Unrestrained 5 ns molecular dynamics simulation of a cisplatin-DNA 1,2-GG adduct provides a rationale for the NMR features and reveals increased conformational flexibility at the platinum binding site

How can the cisplatin analogs with different amine act on DNA during cancer treatment theoretically?

Cisplatin is a widely used anti-cancer drug which inhibits the replication and polymerization of DNA molecule while showing some side effects and drug resistance. For this reason, to enhance its therapeutic index, researchers have synthesized several thousand analogs and tested their properties. In this project, several cisplatin analogs were designed to theoretically study the biological activity and lipophilicity effects on amine changes. The amines of the cisplatin molecule were substituted with aliphatic amines in different analogs. Computational methods such as molecular dynamics simulation, molecular docking, and molecular mechanics Poisson-Boltzmann surface area analysis were performed to investigate the binding of six cisplatin derivatives with DNA. The binding affinity and potential interactions of these drugs with double-strand DNA were analyzed. The stability effect of these drugs was investigated via root-mean-square deviation and root-mean-square fluctuation analysis, which showed that some analogs can break base-pair interaction at the end of DNA and reduced the stability of DNA. Also, the results revealed that the hydrogen bond is one of the most important factors in the binding of cisplatin's adduct to DNA. Molecular mechanics Poisson-Boltzmann surface area analysis indicated that electrostatic and van der Waals interactions are the most important deriving forces to the binding of cisplatin's drug to DNA. Finally, data revealed that cisplatin and the cis-dichloro-dimethylamine-platin tendency for binding to DNA are greater than that of other analogs.

Molecular Dynamics Simulations of Dna and Its Complexes

This article describes how classical molecular simulation methods are being used to gain a molecular-level understanding of the interaction mechanisms responsible for DNA–ligand recognition, and that govern the response of DNA to ligand binding. Case studies using a variety of different ligands—including small pharmaceutical drugs, proteins and lipids—are used to illustrate the power of modern molecular dynamics simulation methods for understanding how we may control the function and structure of DNA.

Conformational rearrangement of 1,2-d(GG) intrastrand cis-diammineplatinum crosslinked DNA is driven by counter-ion penetration within the minor groove of the modified site

The major structural aberrations of DNA induced by a cis-diammineplatinum (II) 1,2-d(GG) intrastrand cross-link (CPT) have been known for decades. To gain deeper insights into the structural dynamics of the sequence-dependent DNA distortions adjacent to the CPT adduct, we employed molecular modeling and molecular dynamics (MD) simulations. The structural dynamics of native (N-DNA) and cisPt 1,2-d(GG) crosslinked (CPT-DNA) in the form of symmetric 36 nt d(G₂T₁₅G*G*T₁₅G₂)●C₂A₁₅CCA₁₅C₂) oligonucleotide duplexes is compared. The selected sequence context enabled tracking of the origin of the DNA axis curvature at the YpR flexible points (N-DNA), the enhancement of axis bending, and further distortions due to steric/electrostatic perturbations arising from the CPT-crosslink. In addition to the known structural distortions of CPT-DNA: helix bend towards the major groove; local helix unwinding; high roll angle between cross-linked guanine bases; and adoption of A-form DNA on the 5'-side of the CPT-crosslink (TpG junction); our results show the existence of a singular irreversible and reproducible conformational rearrangement, not previously observed, resulting in two stable CPT-DNA1 and CPT-DNA2 conformers. The CPT-DNA2 conformation presents an enhanced DNA axis bend and a wider and shallower minor grove with increased solvent accessibility within the modified site. It is concluded that the polymorphous (unstable) DNA environment near the cisPt 1,2-d(GG) unit in synergy with specific dynamic events, such as prolonged minor groove retention of particular Na⁺ ions and water redistribution within the d(TG*G*T) site, together with the formation of extra and more stable H-bonds between Pt(NH₃)₂ amines and neighboring nucleotides, are cooperatively responsible for the initiation of the conformational rearrangement leading to the CPT-DNA2 conformer, which, surprisingly, closely resembles the HMGB1-bound CPT-DNA structure. Graphical abstract Superimposed averaged structures of normal (N-DNA, green) and cisplatin intrastrand cross-linked (CPT-DNA, orange). Global DNA axes: N-DNA (blue beads); CPT-DNA (red beads); PT (yellow sphere); crosslinked dGs viewed from the minor groove (bold).

Molecular dynamics simulation and free energy analysis of the interaction of platinum-based anti-cancer drugs with DNA

Cisplatin and oxaliplatin are two widely-used anti-cancer drugs which covalently bind to a same location in DNA strands. Platinum agents make intrastrand and interstrand cross-links with the N7 atoms of guanine nucleotides which prevent DNA from polymerization by causing a distortion in the double helix. Molecular dynamics simulations and free energy calculations were carried out to investigate the binding of two platinum-based anti-cancer drugs with DNA. We compared the binding of these drugs which differ in their carrier ligands, and hence their potential interactions with DNA. When a platinum agent binds to nucleotides, it causes a high amount of deformation in DNA structure. To find the extent of deformation, torsion angles and base pair and groove parameters of DNA were considered. These parameters were compared with normal B-DNA which was considered as the undamaged DNA. The formation of hydrogen bonds between drugs and DNA nucleotides was examined in solution. It was shown that oxaliplatin forms more hydrogen bonds than cisplatin. Our results confirm that the structure of the platinated DNA rearranges significantly and cisplatin tries to deform DNA more than oxaliplatin. The binding free energies were also investigated to understand the affinities, types and the contributions of interactions between drugs and DNA. It was concluded that oxaliplatin tendency for binding to DNA is more than cisplatin in solvent environment. The binding free energy was calculated based on the MM/PBSA and MM/GBSA methods and the results of QM/MM calculations verified them.

3D-RISM-MP2 Approach to Hydration Structure of Pt(II) and Pd(II) Complexes: Unusual H-Ahead Mode vs Usual O-Ahead One

Solvation of transition metal complexes with water has been one of the fundamental topics in physical and coordination chemistry. In particular, Pt(II) complexes have recently attracted considerable interest for their relation to anticancer activity in cisplatin and its analogues, yet the interaction of the water molecule and the metal center has been obscured. The challenge from a theoretical perspective remains that both the microscopic solvation effect and the dynamical electron correlation (DEC) effect have to be treated simultaneously in a reasonable manner. In this work we derive the analytical gradient for the three-dimensional reference interaction site model Møller-Plesset second order (3D-RISM-MP2) free energy. On the basis of the three-regions 3D-RISM self-consistent field (SCF) method recently proposed by us, we apply a new layer of the Z-vector method to the CP-RISM equation as well as point-charge approximation to the derivatives with respect to the density matrix elements in the RISM-CPHF equation to remarkably reduce the computational cost. This method is applied to study the interaction of H2O with the d(8) square planar transition metal complexes in aqueous solution, trans-[Pt(II)Cl2(NH3)(glycine)] (1a), [Pt(II)(NH3)4](2+) (1b), [Pt(II)(CN)4](2-) (1c), and their Pd(II) analogues 2a, 2b, and 2c, respectively, to elucidate whether the usual H2O interaction through O atom (O-ahead mode) or unusual one through H atom (H-ahead mode) is stable in these complexes. We find that the interaction energy of the coordinating water and the transition metal complex changes little when switching from gas to aqueous phase, but the solvation free energy differs remarkably between the two interaction modes, thereby affecting the relative stability of the H-ahead and O-ahead modes. Particularly, in contrast to the expectation that the O-ahead mode is preferred due to the presence of positive charges in 1b, the H-ahead mode is also found to be more stable. The O-ahead mode is found to be more stable than the H-ahead one only in 2b. The energy decomposition analysis (EDA) at the 3D-RISM-MP2 level revealed that the O-ahead mode is stabilized by the electrostatic (ES) interaction, whereas the H-ahead one is mainly stabilized by the DEC effect. The ES interaction is also responsible for the difference between the Pd(II) and Pt(II) complexes; because the electrostatic potential is more negative along the z-axis in the Pt(II) complex than in the Pd(II) one, the O-ahead mode prefers the Pd(II) complexes, whereas the H-ahead becomes predominant in the Pt(II) complexes.

Atomic level rendering of DNA-drug encounter

Computer simulations have been demonstrated to be important for unraveling atomic mechanisms in biological systems. In this study, we show how combining unbiased molecular dynamic simulations with appropriate analysis tools can successfully describe metal-based drug interactions with DNA. To elucidate the noncovalent affinity of cisplatin's family to DNA, we performed extensive all-atom molecular dynamics simulations (3.7 μs total simulation length). The results show that the parent drug, cisplatin, has less affinity to form noncovalent adducts in the major groove than its aquo complexes. Furthermore, the relative position in which the drugs enter the major groove is dependent on the compound's net charge. Based on the simulations, we estimated noncovalent binding free energies through the use of Markov state models. In addition, and to overcome the lack of experimental information, we employed two additional methods: Molecular Mechanics Poisson-Boltzmann Surface Area (MMPB-SA) and steered molecular dynamics with the Jarzynski estimator, with an overall good agreement between the three methods. All complexes show interaction energies below 3 kcal/mol with DNA but the charged hydrolysis products have slightly more favorable binding free energies than the parent drug. Moreover, this study sets the precedent for future unbiased DNA-ligand simulations of more complex binders.

Non-specificity and synergy at the binding site of the carboplatin-induced DNA adduct via molecular dynamics simulations of the MutSα-DNA recognition complex

MutSα is the most abundant mismatch-binding factor of human DNA mismatch repair (MMR) proteins. MMR maintains genetic stability by recognizing and repairing DNA defects. Failure to accomplish their function may lead to cancer. In addition, MutSα recognizes at least some types of DNA damage making it a target for anticancer agents. Here, complementing scarce experimental data, we report unique hydrogen-bonding motifs associated with the recognition of the carboplatin induced DNA damage by MutSα. These data predict that carboplatin and cisplatin induced damaging DNA adducts are recognized by MutSα in a similar manner. Our simulations also indicate that loss of base pairing at the damage site results in (1) non-specific binding and (2) changes in the atomic flexibility at the lesion site and beyond. To further quantify alterations at MutSα-DNA interface in response to damage recognition, non-bonding interactions and salt bridges were investigated. These data indicate (1) possible different packing and (2) disruption of the salt bridges at the MutSα-DNA interface in the damaged complex. These findings (1) underscore the general observation of disruptions at the MutSα-DNA interface and (2) highlight the nature of the anticancer effect of the carboplatin agent. The analysis was carried out from atomistic simulations.

Mechanism of the cis-[Pt(1R,2R-DACH)(H2O)2]2+ intrastrand binding to the double-stranded (pGpG)·(CpC) dinucleotide in aqueous solution: a computational DFT study

A mechanism of the intrastrand 1,2-cross-link formation between the double-stranded pGpG·CpC dinucleotide (ds(pGpG)) and fully aquated oxaliplatin cis-[Pt(DACH)(H2O)2](2+) (DACH = cyclohexane-1R,2R-diamine) is presented. All structures of the reaction pathways including the transition states (TSs) were fully optimized in water solvent using DFT methodology with dispersion corrections. Both 5' → 3' and 3' → 5' binding directions were considered. In the first step there is a slight kinetic preference for 5'-guanine (5'G) monoadduct formation with an activation Gibbs free energy of 18.7 kcal/mol since the N7 center of the 5'G base is fully exposed to the solvent. On the other hand, the N7 atom of 3'-guanine (3'G) is sterically shielded by 5'G. The lowest energy path for formation of the 3'G monoadduct with an activation barrier of 19.3 kcal/mol is connected with a disruption of the 'DNA-like' structure of ds(pGpG). Monoadduct formation is the rate-determining process. The second step, chelate formation, is kinetically preferred in the 3' → 5' direction. The whole process of the platination is exergonic by up to -18.8 kcal/mol. Structural changes of ds(pGpG), charge transfer effects, and the influence of platination on the G·C base pair interaction strengths are also discussed in detail.

Cisplatin GG-crosslinks within single-stranded DNA: origin of the preference for left-handed helicity

Molecular dynamics (MD) simulations of the single-stranded DNA trinucleotide TG*G*, with the G* guanines crosslinked by the antitumor drug cisplatin, were performed with explicit representation of the water as solvent. The purpose of the simulations was to explain previous NMR observations indicating that in single-stranded cisplatin-DNA adducts, the crosslinked guanines adopt a left-handed helical orientation, whereas in duplexes, the orientation is right-handed. The analysis of the MD trajectory of TG*G* has ascribed a crucial role to hydrogen-bonding (direct or through-water) interactions of the 5'-oriented NH(3) ligand of platinum with acceptor groups at the 5'-side of the crosslink, namely the TpG* phosphate and the terminal 5'-OH group. These interactions bring about some strain into the trinucleotide which is slightly but significantly (1-1.5 kcal.mol(-1)) higher for the right-handed orientation than for the left-handed one. During the unconstrained, 3 ns long MD simulation, left-handed conformations were ~15 times more abundant than the right-handed ones. This sampling difference agrees roughly with the calculated energy difference in strain energy. Overall, these results show that the Pt-GG crosslink within single-stranded DNA is malleable and can access different conformations at a moderate energy cost. This malleability could be of importance in interactions between the platinated DNA and cellular proteins, in which the DNA is locally unwound.

Metal ion-promoted conformational changes of oligonucleotides

The review will discuss the influence of metal ions on conformational changes of oligonucleotides. First, a short definition of the torsion angles is given, followed by a concise yet critical overview of the commonly applied experimental techniques. Finally, the possible role of metals upon the following conformational changes of oligonucleotides is discussed: (i) the denaturation of double-strands, (ii) the transition from B- to A-DNA, (iii) the transition from right- to left-handed DNA and RNA, (iv) the condensation, (v) and other conformational changes. We conclude with a summary and outlook.

Platinum-DNA interstrand crosslinks: molecular determinants of bending and unwinding of the double helix

Platinum diamine complexes are able to crosslink the guanines of d(GC)(2) dinucleotides within double-stranded DNA. The interstrand crosslink thus formed causes a bend of the double helix toward the minor groove and the helical sense changes locally to left-handed, resulting in a considerable unwinding. The bend and unwinding angles have been shown to depend on the platinum ligands. Here, we have used molecular dynamics simulations to investigate the DNA 20-mer d(C(1)T(2)C(3)T(4)C(5)C(6)T(7)T(8)G*(9)C(10)T(11)C(12)T(13)C(14)C(15)T(16)T(17)C(18)T(19)C(20))-d(G(21)A(22)G(23)A(24)A(25)G(26)G(27)A(28)G(29)A(30)G*(31)C(32)A(33)A(34)G(35)G(36)A(37)G(38)A(39)G(40)) with the G* guanines crosslinked by cis-Pt(NH(3))(2)(2+), Pt(R,R-DACH)(2+), or Pt(S,S-DACH)(2+). Previous investigations on cisplatin interstrand adducts indicated that the structure is similar in solid state and in solution; thus, we used the reported X-ray structure of a cisplatin adduct as a starting model. Replacing in the MD-relaxed model for the DNA duplex crosslinked with cis-Pt(NH(3))(2)(2+) the two NH(3) platinum ligands by R,R-DACH or S,S-DACH led to clashes between the DACH residue and the deoxyribose of C(12). Confrontation of MD-derived models with gel shift measurements suggested that these clashes are avoided differently in the adducts of Pt(R,R-DACH)(2+)versus Pt(S,S-DACH)(2+). The R,R-isomer avoids the clash by untwisting the T(11)/A(30)-C(12)/G(29) step, thus increasing the global unwinding. In contrast, the S,S-isomer modifies the shift and slide parameters of this step, which dislocates the helical axis and enhances the bend angle. The clash that leads to the differentiation of the structures as a function of the diamine ligand is related to a hydrogen bond between the platinum complex and the T(11) base and could be characteristic of interstrand crosslinks at d(pyG*Cpy)-d(puG*Cpu) sequences.

Cisplatin interaction with amino acids cysteine and methionine from gas phase to solutions with constant pH

This work is focused on the computational studies of reactions of hydrated forms of cisplatin with sulphur-containing amino acids cysteine and methionine. First, the appropriate model for solvation of the examined complexes was searched for. The suggested procedure employs the B3LYP density functional, 6-311++G(2df,2pd) basis set with Stuttgart-Dresden pseudopotentials on heavy atoms, the D-PCM solvation model and the UAKS cavity which uses more realistic NPA partial charges instead of formal partial charges for platinum ligands. In the second part this model is applied to the evaluation of the Legendre transformed reaction Gibbs free energy of cisplatin with cysteine and methionine in solution at constant pH.

Investigation of transcription factor Ndt80 affinity differences for wild type and mutant DNA: a molecular dynamics study

Molecular dynamics simulations and free energy calculations have been performed on the transcription factor Ndt80 either in complex with the native DNA sequence or with a mutant DNA with a switched central base pair, C5-G5' to G5-C5'. This mutant has been shown to have a 100-fold decrease in binding affinity of Ndt80, and in this study we explain this both structurally and energetically. The major interactions between the protein and the DNA were maintained in the simulations, apart from around the mutation site. The crystal structure of the Ndt80-DNA complex revealed that R177 makes a base specific bidentate interaction with G5' which is part of a conserved 5'-YpG-3' step. In the simulation with the mutant DNA, the side chain of R177 changes conformation and makes three new stable hydrogen bonds to the DNA backbone. This in turn induces a conformational change in the DNA backbone of the T6'-G5' step from the unusual BII state to the canonical BI state. The affinity difference for the protein-DNA complex with the native DNA compared with the mutant DNA is only about 3 kcal/mol. The free energy calculations of the base pair switch indicated a larger difference than what was found experimentally, about 7.7 kcal/mol, but this is explained in structural terms using the 10 ns simulations of the solvated complexes and the rearrangement of the R177 side chain.

Molecular dynamic simulations of cisplatin- and oxaliplatin-d(GG) intrastand cross-links reveal differences in their conformational dynamics

Mismatch repair proteins, DNA damage-recognition proteins and translesion DNA polymerases discriminate between Pt-GG adducts containing cis-diammine ligands (formed by cisplatin (CP) and carboplatin) and trans-RR-diaminocyclohexane ligands (formed by oxaliplatin (OX)) and this discrimination is thought to be important in determining differences in the efficacy, toxicity and mutagenicity of these platinum anticancer agents. We have postulated that these proteins recognize differences in conformation and/or conformational dynamics of the DNA containing the adducts. We have previously determined the NMR solution structure of OX-DNA, CP-DNA and undamaged duplex DNA in the 5'-d(CCTCAGGCCTCC)-3' sequence context and have shown the existence of several conformational differences in the vicinity of the Pt-GG adduct. Here we have used molecular dynamics simulations to explore differences in the conformational dynamics between OX-DNA, CP-DNA and undamaged DNA in the same sequence context. Twenty-five 10 ns unrestrained fully solvated molecular dynamics simulations were performed starting from two different DNA conformations using AMBER v8.0. All 25 simulations reached equilibrium within 4 ns, were independent of the starting structure and were in close agreement with previous crystal and NMR structures. Our data show that the cis-diammine (CP) ligand preferentially forms hydrogen bonds on the 5' side of the Pt-GG adduct, while the trans-RR-diaminocyclohexane (OX) ligand preferentially forms hydrogen bonds on the 3' side of the adduct. In addition, our data show that these differences in hydrogen bond formation are strongly correlated with differences in conformational dynamics, specifically the fraction of time spent in different DNA conformations in the vicinity of the adduct, for CP- and OX-DNA adducts. We postulate that differential recognition of CP- and OX-GG adducts by mismatch repair proteins, DNA damage-recognition proteins and DNA polymerases may be due, in part, to differences in the fraction of time that the adducts spend in a conformation favorable for protein binding.

A pyrazolato-bridged dinuclear platinum(II) complex induces only minor distortions upon DNA-binding

The cytotoxic, pyrazolato-bridged dinuclear platinum(II) complex [(cis-{Pt(NH3)2})2(mu-OH)(mu-pz)]2+ (pz=pyrazolate) has been found to cross-link two adjacent guanines of a double-stranded DNA decamer without destabilizing the duplex and without changing the directionality of the helix axis. A 1H NMR study of the oligonucleotide d(CTCTG*G*TCTC)-d(GAGACCAGAG), cross-linked at the two G* guanines by [(cis-{Pt(NH3)2})2(mu-pz)]3+, and molecular dynamics simulations of the explicitly solvated duplex were performed to characterize the structural details of the adduct. The dinuclear platinum cross-link unwinds the helix by approximately 15 degrees , that is, to a similar extent as the widely used antitumor drug cisplatin, but, in contrast to the latter, induces no significant bend in the helix axis. The Watson-Crick base-pairing remains intact, and the melting temperature of the duplex is unaffected by the cross-link. The helical twist is considerably reduced between the two platinated bases, as becomes manifest in an unusually short sequential H1'-H1' distance. This unwinding also affects the sugar ring of the guanosine in the 3'-position to the cross-link, which presents an N<-->S equilibrium. This is the first cytotoxic platinum complex that has been successfully designed by envisioning the structural consequences of its binding to DNA.

Internal versus terminal metalation of double-helical oligodeoxyribonucleotides

The formation of adducts between cis-[Pt(NH(3))(2)Cl(2)], Zn(II), and Mn(II) and double-stranded oligodeoxynucleotides was studied by 1D and 2D (1)H, (31)P, and (15)N NMR spectroscopy. For labile adducts involving Zn(II) and Mn(II), both (1)H chemical shifts (Zn(II)) and (1)H line-broadening effects (Mn(II)) showed that in the hexamer [d(GGCGCC)](2) I, the terminal G(1)-N7 is the exclusive binding site, while for the dodecamer [d(GGTACCGGTACC)](2) II, which contains both a terminal and internal GG pair, the preference for metal binding is the internal guanine G(7). Zn(II) binding to II was confirmed by natural-abundance 2D [(1)H,(15)N] HMBC NMR spectroscopy, which unambiguously showed that G(7)-N7 is the preferred binding site. The long duplex [d(GGTATATATACCGGTATATATACC)](2) III was expected to have a more pronounced accumulation of electrostatic potential towards the central part of the sequence (vs the terminal part) than does II. However, the Zn(II) titration of III showed no increase in coordination with the internal Gs (vs the terminal Gs), compared with what was observed for II. The reaction between the nonlabile metal complex cis-[PtCl(2)((15)NH(3))(2)] (cisplatin) and II showed a slight preference for the internal GG pair over the terminal GG pair. However, when the diaqua form of cisplatin cis-[Pt((15)NH(3))(2)(H(2)O)(2)] was reacted with II a more pronounced binding preference for the internal GG pair was observed.

Poor base stacking at DNA lesions may initiate recognition by many repair proteins

A fundamental question in DNA repair is how a mismatched or modified base is detected when embedded in millions to billions of normal base pairs. A survey of the literature and structural database reveals a common feature in all repair protein-DNA complexes: the DNA double helix is discontinuous at a lesion site due to base unstacking, kinking and/or nucleotide extrusion. Lesions induce destabilization and distortion of short linear DNAs, and underwinding in negatively supercoiled DNA presumably could compound the reduced stability caused by a lesion. A hypothesis is thus put forward that DNA lesion recognition occurs in two steps. Repair proteins initially recognize the weakened base stacking, and thus a flexible hinge at a DNA lesion. Sampling of flexible hinges rather than all DNA base pairs can reduce the task of finding a lesion by two to three orders of magnitude, from searching millions base pairs to thousands. After the initial encounter, a repair protein scrutinizes the shape, hydrogen bonding and electrostatic potentials of bases at the flexible hinge and dissociates if it is not a correct substrate. MutS, which has a broad range of substrates, actively dissociates from non-specific binding via an ATP-dependent proofreading mechanism. A single lesion may thus be sampled by BER, NER and MMR proteins until repaired. This proposition immediately suggests a mechanism for crosstalk between different repair and signaling pathways. It also raises the possibility that sampling of a lesion by one protein could facilitate loading of another by direct protein-protein or DNA mediated interactions.

Fixing the conformations of diamineplatinum(II)-GpG chelates: NMR and CD signatures of individual rotamers

The bulky, asymmetric analog of the antitumor drug cisplatin, [PtCl(2)(tmen)] (tmen = N,N,N'-trimethylethylenediamine), was used to produce crosslinks with the dinucleotide d(GpG), modeling the most frequent lesions that cisplatin and its analogs cause to DNA. The ligand tmen was chosen because it is expected to constrain the guanine cis to the NMe(2) group in the adduct [Pt(tmen){d(GpG)}](+) to an orientation perpendicular to the coordination plane and to stabilize the other guanine in an oblique orientation, thus maintaining a head-to-head geometry typical of cisplatin-d(GpG) crosslinks within single- and double-stranded DNA. Of the four possible combinations of tmen chirality (R or S symmetry of the coordinated NHMe group) and crosslink direction (5'-G bound cis to the secondary or the tertiary amino group of tmen), two isomers were preponderantly formed, [Pt(R-tmen){d(GpG)}](+) with 5'-G bound cis to NMe(2) and [Pt(S-tmen){d(GpG)}](+) with 5'-G bound cis to NHMe. The former was shown to have a right-handed R2 orientation of guanines similar to that found in duplex DNA, whereas the latter had a left-handed L1 orientation that modeled cisplatin-d(GpG) adducts within single-stranded DNA. The R2 rotamer was found to be in an equilibrium (as observed using EXSY spectroscopy) with a minor fraction (< or =4%) of a Delta-HT rotamer related to R2 by rotation of the 3'-G about the Pt-N7 bond. The major rotamers R2 and L1 were isolated using reverse-phase HPLC, and their NMR and CD signatures were compared to those of the corresponding rotamers of the less hindered adduct [Pt(dmen)(GpG)](+) (dmen = N,N-dimethylethylenediamine). From this and other comparisons with previously reported platinum dinucleotide complexes, and from molecular modeling, it could be concluded that both steric repulsion between guanine and substituents of the cis amino group and N-H...O6 hydrogen bonding are significant effects favoring the oblique orientation of one guanine base typical of the HH rotamers of [Pt(diamine){d(GpG)}](+) and [Pt(diamine)(GpG)](+) complexes.

Modeling anticancer drug–DNA interactions via mixed QM/MM molecular dynamics simulations

The development of anticancer drugs started over four decades ago, with the serendipitous discovery of the antitumor activity of cisplatin and its successful use in the treatment of various cancer types. Despite the efforts made in unraveling the mechanism of the action of cisplatin, as well as in the rational design of new anticancer compounds, in many cases detailed structural and mechanistic information is still lacking. Many of these drugs exert their anticancer activity by covalently binding to DNA inducing a distortion or simply impeding replication, thus triggering a cellular response, which eventually leads to cell death. A detailed understanding of the structural and electronic properties of drug-DNA complexes and their mechanism of binding is the key step in elucidating the principles of their anticancer activity. At the theoretical level, the description of covalent drug-DNA complexes requires the use of state-of-the-art computer simulation techniques such as hybrid quantum/classical molecular dynamics simulations. In this review we provide a general overview on: drugs which covalently bind to DNA duplexes, the basic concepts of quantum mechanics/molecular mechanics (QM/MM), molecular dynamics methods and a list of selected applications of these simulations to the study of drug-DNA adducts. Finally, the potential and the limitations of this approach to the study of such systems are critically evaluated.

Cisplatin interaction with cysteine and methionine, a theoretical DFT study

Interactions of hydrated cisplatin complexes with sulphur-containing amino acids cysteine and methionine were explored. The square-planar cis-[Pt(NH3)2(H2O)X]+ complexes (where X=Cl- and OH-) were chosen as mono- and dihydrated reactants. Calculations using density functional theory (DFT) techniques with B3LYP functional were performed. The isolated molecules and the supermolecular approaches were employed for the determination of the reaction energies. Bond dissociation energies (BDE) were estimated in the model of isolated molecules and supermolecules were used for the determination of the association energies between the two interacting parts. Formation of monodentate complexes by replacing the aqua-ligand with the S, N, and O-sites of both amino acids represents an exothermic process. The highest BDE was found in cysteine structures for the Pt-S coordination. The bonding energy is about 114 kcal/mol, which is comparable with cisplatin-guanine adducts. Analogous BDE for methionine complexes is smaller by about 40 kcal/mol. This correlates well with the known fact that cysteine forms irreversible cisplatin adducts while similar adducts in the methionine case are reversible. The formation of chelate structures is an exothermic reaction only for the hydroxo-form of reactants in the supermolecular approach where additional association interactions between the released water and chelate molecules sufficiently stabilize the final product.

Recognition complex between the HMG domain of LEF-1 and its cognate DNA studied by molecular dynamics simulations with explicit solvation

Molecular dynamics simulations of the complex formed between the HMG box of the lymphoid enhancer-binding factor (LEF-1) and its cognate DNA duplex were carried out with explicit inclusion of water. The simulation started with an NMR-based model (pdb code 2LEF) and the dynamics was pursued for 10 nanoseconds without constraints. It revealed that water intervenes in many ionic/polar interactions, establishing in particular local equilibria between direct and water-mediated hydrogen bonds, and thus increasing the entropy of the complex. Quite unexpectedly, the simulation indicated that a binding pocket for a specific water molecule may be reversibly formed at the apex of the bend induced in the DNA helix by LEF-1 binding, where a methionine side chain intercalates between two destacked adenines. We observed that the specific water molecule can temporarily replace the intercalated S-CH(3) group, acting as a sort of "extension" of the side chain. The residence time of this water molecule was about 3.5 ns. Simulations of the cognate DNA alone showed that this sequence has no intrinsic tendency to bend; therefore, the bending occurs solely as a consequence of the recognition, following the "induced-fit" mechanism.

Chiral differentiation of DNA adducts formed by enantiomeric analogues of antitumor cisplatin is sequence-dependent

1,2-GG intrastrand cross-links formed in DNA by the enantiomeric complexes [PtCl(2)(R,R-2,3-diaminobutane (DAB))] and [PtCl(2)(S,S-DAB)] were studied by biophysical methods. Molecular modeling revealed that structure of the cross-links formed at the TGGT sequence was affected by repulsion between the 5'-directed methyl group of the DAB ligand and the methyl group of the 5'-thymine of the TGGT fragment. Molecular dynamics simulations of the solvated platinated duplexes and our recent structural data indicated that the adduct of [PtCl(2)(R,R-DAB)] alleviated this repulsion by unwinding the TpG step, whereas the adduct of [PtCl(2)(S,S-DAB)] avoided the unfavorable methyl-methyl interaction by decreasing the kink angle. Electrophoretic retardation measurements on DNA duplexes containing 1,2-GG intrastrand cross-links of Pt(R,R-DAB)(2+) or Pt(S,S-DAB)(2+) at a CGGA site showed that in this sequence both enantiomers distorted the double helix to the identical extent similar to that found previously for the same sequence containing the cross-links of the parent antitumor cis-Pt(NH(3))(2)(2+) (cisplatin). In addition, the adducts showed similar affinities toward the high-mobility-group box 1 proteins. Hence, whereas the structural perturbation induced in DNA by 1,2-GG intrastrand cross-links of cisplatin does not depend largely on the bases flanking the cross-links, the perturbation related to GG cross-linking by bulkier platinum diamine derivatives does.

DNA interaction of dioxycyclobutenedione-(1,2-cyclohexanediamine) platinum(II) complex with potential anticancer activity

Dioxycyclobutenedione-(1,2-cyclohexanediamine)platinum(II), (R,R-DC-Pt) was found to have stronger cytotoxicity against six cancer cell lines than cisplatin and its DNA interactions was studied by calorimetric measurements, (13)C NMR. The binding specificity study of DNA base with R,R-DC-Pt was conducted by HPLC. To understand the molecular mechanism of R,R-DC-Pt with stronger cytotoxicity than that of cisplatin, we studied R,R-DC-Pt interaction with an oligonucleotide, d(ACCACGTGGT)(2), which contained c-H-ras gene encoding GGT by NMR spectroscopy. The oligomer DNA double helix was destroyed almost completely upon the R,R-DC-Pt binding. However under the same condition, the cisplatin binding with DNA was not so affected, and instead another conformation was formed, which suggests that larger damage to DNA can be induced by R,R-DC-Pt complex than that by cisplatin.

Protein interactions with platinum-DNA adducts: from structure to function

Because of the efficacy of cisplatin and carboplatin in a wide variety of chemotherapeutic regimens, hundreds of platinum(II) and platinum(IV) complexes have been synthesized and evaluated as anticancer agents over the past 30 years. Of the many third generation platinum compounds evaluated to date, only oxaliplatin has been approved for clinical usage in the United States. Thus, it is important to understand the mechanistic basis for the differences in efficacy, mutagenicity and tumor range between cisplatin and oxaliplatin. Cisplatin and oxaliplain form the same types of adducts at the same sites on DNA. The most abundant adduct for both compounds is the Pt-GG intrastrand diadduct. Cisplatin-GG adducts are preferentially recognized by mismatch repair proteins and some damage-recognition proteins, and this differential recognition of cisplatin- and oxaliplatin-GG adducts is thought to contribute to the differences in cytotoxicity and tumor range of cisplatin and oxaliplatin. A detailed kinetic analysis of the insertion and extension steps of dNTP incorporation in the vicinity of the adduct shows that both pol beta and pol eta catalyze translesion synthesis past oxaliplatin-GG adducts with greater efficiency than past cisplatin-GG adducts. In the case of pol eta, the efficiency and fidelity of translesion synthesis in vitro is very similar to that previously observed with cyclobutane TT dimers, suggesting that pol eta is likely to be involved in error-free bypass of Pt adducts in vivo. This has been confirmed for cisplatin by comparing the cisplatin-induced mutation frequency in human fibroblast cell lines with and without pol eta. Thus, the greater efficiency of bypass of oxaliplatin-GG adducts by pol eta is likely to explain the lower mutagenicity of oxaliplatin compared to cisplatin. The ability of these cellular proteins to discriminate between cisplatin and oxaliplatin adducts suggest that there exist significant conformational differences between the adducts, yet the crystal structures of the cisplatin- and oxaliplatin-GG adducts were very similar. We have recently solved the solution structure of the oxaliplatin-GG adduct and have shown that it is significantly different from the previously published solution structures of the cisplatin-GG adducts. Furthermore, the observed differences in conformation provide a logical explanation for the differential recognition of cisplatin and oxaliplatin adducts by mismatch repair and damage-recognition proteins. Molecular modeling studies are currently underway to analyze the mechanistic basis for the differential bypass of cisplatin and oxaliplatin adducts by DNA polymerases.

NMR solution structure of an oxaliplatin 1,2-d(GG) intrastrand cross-link in a DNA dodecamer duplex

We have determined, at high resolution, the NMR solution structure of an oxaliplatin-GG DNA dodecamer in the AGGC sequence context by 2D NMR studies. Homonuclear assignment strategies resulted in unambiguous assignment of 203 out of 249 protons, which corresponds to assignment of approximately 81% of the protons. Assignments of H5' and H5" protons were tentative due to resonance overlap. The structure of the oxaliplatin duplex was calculated using the program CNS with a simulated annealing protocol. A total of 510 experimental restraints were employed in the structure calculation. Of 20 calculated structures, the 15 with the lowest energy were accepted as a family. The RMSD of the 15 lowest energy structures was 0.68 A, indicating good structural convergence. The theoretical NOESY spectrum obtained by back-calculation from the final average structure showed excellent agreement with the experimental data, indicating that the final structure was in good agreement with the experimental NMR data. Significant conformational differences were observed between the oxaliplatin-GG 12-mer DNA we studied and all previous solution structures of cisplatin-GG DNA duplexes. For example, the oxaliplatin-GG adduct shows much less distortion at the AG base-pair step than the cisplatin-GG adducts. In addition, the oxaliplatin-GG structure also has a narrow minor groove and an overall axis bend of about 31 degrees, both of which are very different from the recent NMR structures for the cisplatin-GG adducts. These structural differences may explain some of the biological differences between oxaliplatin- and cisplatin-GG adducts.

Discrimination between BI and BII conformational substates of B-DNA based on sugar-base interproton distances

Molecular dynamics (MD) simulations of four water-solvated DNA duplexes were used to generate a database of approximately 27000 dinucleotide conformations. Analyzing this database, we investigated the relationship between so-called BI-BII transitions and short-range interproton distances. Four H-H distances were found particularly sensitive to BI-BII transitions: internucleotide H1'(n)-H68(n+1), H2'(n)-H68(n+1), and H2"(n)-H68(n+1), and intranucleotide H2"(n)-H68(n). Determination of these distances using classical NOESY spectroscopy can thus provide valuable indications on the existence of BII substates, complementing the existing method based on (31)P chemical shifts and (31)P-(1)H spin-spin coupling constants.

Simulating structural and thermodynamic properties of carcinogen-damaged DNA

A pair of stereoisomeric covalent adducts to guanine in double-stranded DNA, derived from the reaction of mutagenic and tumorigenic metabolites of benzo[a]pyrene, have been well characterized structurally and thermodynamically. Both high-resolution NMR solution structures and an array of thermodynamic data are available for these 10S (+)- and 10R (-)-trans-anti -[BP]-N(2)-dG adducts in double-stranded deoxyoligonucleotides. The availability of experimentally well-characterized duplexes containing these two stereoisomeric guanine adducts provides an opportunity for evaluating the molecular mechanics Poisson-Boltzmann surface area (MM-PBSA) method for computing thermodynamic properties from molecular dynamics ensembles. We have carried out 3-ns molecular dynamics simulations, using NMR solution structures as the starting models for the 10S (+)- and 10R (-)-trans-anti-dG adducts in a DNA duplex 11-mer using AMBER 6.0. We employed the MM-PBSA method to compute the free energies, enthalpies, and entropies of the two adducts. Our complete thermodynamic analysis agrees quite well with the full experimental thermodynamic characterization of these adducts, showing essentially equal stabilities of the two adducts. We also calculated the nuclear Overhauser effect (NOE) distances from the molecular dynamics trajectories, and compared them against the experimental NMR-derived NOE distances. Our results showed that the simulated structures are in good agreement with the NMR experimental NOE data. Furthermore, the molecular dynamics simulations provided new structural and biological insights. Specifically, the puzzling observation that the BP aromatic ring system in the 10S (+)-trans-anti-dG adduct is more exposed to the aqueous solvent than the 10R (-)-trans-anti-dG adduct, is rationalized in terms of the adduct structures. The structural and thermodynamic features of these stereoisomeric adducts are also discussed in relation to their reported low susceptibilities to nucleotide excision repair.