the target in the present study is as follows. First, the H4 0 and the H5 0 are the two most accessible sites in B-form DNA. Secondly, the resultant C4 0 -based sugar radical is most stable (Table S3 in ESI†), indicating that the H4 0 -abstraction should be most favorable in energetics. This is because the other component of the products, namely HPt(NH3)2Cl, being the same, the reaction energies of the hydrogen abstractions are expected to be principally determined by the stability of the resultant sugar radicals. In addition, the transferred hydrogen is already far from the carbon center in the transition state (see TS2 in Fig. 2), implying that the sugar moiety in the transition state bears a substantial radical character. Accordingly, the reaction barriers are also expected to correlate with the stability of sugar radicals. In fact, we have calculated the reaction of H5 0 abstraction (Fig. 3), and the resultant barrier and reaction energy are indeed larger than that of H4 0 -abstraction, supporting the aforementioned inference. 3.2.3 H-abstraction from the methyl group of thymine. Experimental studies have shown that among four nucleobases only the methyl group of thymine can undergo the hydrogen abstraction reaction with hydroxyl radicals.30,31,58 This experimental result is in harmony with the calculations of bond energies; the C–H bond of the methyl group of thymine is characterized by a relatively small bond dissociation energy compared to other C–H and N–H bonds of nucleobases (Table S5 in ESI†). Accordingly, we investigated the reaction of [Pt(NH3)2Cl] to abstract the hydrogen atom from the methyl group of thymine (Fig. 4).
Phys. Chem. Chem. Phys., 2014, 16, 19290--19297 | 19293
View Article Online
Published on 01 August 2014. Downloaded by University of Newcastle on 12/09/2014 14:05:04.
PCCP
Fig. 3 SMD/M06/BS-2 free energy profile (in kcal mol1) for H5 0 abstraction from nucleotide by [Pt(NH3)2Cl] . The values in parentheses are results without Wertz corrections.
The Wertz-corrected results reveal that the [Pt(NH3)2Cl] can attach to thymine through the hydrogen bond between NH3 of [Pt(NH3)2Cl] and O4 of thymine; the RA is more stable than the SR by 2.1 kcal mol1. Then the Pt(I) center approaches and abstracts the hydrogen atom of the methyl group with an activation energy of 19.2 kcal mol1. Without the Wertz corrections, the complex of [Pt(NH3)2Cl] and thymine was predicted to be instable, and the activation energy was estimated to be 23.4 kcal mol1. Although this H-abstraction reaction is endergonic by 11 kcal mol1, it is possible that the reaction could be further driven by thermodynamically favorable reactions of the methyl radical of thymine. For example, there has been
Fig. 4 SMD/M06/BS-2 free energy profile (in kcal mol1) for H-abstraction from the methyl group of thymine by [Pt(NH3)2Cl] . The values in parentheses are results without Wertz corrections.
19294 | Phys. Chem. Chem. Phys., 2014, 16, 19290--19297
Paper
experimental evidence showing that the methyl radical of thymine can form covalent bonds with adjacent nucleobases, leading to intrastrand59,60 and interstrand61–63 cross-link lesions. Note that the activation energy of H-abstraction from the methyl group of thymine is higher than that from the sugar moiety. This result might partially explain why the major forms of DNA damage caused by cisplatin plus electrons are strand breaks rather than cross links. Considering the steric factor, the methyl group of thymine is exposed to the major groove of DNA and is not sterically hindered by neighboring bases and, therefore, can be accessed by [Pt(NH3)2Cl] . 3.2.4 Binding to the N7 site of guanine and adenine. It is well-known that cisplatin, upon hydrolysis, is highly inclined to bind with the N7 position of purines.1–10 To understand whether or not the reduced cisplatin still has this characteristic property, the binding free energies of [Pt(NH3)2Cl] with guanine–cytosine (GC) and adenine–thymine (AT) base pairs were calculated. The optimized structures are shown in Fig. 5. The calculations reveal that the high affinity of cisplatin for the N7 position of purines is almost vanished upon reduction. The binding free energies of [Pt(NH3)2Cl] with GC and AT are only 0.5 and 1.7 kcal mol1, respectively, even lower than 2.3 kcal mol1 of [Pt(NH3)2Cl] and the phosphate backbone (RA in Fig. 2). This rationalizes why the [Pt(NH3)2Cl] would not bind to purine bases but instead attack the DNA backbone and cause strand breaks. For comparison, the binding free energies of the hydrolysis product [Pt(NH3)2Cl]+ with GC and AT were calculated as well, and the values are, respectively, 37.2 and 35.6 kcal mol1, supporting the high affinity of cisplatin for purines. 3.3
Reactivity of two-electron reduced product [Pt(NH3)Cl]
With respect to the reactivity of the two-electron reduced product [Pt(NH3)Cl], we did not consider the possibility of reaction with DNA due to the negatively charged nature of the
Fig. 5 SMD/M06/BS-2 optimized structures of (a) [Pt(NH3)2Cl]+GC (b) [Pt(NH3)2Cl] GC (c) [Pt(NH3)2Cl]+AT (d) [Pt(NH3)2Cl] AT.
This journal is © the Owner Societies 2014
View Article Online
Published on 01 August 2014. Downloaded by University of Newcastle on 12/09/2014 14:05:04.
Paper
PCCP
Fig. 6 SMD/M06/BS-2 free energy profile (in kcal mol1) for protonation of [Pt(NH3)Cl] by water. The values in parentheses are results without Wertz corrections.
DNA backbone. Instead, we explored the reaction of [Pt(NH3)Cl] with water, inspired by the observation of strong O–H Pt interactions. The calculations reveal that the electron-rich Pt(0) center of [Pt(NH3)Cl] can be protonated by water via a concerted double proton transfer involving the water molecules in the first and second hydration layers; the activation and reaction energies are 14.6 and 13.1 kcal mol1, respectively (Fig. 6). Through this mechanism the resultant OH anion can be conveyed to the second hydration layer and fully stabilized by three water molecules, which has the advantage in lowering the reaction energy.64 A similar mechanism has been observed in the protonation of nucleobase anions by water.64 The example of the Pt(0) complex able to activate the O–H bond of water has been previously reported.65 The OH anion is a strong base and it can damage DNA by deprotonating nucleobases. Computational studies have shown that deprotonation of amine groups of adenine and cytosine by OH is a highly exothermic and barrier-free reaction.66,67 In addition, OH may also act as a nucleophile to attack the C1 0 site of sugar and lead to glycosidic bond cleavage through the SN2 mechanism.68,69 On the other hand, it is not clear yet whether the platinum hydride plays a role in DNA damage. While myriads of studies on metal hydride complexes have been appeared, most of these studies address the catalytic issue. After literature searching, we only found one paper reporting that the iridium hydrides can interact with DNA and induce cell apoptosis and, therefore, exhibit antitumor activity.70 More experimental and computational studies are needed to shed light on the biological activities of metal hydrides.
4. Conclusions The mechanism of action of combining cisplatin and electron sources is distinct from that of treating with cisplatin alone. Our DFT study demonstrates that the one-electron reduced product of cisplatin, [Pt(NH3)2Cl] , loses high affinity for binding with purines but damages DNA by abstracting hydrogens from ribose or methyl
This journal is © the Owner Societies 2014
Fig. 7 SMD/M06/BS-2 simulated IR spectrum of HPt(NH3)2Cl, cisplatin, and 2 0 -deoxycitidine-3 0 -monophosphate (3 0 -dCMP).
groups of thymine instead. The activation energies of these H-abstraction reactions (16 and 19 kcal mol1) are considerably lower than that of the hydrolysis of cisplatin (ca. 24 kcal mol1), a prerequisite step in the normal mechanism of action, explaining why the combination of cisplatin and electron sources can significantly enhance DNA damage and the cell-killing rate. The mechanisms proposed in the present study remain to be verified by experiments. A feasible experimental test is to detect the formation of the platinum hydride complex in the H-abstraction reactions. This could be achieved by monitoring the characteristic IR absorption of the platinum hydride complex. The vibrational frequency analyses show that the HPt(NH3)2Cl complex displays an intense absorption at 2155 cm1, which corresponds to the stretching of the Pt–H bond (Fig. 7). As cisplatin and DNA do not absorb at 2000–3000 cm1, the unique absorption of the Pt–H bond might be exploited to probe the H-abstraction reactions from DNA by [Pt(NH3)2Cl] .
Acknowledgements We thank the Ministry of Science and Technology of Taiwan for financial support and the National Center for High-performance Computing for computer time and facilities.
References 1 D. Wang and S. J. Lippard, Nat. Rev. Drug Discovery, 2005, 4, 307.
Phys. Chem. Chem. Phys., 2014, 16, 19290--19297 | 19295
View Article Online
Published on 01 August 2014. Downloaded by University of Newcastle on 12/09/2014 14:05:04.
PCCP
2 R. A. Alderden, M. D. Hall and T. W. Hambley, J. Chem. Educ., 2006, 83, 728. 3 L. Kelland, Nat. Rev. Cancer, 2007, 7, 573. 4 E. R. Jamieson and S. J. Lippard, Chem. Rev., 1999, 99, 2467. 5 J. Reedijk, Chem. Rev., 1999, 99, 2499. ´rez, Chem. Rev., 2003, 6 M. A. Fuertes, C. Alonso and J. M. Pe 103, 645. 7 Y. Jung and S. J. Lippard, Chem. Rev., 2007, 107, 1387. 8 M. H. Baik, R. A. Friesner and S. J. Lippard, J. Am. Chem. Soc., 2003, 125, 14082. 9 A. Robertazzi and J. A. Platts, Chem. – Eur. J., 2006, 12, 5747. 10 T. Matsui, Y. Shigeta and K. Hirao, J. Phys. Chem. B, 2007, 111, 1176. 11 J. R. Perumareddi and A. W. Adamson, J. Phys. Chem., 1968, 72, 414. 12 J. W. Reishus and D. S. Martin, J. Am. Chem. Soc., 1961, 83, 2457. 13 R. N. Bose, R. E. Viola and R. D. Cornelius, J. Am. Chem. Soc., 1984, 106, 3336. 14 P. Carloni, M. Sprik and W. Andreoni, J. Phys. Chem. B, 2000, 104, 823. 15 Y. Zhang, Z. Guo and X. Z. You, J. Am. Chem. Soc., 2001, 123, 9378. 16 J. V. Burda, M. Zeizinger and J. Leszczynski, J. Comput. Chem., 2005, 26, 907. 17 J. K. C. Lau and D. V. Deubel, J. Chem. Theory Comput., 2006, 2, 103. 18 J. K. C. Lau and B. Ensing, Phys. Chem. Chem. Phys., 2010, 12, 10348. 19 A. Melchior, E. S. Marcos, R. R. Pappalardo and J. M. Martı´nez, Theor. Chem. Acc., 2011, 128, 627. ´n-Carrasco, D. Jacquemin and E. Caue ¨t, Phys. 20 J. P. Cero Chem. Chem. Phys., 2012, 14, 12457. 21 V. Cepeda, M. A. Fuertes, J. Castilla, C. Alonso, C. Quevedo ´rez, Anti-Cancer Agents Med. Chem., 2007, 7, 3. and J. M. Pe 22 Y. Zheng, D. J. Hunting, P. Ayotte and L. Sanche, Phys. Rev. Lett., 2008, 100, 198101. 23 L. Sanche, Chem. Phys. Lett., 2009, 474, 1. 24 M. Rezaee, L. Sanche and D. J. Hunting, Radiat. Res., 2013, 179, 323. 25 F. Xiao, X. Luo, X. Fu and Y. Zheng, J. Phys. Chem. B, 2013, 117, 4893. 26 Q. B. Lu, S. Kalantari and C. R. Wang, Mol. Pharm., 2007, 4, 624. 27 Q. B. Lu, J. Med. Chem., 2007, 50, 2601. 28 Q. B. Lu, Mutat. Res., Rev. Mutat. Res., 2010, 704, 190. 29 T. Luo, J. Yu, J. Nguyen, C. R. Wang, R. G. Bristow, D. A. Jaffray, X. Z. Zhou, K. P. Lu and Q. B. Lu, Proc. Natl. Acad. Sci. U. S. A., 2012, 109, 10175. 30 C. J. Burrows and J. G. Muller, Chem. Rev., 1998, 98, 1109. 31 C. von Sonntag, Free-Radical-Induced DNA Damage and Its Repair, A Chemical Perspective, Springer-Verlag, Berlin, Heidelberg, New York, 2006. 32 R. B. Zhang and L. A. Eriksson, J. Phys. Chem. B, 2007, 111, 6571. 33 B. Boudaı¨ffa, P. Cloutier, D. Hunting, M. A. Huels and L. Sanche, Science, 2000, 287, 1658.
19296 | Phys. Chem. Chem. Phys., 2014, 16, 19290--19297
Paper
34 J. Nguyen, Y. Ma, T. Luo, R. G. Bristow, D. A. Jaffray and Q. B. Lu, Proc. Natl. Acad. Sci. U. S. A., 2011, 108, 11778. 35 M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, G. Scalmani, V. Barone, B. Mennucci, G. A. Petersson, H. Nakatsuji, M. Caricato, X. Li, H. P. Hratchian, A. F. Izmaylov, J. Bloino, G. Zheng, J. L. Sonnenberg, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, T. Vreven, J. A. Montgomery, Jr., J. E. Peralta, F. Ogliaro, M. Bearpark, J. J. Heyd, E. Brothers, K. N. Kudin, V. N. Staroverov, R. Kobayashi, J. Normand, K. Raghavachari, A. Rendell, J. C. Burant, S. S. Iyengar, J. Tomasi, M. Cossi, N. Rega, J. M. Millam, M. Klene, J. E. Knox, J. B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R. E. Stratmann, O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski, R. L. Martin, K. Morokuma, V. G. Zakrzewski, G. A. Voth, P. Salvador, J. J. Dannenberg, S. Dapprich, A. D. Daniels, O. Farkas, J. B. Foresman, J. V. Ortiz, J. Cioslowski and D. J. Fox, Gaussian, 09, Revision A.02, Gaussian, Inc., Wallingford CT, 2009. 36 Y. Zhao and D. G. Truhlar, Theor. Chem. Acc., 2008, 120, 215. 37 Y. Zhao and D. G. Truhlar, Acc. Chem. Res., 2008, 41, 157. 38 J. V. Burda, M. Zeizinger, J. Sponer and J. Leszczynski, J. Chem. Phys., 2000, 113, 2224. 39 A. V. Marenich, C. J. Cramer and D. G. Truhlar, J. Phys. Chem. B, 2009, 113, 6378. 40 K. Fukui, J. Phys. Chem., 1970, 74, 4161. 41 D. H. Wertz, J. Am. Chem. Soc., 1980, 102, 5316. 42 M. H. Abraham, J. Am. Chem. Soc., 1981, 103, 6742. 43 Y. Liang, S. Liu, Y. Xia, Y. Li and Z. X. Yu, Chem. – Eur. J., 2008, 14, 4361. 44 F. Huang, G. Lu, L. Zhao, H. Li and Z. X. Wang, J. Am. Chem. Soc., 2010, 132, 12388. ´n and A. Warshel, J. Phys. Chem. B, 1998, 102, 719. 45 J. Floria 46 Z. X. Yu and K. N. Houk, J. Am. Chem. Soc., 2003, 125, 13825. 47 J. Cooper and T. Ziegler, Inorg. Chem., 2002, 41, 6614. 48 J. Kua, S. W. Hanley and D. O. De Haan, J. Phys. Chem. A, 2008, 112, 66. 49 Y. Lattach, P. Archirel and S. Remita, J. Phys. Chem. B, 2012, 116, 1467. `s, R. Attias and J. Fraitag, Angew. Chem., 50 J. Kozelka, J. Berge Int. Ed., 2000, 39, 198. `s, J. Caillet, J. Langlet and J. Kozelka, Chem. Phys. 51 J. Berge Lett., 2001, 344, 573. 52 C. R. Wang, T. Luo and Q. B. Lu, Phys. Chem. Chem. Phys., 2008, 10, 4463. 53 Y. Tang, H. Shen, K. Sekiguchi, N. Kurahashi, T. Mizuno, Y. I. Suzuki and T. Suzuki, Phys. Chem. Chem. Phys., 2010, 12, 3653. 54 H. Y. Chen, S. C. N. Hsu and C. L. Kao, Phys. Chem. Chem. Phys., 2010, 12, 1253. 55 A. Melchior, J. M. Martı´nez, R. R. Pappalardo and E. S. Marcos, J. Chem. Theory Comput., 2013, 9, 4562. 56 A. D. Becke, J. Chem. Phys., 1993, 98, 5648. 57 C. Lee, W. Yang and R. G. Parr, Phys. Rev. B: Condens. Matter Mater. Phys., 1988, 37, 785.
This journal is © the Owner Societies 2014
View Article Online
Published on 01 August 2014. Downloaded by University of Newcastle on 12/09/2014 14:05:04.
Paper
´ and G. Pratviel, Chem. Rev., 2010, 110, 1018. 58 M. Pitie 59 Q. Zhang and Y. Wang, J. Am. Chem. Soc., 2003, 125, 12795. 60 Q. Zhang and Y. Wang, Nucleic Acids Res., 2005, 33, 1593. 61 I. S. Hong and M. M. Greenberg, J. Am. Chem. Soc., 2005, 127, 3692. 62 H. Ding, A. Majumdar, J. R. Tolman and M. M. Greenberg, J. Am. Chem. Soc., 2008, 130, 17981. 63 X. Peng, Y. Z. Pigli, P. A. Rice and M. M. Greenberg, J. Am. Chem. Soc., 2008, 130, 12890. 64 S. C. N. Hsu, T. P. Wang, C. L. Kao, H. F. Chen, P. Y. Yang and H. Y. Chen, J. Phys. Chem. B, 2013, 117, 2096.
This journal is © the Owner Societies 2014
PCCP
65 L. Schwartsburd, R. Cohen, L. Konstantinovski and D. Milstein, Angew. Chem., Int. Ed., 2008, 47, 3603. 66 M. H. Almatarneh, C. G. Flinn and R. A. Poirier, J. Chem. Inf. Model., 2008, 48, 831. 67 A. I. Alrawashdeh, M. H. Almatarneh and R. A. Poirier, Can. J. Chem., 2013, 91, 518. 68 A. L. Millen, L. A. B. Archibald, K. C. Hunter and S. D. Wetmore, J. Phys. Chem. B, 2007, 111, 3800. 69 S. A. P. Lenz, J. L. Kellie and S. D. Wetmore, J. Phys. Chem. B, 2012, 116, 14275. 70 X. Song, Y. Qian, R. Ben, X. Lu, H. L. Zhu, H. Chao and J. Zhao, J. Med. Chem., 2013, 56, 6531.
Phys. Chem. Chem. Phys., 2014, 16, 19290--19297 | 19297
Published on 01 August 2014. Downloaded by University of Newcastle on 12/09/2014 14:05:04.
PAPER
Cite this: Phys. Chem. Chem. Phys., 2014, 16, 19290
View Journal | View Issue
Interaction of electrons with cisplatin and the subsequent effect on DNA damage: a density functional theory study† Hsing-Yin Chen,* Hui-Fen Chen, Chai-Lin Kao, Po-Yu Yang and Sodio C. N. Hsu Cisplatin, Pt(NH3)2Cl2, is a leading chemotherapeutic agent that has been widely used for various cancers. Recent experiments show that combining cisplatin and electron sources can dramatically enhance DNA damage and the cell-killing rate and, therefore, is a promising way to overcome the side effects and the resistance of cisplatin. However, the molecular mechanisms underlying this phenomenon are not clear yet. By using density functional theory calculations, we confirm that cisplatin can efficiently capture the prehydrated electrons and then undergo dissociation. The first electron attachment triggers a spontaneous departure of the chloride ion, forming a T-shaped [Pt(NH3)2Cl] neutral radical, whereas the second electron attachment leads to a spontaneous departure of ammine, forming a linear [Pt(NH3)Cl] anion. We further recognize that the one-electron reduced product [Pt(NH3)2Cl] is extremely harmful to DNA. It can abstract hydrogen atoms from the C–H bonds of the ribose moiety and the methyl group of thymine, which in turn leads to DNA strand breaks and cross-link
Received 26th May 2014, Accepted 24th July 2014
lesions. The activation energies of these hydrogen abstraction reactions are relatively small compared to the hydrolysis of cisplatin, a prerequisite step in the normal mechanism of action of cisplatin. These
DOI: 10.1039/c4cp02306d
results rationalize the improved cytotoxicity of cisplatin by supplying electrons. Although the biological effects of the two-electron reduced product [Pt(NH3)Cl] are not clear at this stage, our calculations
www.rsc.org/pccp
indicate that it might be protonated by the surrounding water.
1. Introduction Cisplatin, Pt(NH3)2Cl2, is one of the most widely used anticancer drugs nowadays. It has been successfully applied to the treatment of various cancers such as testicular, ovarian, head/neck, and small-cell lung cancers.1–3 It is generally accepted that the target of cisplatin is DNA: it prefers binding to the N7 sites of two consecutive guanines, leading to intrastrand cross links and the subsequent cell apoptosis.1–10 Prior to binding with DNA, cisplatin has to undergo a hydrolysis reaction in which one or two chlorides are replaced by water molecules. Experimental measurements demonstrate that the hydrolysis of cisplatin is a slow process with an activation energy of 23–24 kcal mol1,11–13 consistent with the theoretical predictions.14–19 The bioactivity of cisplatin might also involve many other reactions such as Department of Medicinal and Applied Chemistry, Kaohsiung Medical University, Kaohsiung 807, Taiwan. E-mail: [email protected]; Fax: +886 7 3125339; Tel: +886 7 3121101 ext. 2807 † Electronic supplementary information (ESI) available: Calculations of 17 explicit water molecules plus the SMD bulk hydration model, B3LYP results, relative stability of sugar radicals, the temperature effect, the cavity effect, bond energies of nucleobases, Cartesian coordinates of optimized structures. See DOI: 10.1039/ c4cp02306d
19290 | Phys. Chem. Chem. Phys., 2014, 16, 19290--19297
tautomerization of nucleobases10,20 and interactions with nonDNA targets.21 Despite the success of cisplatin in the cancer therapy, it has two main disadvantages: severe side effects and intrinsic/ acquired resistance. To overcome these drawbacks, enormous efforts and resources have been invested in developing new Pt-based anticancer drugs in the past 35 years. To date, two new platinum drugs, carboplatin and oxaliplatin, have been approved by the US Food and Drug Administration (FDA).1–3 Compared with cisplatin, carboplatin has the advantage of lower toxicity and is thus safer, whereas oxaliplatin displays a broader spectrum of antitumour activity.3 In addition to developing new drugs, an alternative way is to use an old drug but modify its mechanism of action by certain chemical or physical approaches. Recently, a series of experiments have demonstrated that this strategy works.22–29 Zheng et al.22 and Xiao et al.25 examined the influences of cisplatin on the sensitization of DNA toward secondary low-energy electrons (LEEs). They observed that the yields of DNA strand breaks and glycosidic bond cleavages induced by LEEs were enhanced in the solid films of cisplatin-DNA adducts in comparison with the solid films consisting of DNA alone.22,23,25 A similar radiosensitization effect of cisplatin in the aqueous solution was
This journal is © the Owner Societies 2014
View Article Online
Published on 01 August 2014. Downloaded by University of Newcastle on 12/09/2014 14:05:04.
Paper
PCCP
reported by Rezaee et al. as well;24 this work demonstrates that the formation of cisplatin-DNA adducts boosts DNA strand breaks induced by both hydrated electrons and hydroxyl radicals, which are the two major species produced in the radiolysis of water. The hydroxyl radical plays a pivotal role in DNA damage and the related mechanisms have been extensively studied. It mainly undergoes two types of reactions as the first step toward DNA strand breaks: addition to C–C and C–N double bonds in nucleobases and hydrogen abstraction from the sugar moiety.30–32 In contrast, the effect of LEEs on DNA damage has long been neglected until 2000.33 Recently, there has been experimental evidence showing that LEEs might be more efficient than hydroxyl radicals in causing DNA strand breaks.34 Lu and coworkers investigated the interactions of cisplatin with electrons by time-resolved femtosecond laser spectroscopy. They discovered that cisplatin is highly reactive toward the dissociative electron transfer reaction with weakly bound electrons that may come from two-photon absorption of water,26 excitation of photosensitizers,27 or reducing agents.29 More importantly, they further showed that the combination of cisplatin and electron sources can significantly improve the yield of DNA strand breaks and the rate of cell killing, even for the cisplatin-resistant cell lines.27,29 Since no incubation time of cisplatin with DNA was needed to detect DNA strand breaks in Lu’s experiments,27,28 the enhancement of DNA damage must be a consequence of the interactions of electrons with free cisplatin rather than with cisplatin-DNA adduct. Here, we report a density functional theory study which aims to provide the molecular level understanding of Lu’s experimental findings. The dissociative behaviors of cisplatin upon one- and two-electron attachments were investigated. In addition, a number of possible reactions that cause DNA damage by the dissociative products were studied.
2. Computational methods All calculations were accomplished by the Gaussian 09 program.35 Hybrid meta functional M0636 was employed in the present study. This functional is recommended for dealing with the chemical reactions where both organic and transition-metal bonds are formed or broken.37 In addition, the capability of M06 functional to provide satisfactory description for noncovalent interactions has been verified as well.37 The 6-31+G(d) or 6-31++G(d,p) basis sets were used for the main group elements. The Pt atom was described by the Stuttgart–Dresden pseudopotential (SDD) and
Fig. 1
the corresponding basis set augmented by a set of polarization (af = 0.98) and diffuse functions (as = 0.0075, ap = 0.013, ad = 0.025).38 The combinations of the augmented SDD with the 6-31+G(d) and with the 6-31++G(d,p) are denoted by BS-1 and BS-2, respectively. The SMD continuum solvation model39 was employed to simulate the bulk hydration environment (e = 78.3553). Geometry optimizations and vibrational frequency calculations were carried out using the SMD solvation model. An ultrafine grid was used for numerical integrations. Free energy corrections were performed under the standard conditions of 298.15 K and 1 atm. Intrinsic reaction coordinate (IRC) calculations40 were performed for the transition state structures to confirm that they connect the corresponding reactants and products. It has been pointed out by Wertz41 and Abraham42 that all molecules lose the same fraction (B0.5) of their entropies when they are transferred from gas phase into aqueous solution. This entropy loss results from the fact that the free volume of a molecule in solution is reduced, and its translational, rotational, and probably parts of vibrational motions are subjected to the restriction imposed by the solvent cage. Continuum solvation models cannot provide a correct description of this effect; as a result, the entropic penalties and, thus, the activation and reaction free energies for multimolecular reactions in solutions are usually overestimated.43–46 To take into account the effect of restrained motions of solutes in solution, the free energies were evaluated by G = H 0.5TS. The empirical Wertz correction has been successfully applied to reproduce experimental measurements in many studies.17,47–49
3. Results and discussion 3.1
Dissociation of cisplatin by electron attachment
To explore the dissociative behavior of cisplatin in the aqueous environment, we employed a microhydrated structure of cisplatin including 15 explicit water molecules in the first hydration layer; this microhydrated model has been previously used to study the hydrolysis of cisplatin.19 We further improved the hydration model by immersing this supermolecule of cisplatin plus 15H2O in the bulk hydration environment described by the SMD continuum solvation model. The optimized structures of cisplatin and its one- and twoelectron adducts in the aqueous environment are shown in Fig. 1 and the relevant bond distances are presented in Table 1. For the neutral cisplatin, the optimized distances of Pt–Cl and
SMD/M06/BS-1 optimized structures of cisplatin and its one- and two-electron adducts.
This journal is © the Owner Societies 2014
Phys. Chem. Chem. Phys., 2014, 16, 19290--19297 | 19291
View Article Online
PCCP
Paper
Table 1 Bond distances (Å) of cisplatin and its one- and two-electron adducts optimized in the presence of 15 explicit water molecules plus a SMD bulk hydration environment
Pt(NH3)2Cl2 Pt–ClL a Pt–Cl Pt–NL a Pt–N
2.39 2.38 2.04 2.06
(2.39)b (2.37)b (2.04)b (2.05)b
[Pt(NH3)2Cl]
[Pt(NH3)Cl]
3.82 2.41 2.26 2.10
4.11 2.37 3.59 2.08
Published on 01 August 2014. Downloaded by University of Newcastle on 12/09/2014 14:05:04.
a
The subscript L denotes the leaving groups. b The values in parentheses are gas-phase optimized distances in the presence of 15 explicit water molecules (ref. 19).
Pt–N are 2.38, 2.39 and 2.04, 2.06 Å respectively. These values are very close to the previous gas-phase results incorporating 15 explicit H2O.19 Geometry optimizations revealed that when an extra electron was added to the neutral cisplatin one of the chlorides spontaneously left (the distance of Pt–ClL is 3.82 Å), resulting in the formation of a T-shaped [Pt(NH3)2Cl] neutral radical. It is worth noting that the distance between the Pt center and the NH3 opposite to the vacant site (i.e., Pt–NL) is substantially elongated to 2.26 Å. When the second electron was added to the [Pt(NH3)2Cl] , this ammonia immediately departed (the distance of Pt–NL is 3.59 Å), giving a linear [Pt(NH3)Cl] complex anion with the bond lengths of Pt–Cl and Pt–N being 2.37 and 2.08 Å, respectively. Notice that the leaving ammonia interacts with the Pt(0) center by a noncovalent NL–H Pt interaction (the distance of H Pt is 2.56 Å). If there exists a substantial barrier in the dissociative processes, geometry optimizations starting from Pt(NH3)2Cl2 + e and [Pt(NH3)2Cl] + e should, respectively, converge to the local minima corresponding to [Pt(NH3)2Cl2] and [Pt(NH3)2Cl], rather than directly converging to the dissociative products [Pt(NH3)2Cl] + Cl and [Pt(NH3)Cl] + NH3. However, this is not the truth, implying that the barriers, if exist, are small, and the cisplatin readily dissociates upon electron attachment. This result is consistent with the experimental observation that the transient absorption of cisplatin decays within a few tens of picoseconds after interacting with electrons.26 We also notice that the water molecules interact with the Pt center by their O–H bonds rather than by oxygen atoms. These O–H Pt interactions (blue dot line in Fig. 1) are characterized by the bond lengths considerably shorter than the sum of the van der Waals radii of hydrogen (1.20 Å) and platinum (1.75 Å), varying from 2.29 to 2.67 Å. In addition, the O–H Pt interactions display directionality with the bonding angles being in the range of 1571–1691. The number and strength (judging by bond length) of the O–H Pt interaction increase with the reduction state. This observation gives us a hint that the electron-rich Pt(0) center of [Pt(NH3)Cl] could be able to activate the polar bonds such as the O–H bond of water. In fact, the O–H Pt interaction has been recognized in many crystal structures and computational analyses have shown that the dispersion force is responsible for the formation of such an interaction.50,51 To assess the electron-accepting capability of cisplatin, the vertical electron affinities (EAv) of Pt(NH3)2Cl2 and [Pt(NH3)2Cl]
19292 | Phys. Chem. Chem. Phys., 2014, 16, 19290--19297
were calculated. The first and the second EAv values of cisplatin were evaluated to be 1.895 and 2.400 eV, respectively. Comparing these values with the electron binding energies of the prehydrated electron (1.0–1.5 eV)52 and the fully hydrated electron (3.3 eV)53 suggests that cisplatin can only be reduced by the former. It is remarkable that the EAv value of cisplatin is larger than that of the guanine–cytosine base pair (1.433 eV).54 Since cytosine is the most electron-affinic site in DNA, this result indicates that cisplatin has a higher ability to capture electrons than DNA. The hydration number for the first coordination shell of cisplatin might be higher as recently pointed out by a classical molecular dynamics simulations study.55 To test the influence of the number of explicit water molecules, we further augmented the solvation model by including two more water molecules around the ammine ligands (i.e., 17H2O + SMD model). In addition, in order to check the DFT functional dependence we also calculated the interaction of electrons with cisplatin by the B3LYP method.56,57 All these calculations revealed that while the optimized geometries are somewhat different, the main results and conclusions regarding the dissociation behavior, the characteristic interactions between water and platinum, and the electron affinity are not changed (Fig. S1 and S2 and Tables S1 and S2 in ESI†). 3.2
Reactivity of one-electron reduced product [Pt(NH3)2Cl]
3.2.1 DNA strand break by direct attack on the C3 0 –O bond. The next question is how these dissociative products damage DNA? According to the experimental results, the major forms of DNA damage induced by the simultaneous treatment of cisplatin and electrons are strand breaks, while the cross links are minor forms.27 We found that the [Pt(NH3)2Cl] can cause DNA strand break via direct oxidative addition to the C3 0 – O bond (red path in Fig. 2). According to the results of Wertzcorrected free energies, initially a stable reactant adduct (RA), which lies 2.3 kcal mol1 below the separated reactants (SR), can be formed via the hydrogen bonding between NH3 of [Pt(NH3)2Cl] and the O atom of the phosphate group. Then the Pt(I) center attacks the O atom connecting to the C3 0 (TS1), leading to a homolytic cleavage of the C30 –O bond and releasing the separated products of [Pt(NH3)2(HPO4)Cl] and C30 -sugarradical (SP1). Although this reaction is only slightly endergonic by 3.6 kcal mol1, it has a large energetic barrier of 33.6 kcal mol1. Without the Wertz correction, RA was predicted to be unstable (3.6 kcal mol1 higher than SR). If RA is not a stable complex the energetic barrier of the oxidative addition of [Pt(NH3)2Cl] to the C3 0 –O bond should be evaluated by the free-energy difference between the TS1 and the SR (rather than RA), and the resultant value becomes 38.2 kcal mol1, substantially larger than the prediction including Wertz corrections. However, the Wertz correction has a minor effect on the evaluation of the reaction free energy. This is because the molecular number of SP1 and SR is equal and, thus, the effect of Wertz corrections is largely cancelled out. Both results, no matter with or without Wertz corrections, suggest that the reaction of direct oxidative addition to the C–O
This journal is © the Owner Societies 2014
View Article Online
Published on 01 August 2014. Downloaded by University of Newcastle on 12/09/2014 14:05:04.
Paper
PCCP
Fig. 2 SMD/M06/BS-2 free energy profile (in kcal mol1) for C3 0 –O bond cleavage induced by [Pt(NH3)2Cl] . The values in parentheses are results without Wertz corrections.
bond is too slow to account for the observed high efficiency of cisplatin plus electrons in causing DNA strand breaks. 3.2.2 DNA strand break by H4 0 -abstraction. An alternative pathway leading to DNA strand break is the hydrogen abstraction from the ribose moiety (blue path in Fig. 2). Starting from the same reactant adduct, the Pt(I) center could attack and abstract the H40 atom (TS2), resulting in the formation of C40 -nucleotide-radical (1). The Wertz-corrected activation free energy of the H40 -abstraction is 16 kcal mol1 (20.3 kcal mol1 without Wertz corrections). The C30 – O bond of the radical 1 is significantly weakened and immediately undergoes a heterolytic cleavage to release [HPt(NH3)2Cl]HPO42 and sugar radical cation (2); the Wertz-corrected dissociation free energy of this bond was estimated to be only 14.2 kcal mol1 (7.1 kcal mol1 without Wertz corrections), considerably lower than 80–90 kcal mol1 for the typical C–O bonds. The unusual fragility of the C30 –O bond in radical 1 can be rationalized by a high stability of the resulting sugar radical cation due to the presence of three resonance forms (cf. chemical structures of 2 in Fig. 2).58 The [HPt(NH3)2Cl]HPO42 and the sugar radical cation might recombine to form intermediate 3 through the Pt C30 interaction. A rapid hydride transfer from Pt to C30 (TS3) subsequently takes place, producing the separated products of [Pt(NH3)2(HPO4)Cl] and C40 sugar-radical (SP2); the barrier of this step is only 8 kcal mol1. The overall reaction is almost thermodynamically neutral. The activation energies along this reaction pathway do not exceed 16 kcal mol1 (or 20 kcal mol1 without Wertz corrections) and might be responsible for the high yield of DNA strand breaks caused by cisplatin plus electrons. We would like to emphasize that the possibilities of hydrogen abstractions from other sites of sugar cannot be excluded.
This journal is © the Owner Societies 2014
The resultant sugar radicals have been shown to be highly reactive and will undergo various reactions that eventually lead to strand breaks.58 The reason we selected the 4 0 -hydrogen atom as the target in the present study is as follows. First, the H4 0 and the H5 0 are the two most accessible sites in B-form DNA. Secondly, the resultant C4 0 -based sugar radical is most stable (Table S3 in ESI†), indicating that the H4 0 -abstraction should be most favorable in energetics. This is because the other component of the products, namely HPt(NH3)2Cl, being the same, the reaction energies of the hydrogen abstractions are expected to be principally determined by the stability of the resultant sugar radicals. In addition, the transferred hydrogen is already far from the carbon center in the transition state (see TS2 in Fig. 2), implying that the sugar moiety in the transition state bears a substantial radical character. Accordingly, the reaction barriers are also expected to correlate with the stability of sugar radicals. In fact, we have calculated the reaction of H5 0 abstraction (Fig. 3), and the resultant barrier and reaction energy are indeed larger than that of H4 0 -abstraction, supporting the aforementioned inference. 3.2.3 H-abstraction from the methyl group of thymine. Experimental studies have shown that among four nucleobases only the methyl group of thymine can undergo the hydrogen abstraction reaction with hydroxyl radicals.30,31,58 This experimental result is in harmony with the calculations of bond energies; the C–H bond of the methyl group of thymine is characterized by a relatively small bond dissociation energy compared to other C–H and N–H bonds of nucleobases (Table S5 in ESI†). Accordingly, we investigated the reaction of [Pt(NH3)2Cl] to abstract the hydrogen atom from the methyl group of thymine (Fig. 4).
Phys. Chem. Chem. Phys., 2014, 16, 19290--19297 | 19293
View Article Online
Published on 01 August 2014. Downloaded by University of Newcastle on 12/09/2014 14:05:04.
PCCP
Fig. 3 SMD/M06/BS-2 free energy profile (in kcal mol1) for H5 0 abstraction from nucleotide by [Pt(NH3)2Cl] . The values in parentheses are results without Wertz corrections.
The Wertz-corrected results reveal that the [Pt(NH3)2Cl] can attach to thymine through the hydrogen bond between NH3 of [Pt(NH3)2Cl] and O4 of thymine; the RA is more stable than the SR by 2.1 kcal mol1. Then the Pt(I) center approaches and abstracts the hydrogen atom of the methyl group with an activation energy of 19.2 kcal mol1. Without the Wertz corrections, the complex of [Pt(NH3)2Cl] and thymine was predicted to be instable, and the activation energy was estimated to be 23.4 kcal mol1. Although this H-abstraction reaction is endergonic by 11 kcal mol1, it is possible that the reaction could be further driven by thermodynamically favorable reactions of the methyl radical of thymine. For example, there has been
Fig. 4 SMD/M06/BS-2 free energy profile (in kcal mol1) for H-abstraction from the methyl group of thymine by [Pt(NH3)2Cl] . The values in parentheses are results without Wertz corrections.
19294 | Phys. Chem. Chem. Phys., 2014, 16, 19290--19297
Paper
experimental evidence showing that the methyl radical of thymine can form covalent bonds with adjacent nucleobases, leading to intrastrand59,60 and interstrand61–63 cross-link lesions. Note that the activation energy of H-abstraction from the methyl group of thymine is higher than that from the sugar moiety. This result might partially explain why the major forms of DNA damage caused by cisplatin plus electrons are strand breaks rather than cross links. Considering the steric factor, the methyl group of thymine is exposed to the major groove of DNA and is not sterically hindered by neighboring bases and, therefore, can be accessed by [Pt(NH3)2Cl] . 3.2.4 Binding to the N7 site of guanine and adenine. It is well-known that cisplatin, upon hydrolysis, is highly inclined to bind with the N7 position of purines.1–10 To understand whether or not the reduced cisplatin still has this characteristic property, the binding free energies of [Pt(NH3)2Cl] with guanine–cytosine (GC) and adenine–thymine (AT) base pairs were calculated. The optimized structures are shown in Fig. 5. The calculations reveal that the high affinity of cisplatin for the N7 position of purines is almost vanished upon reduction. The binding free energies of [Pt(NH3)2Cl] with GC and AT are only 0.5 and 1.7 kcal mol1, respectively, even lower than 2.3 kcal mol1 of [Pt(NH3)2Cl] and the phosphate backbone (RA in Fig. 2). This rationalizes why the [Pt(NH3)2Cl] would not bind to purine bases but instead attack the DNA backbone and cause strand breaks. For comparison, the binding free energies of the hydrolysis product [Pt(NH3)2Cl]+ with GC and AT were calculated as well, and the values are, respectively, 37.2 and 35.6 kcal mol1, supporting the high affinity of cisplatin for purines. 3.3
Reactivity of two-electron reduced product [Pt(NH3)Cl]
With respect to the reactivity of the two-electron reduced product [Pt(NH3)Cl], we did not consider the possibility of reaction with DNA due to the negatively charged nature of the
Fig. 5 SMD/M06/BS-2 optimized structures of (a) [Pt(NH3)2Cl]+GC (b) [Pt(NH3)2Cl] GC (c) [Pt(NH3)2Cl]+AT (d) [Pt(NH3)2Cl] AT.
This journal is © the Owner Societies 2014
View Article Online
Published on 01 August 2014. Downloaded by University of Newcastle on 12/09/2014 14:05:04.
Paper
PCCP
Fig. 6 SMD/M06/BS-2 free energy profile (in kcal mol1) for protonation of [Pt(NH3)Cl] by water. The values in parentheses are results without Wertz corrections.
DNA backbone. Instead, we explored the reaction of [Pt(NH3)Cl] with water, inspired by the observation of strong O–H Pt interactions. The calculations reveal that the electron-rich Pt(0) center of [Pt(NH3)Cl] can be protonated by water via a concerted double proton transfer involving the water molecules in the first and second hydration layers; the activation and reaction energies are 14.6 and 13.1 kcal mol1, respectively (Fig. 6). Through this mechanism the resultant OH anion can be conveyed to the second hydration layer and fully stabilized by three water molecules, which has the advantage in lowering the reaction energy.64 A similar mechanism has been observed in the protonation of nucleobase anions by water.64 The example of the Pt(0) complex able to activate the O–H bond of water has been previously reported.65 The OH anion is a strong base and it can damage DNA by deprotonating nucleobases. Computational studies have shown that deprotonation of amine groups of adenine and cytosine by OH is a highly exothermic and barrier-free reaction.66,67 In addition, OH may also act as a nucleophile to attack the C1 0 site of sugar and lead to glycosidic bond cleavage through the SN2 mechanism.68,69 On the other hand, it is not clear yet whether the platinum hydride plays a role in DNA damage. While myriads of studies on metal hydride complexes have been appeared, most of these studies address the catalytic issue. After literature searching, we only found one paper reporting that the iridium hydrides can interact with DNA and induce cell apoptosis and, therefore, exhibit antitumor activity.70 More experimental and computational studies are needed to shed light on the biological activities of metal hydrides.
4. Conclusions The mechanism of action of combining cisplatin and electron sources is distinct from that of treating with cisplatin alone. Our DFT study demonstrates that the one-electron reduced product of cisplatin, [Pt(NH3)2Cl] , loses high affinity for binding with purines but damages DNA by abstracting hydrogens from ribose or methyl
This journal is © the Owner Societies 2014
Fig. 7 SMD/M06/BS-2 simulated IR spectrum of HPt(NH3)2Cl, cisplatin, and 2 0 -deoxycitidine-3 0 -monophosphate (3 0 -dCMP).
groups of thymine instead. The activation energies of these H-abstraction reactions (16 and 19 kcal mol1) are considerably lower than that of the hydrolysis of cisplatin (ca. 24 kcal mol1), a prerequisite step in the normal mechanism of action, explaining why the combination of cisplatin and electron sources can significantly enhance DNA damage and the cell-killing rate. The mechanisms proposed in the present study remain to be verified by experiments. A feasible experimental test is to detect the formation of the platinum hydride complex in the H-abstraction reactions. This could be achieved by monitoring the characteristic IR absorption of the platinum hydride complex. The vibrational frequency analyses show that the HPt(NH3)2Cl complex displays an intense absorption at 2155 cm1, which corresponds to the stretching of the Pt–H bond (Fig. 7). As cisplatin and DNA do not absorb at 2000–3000 cm1, the unique absorption of the Pt–H bond might be exploited to probe the H-abstraction reactions from DNA by [Pt(NH3)2Cl] .
Acknowledgements We thank the Ministry of Science and Technology of Taiwan for financial support and the National Center for High-performance Computing for computer time and facilities.
References 1 D. Wang and S. J. Lippard, Nat. Rev. Drug Discovery, 2005, 4, 307.
Phys. Chem. Chem. Phys., 2014, 16, 19290--19297 | 19295
View Article Online
Published on 01 August 2014. Downloaded by University of Newcastle on 12/09/2014 14:05:04.
PCCP
2 R. A. Alderden, M. D. Hall and T. W. Hambley, J. Chem. Educ., 2006, 83, 728. 3 L. Kelland, Nat. Rev. Cancer, 2007, 7, 573. 4 E. R. Jamieson and S. J. Lippard, Chem. Rev., 1999, 99, 2467. 5 J. Reedijk, Chem. Rev., 1999, 99, 2499. ´rez, Chem. Rev., 2003, 6 M. A. Fuertes, C. Alonso and J. M. Pe 103, 645. 7 Y. Jung and S. J. Lippard, Chem. Rev., 2007, 107, 1387. 8 M. H. Baik, R. A. Friesner and S. J. Lippard, J. Am. Chem. Soc., 2003, 125, 14082. 9 A. Robertazzi and J. A. Platts, Chem. – Eur. J., 2006, 12, 5747. 10 T. Matsui, Y. Shigeta and K. Hirao, J. Phys. Chem. B, 2007, 111, 1176. 11 J. R. Perumareddi and A. W. Adamson, J. Phys. Chem., 1968, 72, 414. 12 J. W. Reishus and D. S. Martin, J. Am. Chem. Soc., 1961, 83, 2457. 13 R. N. Bose, R. E. Viola and R. D. Cornelius, J. Am. Chem. Soc., 1984, 106, 3336. 14 P. Carloni, M. Sprik and W. Andreoni, J. Phys. Chem. B, 2000, 104, 823. 15 Y. Zhang, Z. Guo and X. Z. You, J. Am. Chem. Soc., 2001, 123, 9378. 16 J. V. Burda, M. Zeizinger and J. Leszczynski, J. Comput. Chem., 2005, 26, 907. 17 J. K. C. Lau and D. V. Deubel, J. Chem. Theory Comput., 2006, 2, 103. 18 J. K. C. Lau and B. Ensing, Phys. Chem. Chem. Phys., 2010, 12, 10348. 19 A. Melchior, E. S. Marcos, R. R. Pappalardo and J. M. Martı´nez, Theor. Chem. Acc., 2011, 128, 627. ´n-Carrasco, D. Jacquemin and E. Caue ¨t, Phys. 20 J. P. Cero Chem. Chem. Phys., 2012, 14, 12457. 21 V. Cepeda, M. A. Fuertes, J. Castilla, C. Alonso, C. Quevedo ´rez, Anti-Cancer Agents Med. Chem., 2007, 7, 3. and J. M. Pe 22 Y. Zheng, D. J. Hunting, P. Ayotte and L. Sanche, Phys. Rev. Lett., 2008, 100, 198101. 23 L. Sanche, Chem. Phys. Lett., 2009, 474, 1. 24 M. Rezaee, L. Sanche and D. J. Hunting, Radiat. Res., 2013, 179, 323. 25 F. Xiao, X. Luo, X. Fu and Y. Zheng, J. Phys. Chem. B, 2013, 117, 4893. 26 Q. B. Lu, S. Kalantari and C. R. Wang, Mol. Pharm., 2007, 4, 624. 27 Q. B. Lu, J. Med. Chem., 2007, 50, 2601. 28 Q. B. Lu, Mutat. Res., Rev. Mutat. Res., 2010, 704, 190. 29 T. Luo, J. Yu, J. Nguyen, C. R. Wang, R. G. Bristow, D. A. Jaffray, X. Z. Zhou, K. P. Lu and Q. B. Lu, Proc. Natl. Acad. Sci. U. S. A., 2012, 109, 10175. 30 C. J. Burrows and J. G. Muller, Chem. Rev., 1998, 98, 1109. 31 C. von Sonntag, Free-Radical-Induced DNA Damage and Its Repair, A Chemical Perspective, Springer-Verlag, Berlin, Heidelberg, New York, 2006. 32 R. B. Zhang and L. A. Eriksson, J. Phys. Chem. B, 2007, 111, 6571. 33 B. Boudaı¨ffa, P. Cloutier, D. Hunting, M. A. Huels and L. Sanche, Science, 2000, 287, 1658.
19296 | Phys. Chem. Chem. Phys., 2014, 16, 19290--19297
Paper
34 J. Nguyen, Y. Ma, T. Luo, R. G. Bristow, D. A. Jaffray and Q. B. Lu, Proc. Natl. Acad. Sci. U. S. A., 2011, 108, 11778. 35 M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, G. Scalmani, V. Barone, B. Mennucci, G. A. Petersson, H. Nakatsuji, M. Caricato, X. Li, H. P. Hratchian, A. F. Izmaylov, J. Bloino, G. Zheng, J. L. Sonnenberg, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, T. Vreven, J. A. Montgomery, Jr., J. E. Peralta, F. Ogliaro, M. Bearpark, J. J. Heyd, E. Brothers, K. N. Kudin, V. N. Staroverov, R. Kobayashi, J. Normand, K. Raghavachari, A. Rendell, J. C. Burant, S. S. Iyengar, J. Tomasi, M. Cossi, N. Rega, J. M. Millam, M. Klene, J. E. Knox, J. B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R. E. Stratmann, O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski, R. L. Martin, K. Morokuma, V. G. Zakrzewski, G. A. Voth, P. Salvador, J. J. Dannenberg, S. Dapprich, A. D. Daniels, O. Farkas, J. B. Foresman, J. V. Ortiz, J. Cioslowski and D. J. Fox, Gaussian, 09, Revision A.02, Gaussian, Inc., Wallingford CT, 2009. 36 Y. Zhao and D. G. Truhlar, Theor. Chem. Acc., 2008, 120, 215. 37 Y. Zhao and D. G. Truhlar, Acc. Chem. Res., 2008, 41, 157. 38 J. V. Burda, M. Zeizinger, J. Sponer and J. Leszczynski, J. Chem. Phys., 2000, 113, 2224. 39 A. V. Marenich, C. J. Cramer and D. G. Truhlar, J. Phys. Chem. B, 2009, 113, 6378. 40 K. Fukui, J. Phys. Chem., 1970, 74, 4161. 41 D. H. Wertz, J. Am. Chem. Soc., 1980, 102, 5316. 42 M. H. Abraham, J. Am. Chem. Soc., 1981, 103, 6742. 43 Y. Liang, S. Liu, Y. Xia, Y. Li and Z. X. Yu, Chem. – Eur. J., 2008, 14, 4361. 44 F. Huang, G. Lu, L. Zhao, H. Li and Z. X. Wang, J. Am. Chem. Soc., 2010, 132, 12388. ´n and A. Warshel, J. Phys. Chem. B, 1998, 102, 719. 45 J. Floria 46 Z. X. Yu and K. N. Houk, J. Am. Chem. Soc., 2003, 125, 13825. 47 J. Cooper and T. Ziegler, Inorg. Chem., 2002, 41, 6614. 48 J. Kua, S. W. Hanley and D. O. De Haan, J. Phys. Chem. A, 2008, 112, 66. 49 Y. Lattach, P. Archirel and S. Remita, J. Phys. Chem. B, 2012, 116, 1467. `s, R. Attias and J. Fraitag, Angew. Chem., 50 J. Kozelka, J. Berge Int. Ed., 2000, 39, 198. `s, J. Caillet, J. Langlet and J. Kozelka, Chem. Phys. 51 J. Berge Lett., 2001, 344, 573. 52 C. R. Wang, T. Luo and Q. B. Lu, Phys. Chem. Chem. Phys., 2008, 10, 4463. 53 Y. Tang, H. Shen, K. Sekiguchi, N. Kurahashi, T. Mizuno, Y. I. Suzuki and T. Suzuki, Phys. Chem. Chem. Phys., 2010, 12, 3653. 54 H. Y. Chen, S. C. N. Hsu and C. L. Kao, Phys. Chem. Chem. Phys., 2010, 12, 1253. 55 A. Melchior, J. M. Martı´nez, R. R. Pappalardo and E. S. Marcos, J. Chem. Theory Comput., 2013, 9, 4562. 56 A. D. Becke, J. Chem. Phys., 1993, 98, 5648. 57 C. Lee, W. Yang and R. G. Parr, Phys. Rev. B: Condens. Matter Mater. Phys., 1988, 37, 785.
This journal is © the Owner Societies 2014
View Article Online
Published on 01 August 2014. Downloaded by University of Newcastle on 12/09/2014 14:05:04.
Paper
´ and G. Pratviel, Chem. Rev., 2010, 110, 1018. 58 M. Pitie 59 Q. Zhang and Y. Wang, J. Am. Chem. Soc., 2003, 125, 12795. 60 Q. Zhang and Y. Wang, Nucleic Acids Res., 2005, 33, 1593. 61 I. S. Hong and M. M. Greenberg, J. Am. Chem. Soc., 2005, 127, 3692. 62 H. Ding, A. Majumdar, J. R. Tolman and M. M. Greenberg, J. Am. Chem. Soc., 2008, 130, 17981. 63 X. Peng, Y. Z. Pigli, P. A. Rice and M. M. Greenberg, J. Am. Chem. Soc., 2008, 130, 12890. 64 S. C. N. Hsu, T. P. Wang, C. L. Kao, H. F. Chen, P. Y. Yang and H. Y. Chen, J. Phys. Chem. B, 2013, 117, 2096.
This journal is © the Owner Societies 2014
PCCP
65 L. Schwartsburd, R. Cohen, L. Konstantinovski and D. Milstein, Angew. Chem., Int. Ed., 2008, 47, 3603. 66 M. H. Almatarneh, C. G. Flinn and R. A. Poirier, J. Chem. Inf. Model., 2008, 48, 831. 67 A. I. Alrawashdeh, M. H. Almatarneh and R. A. Poirier, Can. J. Chem., 2013, 91, 518. 68 A. L. Millen, L. A. B. Archibald, K. C. Hunter and S. D. Wetmore, J. Phys. Chem. B, 2007, 111, 3800. 69 S. A. P. Lenz, J. L. Kellie and S. D. Wetmore, J. Phys. Chem. B, 2012, 116, 14275. 70 X. Song, Y. Qian, R. Ben, X. Lu, H. L. Zhu, H. Chao and J. Zhao, J. Med. Chem., 2013, 56, 6531.
Phys. Chem. Chem. Phys., 2014, 16, 19290--19297 | 19297
