Azole-Bridged Diplatinum Anticancer Compounds. Modulating DNA

11873

2007, 111, 11873-11876 Published on Web 10/11/2007

Azole-Bridged Diplatinum Anticancer Compounds. Modulating DNA Flexibility to Escape Repair Mechanism and Avoid Cross Resistance Katrin Spiegel,† Alessandra Magistrato,*,‡ Paolo Carloni,‡ Jan Reedijk,§ and Michael L. Klein† Center for Molecular Modeling, UniVersity of PennsylVania, 231 South 34th Street, Philadelphia, PennsylVania 19104, CNR-INFM-Democritos and International School of AdVanced Studies (SISSA/ISAS), Trieste, Italy, and Leiden Institute of Chemistry, Leiden UniVersity, P.O. Box 9502, 2300 RA Leiden, The Netherlands ReceiVed: August 3, 2007; In Final Form: September 4, 2007

Dinuclear azole-bridged Pt compounds bind to DNA helices, forming intrastrand crosslinks between adjacent guanines in a similar way to cisplatin. Their cytotoxic profile is, however, different from that of first and second generation Pt drugs in that they lack cross resistance in cisplatin-resistant cell lines. In contrast to cisplatin, which induces a large kink in DNA duplex, structural NMR studies and molecular dynamics simulations have shown that azole-bridged diplatinum compounds induce only small structural changes in double-stranded DNA. These structural differences have been invoked to explain the different cytotoxic profile of these compounds. Here, we show that in addition to the small structural changes in DNA, dinuclear Pt compounds also affect DNA minor groove flexibility in a different way than cisplatin. Free-energy calculations on azole-bridged diplatinum DNA adducts reveal that opening of the minor groove requires a higher freeenergy cost (∆G ∼ 7-15 kcal/mol) than in the corresponding cisplatin-DNA adduct (∆G ∼ 0 kcal/mol). This could prevent minor groove binding proteins from binding to diplatinum-DNA adducts thus leading to a different cellular response than cisplatin and possibly decreasing the activity of excision repair enzymes. Although the development of drug resistance is a highly complex mechanism, our findings provide an additional rationale for the improved cytotoxic activity of these compounds in cell lines resistant to cisplatin.

Cisplatin is one of the most successful anticancer drugs with cure rates up to 90% for certain cancer types.1-3 However, its use is limited due to intrinsic and acquired resistance.4 Intensive research is dedicated to understand and circumvent resistance problems, potentially deriving from the structural distortions that these drugs induce in the DNA double helix.4-6 It is well known that cisplatin-DNA adducts are mainly characterized by a relatively large kink toward the major groove and opening of the minor groove.2,6 Cross resistance in Pt-derived drugs, such as carboplatin and oxaliplatin, is believed to originate from the similar structural distortions these drugs induce in DNA upon binding.2,7 An increased activity of the DNA repair machinery, which is able to recognize cisplatin-induced DNA deformations, has been shown to be in part responsible for resistance development.3,8-11 Current research focuses therefore on the design of compounds, which upon DNA binding leave the double helix almost unaltered, being in consequence less prone to recognition of DNA-binding proteins.12-15 In this respect, azole-bridged diplatinum compounds 1 and 2 (Figure 1) are promising candidates, because they only induce small structural changes in DNA.13,16,17 These compounds form a N7(G)-N7(G) intrastrand * To whom correspondence should be addressed. E-mail: [email protected]. † University of Pennsylvania. ‡ CNR-INFM-Democritos and International School of Advanced Studies (SISSA/ISAS). § Leiden University.

10.1021/jp0762323 CCC: $37.00

Figure 1. The azole-bridged diplatinum compounds [Pt2(µ-OH)(µpz)]2+ (1), [Pt2(µ-OH)(µ-1,2,3ta)]2+ (2), and the atom numbering in the rings; the amines are labeled cis and trans with respect to the bridging OH ligand. Two different binding modes of 2 to DNA yield to adducts B and C.

cross link similar to cisplatin. In vitro studies have shown that both 1 and 2 have cytotoxic activities comparable to cisplatin,12 but interestingly 2 is more active in cisplatin-resistant cell lines.17 This different cytotoxic profile has been attributed to the fact that upon binding to DNA compound 2 can undergo a migration mechanism in which Pt2 migrates from N2 to N3 (2′) on the triazolate (Figure 1),12,17 leading to two different Pt-DNA © 2007 American Chemical Society

11874 J. Phys. Chem. B, Vol. 111, No. 41, 2007 adducts 2-DNA (B) and 2′-DNA (C). The NMR structure of compound 1 in complex to DNA is known,16 whereas for the 2- and 2′-DNA adducts classical and hybrid quantum mechanical/molecular mechanical (QM/MM) simulations have provided the first insights into structural rearrangements of DNA upon binding of these compounds.16,18,19,20 Besides the structural characteristics of the platinated DNA duplex, also groove flexibility might influence the binding affinity of proteins to DNA. Groove flexibility plays an important role in many DNA biological functions and in particular groove deformability can influence DNA recognition by many DNA-binding proteins, including those proteins involved in repair mechanism.21-23 Generally, the binding of proteins to the DNA minor groove results in a significant deformation from the standard B-DNA conformation, namely an opening of the minor groove and a bending of DNA toward the major grove.24,25 Zacharias et al. have shown that the free energy necessary to distort the minor groove can be related to the binding affinity of minor groove binding proteins.21 So far, this aspect has not been considered in the case of Pt drugs but could be important, since it has been demonstrated that the use of cisplatin enhances protein-DNA recognition.21 An increased activity of proteins involved in the excision repair mechanism could ultimately lead to cisplatin resistance. Herein, we investigate the effect of cisplatin and dinuclear Pt drugs on minor groove flexibility by performing classical molecular dynamics simulations (MD) with appropriately derived force field parameters for the diplatinum moiety20,26 and in combination with an adaptive bias force (ABF) method to calculate free-energy profiles.27 We show that dinuclear Pt compounds impose a low strain on the DNA helix but increase the energetic barrier associated to minor groove opening. We carried out classical MD simulations on the three azolebridged Pt-DNA adducts, resulting from binding of compound 1, 2, and 2′ to the DNA hexadecamer with the sequence (5′d(CpCpUpCpTpCpTpG*pG*pACpCpTpTpCpCp)-3′), leading to complexes A (1-DNA), B (2-DNA), and C (2′-DNA) (Figure 1). We selected this DNA sequence because the X-ray structure of the ternary complex of cisplatined DNA with this sequence in the complex with HMG1 protein (DNA-HMG, hereafter) was solved by X-ray crystallography8 thus allowing direct comparison to the experimental structure. Two additional simulations on the cisplatin-DNA complex (D) and the free oligonucleotide (E) were carried out. Classical parameters for the nonstandard Pt moieties were obtained by a force-matching procedure and described previously.20,26 The AMBER parm99 force field was used for simulating DNA.28,29 Potassium counterions were added to achieve overall neutrality, and the systems were solvated in a box of TIP3P water.30 We have used potassium counterions for consistency with our previous studies, although we are aware that the different type of counterions may affect the structure of DNA.31 In addition, K+ is routinely used in classical MD simulations.21,32 Each system was minimized and brought to 300 K by temperature rescaling. The production runs were conducted at constant temperature of 300 K and constant pressure of 1 atm for 10 ns. Long-range electrostatic interactions were treated using the smooth-particle mesh Ewald method.33 For the nonbonded short-range interactions, we employed a cutoff of 12 Å. The integration time step was 1 fs, and all the bonds were flexible except for the ones involved in water molecules. Ten nanosecond trajectories were obtained using the NAMD simulation package.34,35 The trajectories were analyzed in terms of helical parameters, using the program Curves 5.0.36 Here, we report the values for roll, rise,

Letters TABLE 1: Roll [°], Rise [Å], Tilt [°], Angle [°], and Twist [°] at the G8-G9 Base-Step of the Models Investigated in This Studya A B C D E (1-DNA) (2-DNA) (2′-DNA) (cisplatin-DNA) (free DNA) roll 12.9 (5.2) 6.8 (4.6) -8.0 (5.0) rise 3.0 (0.3) 3.3 (0.2) 3.7 (0.3) tilt 14.9 (4.5) 12.0 (4.8) 21.7 (6.4) angle 6.6 (2.4) 5.7 (2.6) 9.1 (3.6) twist 20.0(6.5) 29.9(5.4) 35.7(4.0) minor groove 6.7 (1.4) 5.7 (1.3) 5.5 (1.4) 4.8 (0.7) 5.6 (0.8) 5.6 (1.0) major groove 14.0 (2.1) 15.0 (2.0) 16.5 (2.2) 6.4 (1.9) 5.6 (1.7) 3.8 (1.9)

64.4 (6.5) 4.6 (0.5) -14.4 (5.8) 30.7 (4.7) 13.6(7.2) 10.0 (1.4) 2.6 (1.0) 9.9 (2.3) 9.3 (2.1)

1.9 (6.9) 3.3 (0.3) 3.1 (6.7) 5.6(2.5) 30.7 (8.6) 6.3 (1.5) 4.9 (0.7) 13.5 (1.6) 5.5 (1.8)

a Groove width and depth (in italic) at G9 [Å]. Standard deviations are given in parentheses.

tilt, and angle, where the latter refers to the angle between the two normal vectors of two adjacent nucleobase planes. For the free-energy simulations, we used the ABF methodology described by Chipot et al.27 In this method, a bias force opposite to the force along the reaction coordinate is added during the MD simulation. This bias force is continuously adapted to match the actual force resulting from the molecular force field. This leads to a smoothening of the potential energy surface along the reaction coordinate and to a homogeneous sampling. To further increase the efficiency of the sampling, the entire distance of the reaction coordinate ξ from 9 to 17 Å was divided into several windows of a width of 2 Å. The free energy was collected in bins of width dξ ) 0.1 Å, and for each bin 800 steps of unbiased MD were carried out to accrue the initial adaptive force. A total of 5 ns in the NPT ensemble were simulated to obtain each profile. Classical MD simulations carried out on A-C showed that the dinuclear Pt compounds induced only small structural rearrangements at the lesion site, in agreement with QM/MM and classical MD simulations of azole-bridged diplatinum-DNA adducts with a different DNA sequence. The most significant changes included (i) a moderate increase in rise, (ii) a small or negative roll angle, (iii) a small local axis bend, (iv) a rather narrow minor groove, and (v) a significantly larger major groove (Table 1).18,19 In general, local and global DNA parameters were quite close to those of canonical B-DNA but in strong contrast to cisplatin-DNA (Table 1).37 In fact, structural parameters of the lesion site in cisplatinated DNA were well characterized and showed (i) a large rise, (ii) an increase in roll angle, (iii) a large axis bend, (iv) a narrow minor groove, and (v) a large major groove (Table 1).37,38 The different structural response of DNA to 1 and 2 binding has been invoked as a rationale for the lack of cross resistance to cisplatin.12,17 Interestingly, 1 and 2 had similar activity in most cell lines, but 2 showed better activity in cisplatin-resistant cell lines. This indicates that the 2-DNA or 2′-DNA complex are more resilient to the resistance mechanism developed in cisplatin-resistant cell lines. This could be due to the observed migration mechanism of Pt along the triazolate ring from N2 to N3. This mechanism can only occur in compound 2 and leads to 2′. The 2′-DNA adduct, C, showed different structural characteristics with respect to A and B.18,20 In particular, in C the rise was larger and the roll between the two platinated base pairs was larger and negative. The static picture described here might indeed provide a first explanation for the different cellular response of these two compounds, but we suggest that it is not the sole cause for the different cytotoxic profiles observed for 1 and 2.12,17 To provide a more accurate picture of the differences between 1, 2, and cisplatin, we investigated how these compounds affect groove flexibility. In

Letters

J. Phys. Chem. B, Vol. 111, No. 41, 2007 11875

Figure 2. (Left) Backbone atoms and their centers of mass (transparent and solid blue spheres, respectively) defining the reaction coordinate ξ. (Right) Free-energy profiles for A-E. Small arrows indicate the free-energy minima, and the yellow bar shows the minor groove width of DNAHMG.

Figure 3. Superposition of average MD structures (ξ ) 15-17 Å) and DNA-HMG (brown). (Left) C (blue), rmsd ) 3.1 Å (central 8 residues). (Right) D (pink), rmsd ) 1.2 Å (central 8 residues).

Figure 4. Three snapshots from ABF simulation for ξ ) 15-17 Å, showing that in B the Watson-Crick H-bond pattern is maintained, whereas in C the H-bond pattern between G8-C25 is broken sometimes, and the two bases are in a staggered conformation.

particular, we calculated the free-energy profiles for minor groove opening via ABF MD simulations,27 choosing as reaction coordinate ξ, which is the distance between the center of mass of the backbone atoms of G9 and A10 on the first strand and to C25 and A26 on the second strand (Figure 2). We first performed several test simulations with slightly different choices of ξ to assess the dependency of the results on the choice of reaction coordinate (data not shown). We found that the results are all very similar and in qualitative agreement. The final reaction coordinate was the one that leads to the best structural agreement between the averaged ABF MD structure of D and the X-ray structure of DNA-HMG (pdb-entry 1CKT)8 (Figure 3). To compare the free-energy profiles of A-E, the minimum in each profile was set the to 0 kcal/mol, and a reference value of ξ was chosen along the reaction coordinate to compare relative free-energy profiles. In particular, we took ξ ) 16 Å, which corresponds to the minor groove width of DNA in DNAHMG (yellow bar in Figure 2). A low energetic cost during minor groove opening (∆G) indicated favorable protein-DNA

TABLE 2: Calculated Free-Energy Values (kcal/mol) Associated to the Minor Groove Opening at ξ ) 16 Å ∆G

A

B

C

D

E

7.3

7.8

14.8

0

5.0

complex formation. From Figure 2, it appeared that cisplatin favors protein binding, because a minor groove opening did basically happen at no energetic cost. In comparison, minor groove opening in free DNA (E) requires an energetic cost of ∆G ) +5 kcal/mol. A and B showed a significant increase in free energy (∆G ∼ 7-8 kcal/mol) with respect to D, although the free-energy cost of minor groove opening was only slightly higher than that of E (∆G ∼ 2-3 kcal/mol). Remarkably, in C the energetic cost associated to minor groove opening was significantly higher, namely ∆G ∼ +15 and 10 kcal/mol with respect to cisplatinated and free DNA (Table 2). Thus, the freeenergy profiles of A and B were very similar to E, suggesting that these Pt-DNA adducts are difficult to distinguish from free DNA, whereas C led to the most significant changes both in structure and minor groove flexibility. All three azole-bridged

11876 J. Phys. Chem. B, Vol. 111, No. 41, 2007 Pt adducts A-C showed a larger free-energy cost to open the minor groove, which was probably due to a decreased flexibility of the G8-G9 base-pair step upon drug binding. In D and E, the minor groove opening came along with an increase in roll and local angles at the G8-G9 base step, while only a moderate increase of these parameters was observed for A-C. A close look at the structural properties of the Pt lesion along the MD simulations suggests that the azole-bridged diplatinum compounds act like a clamp, imposing a small or negative roll at the G8-G9 site and opposing to minor groove opening (average structures for each ABF sampling windows are shown in Supporting Information, superimposed to the average unconstrained structure). This effect was most pronounced in C, which led to the largest negative roll at the G8-G9 step and to the largest structural differences with respect to DNA-HMG (Figure 4). Indeed, significant distortions in Watson-Crick hydrogen bonding occurred in C at the G8-C25 base pair at large minor groove widths (ξ > 15 Å) (Figure 4), resulting in an unfavorable minor groove opening. In A and B, the hydrogen pattern was maintained, but the large roll, angle, and tilt values at adjacent base-pair steps indicated unfavorable distortions due to minor groove opening (Figure 3, and Supporting Information), although less pronounced than in C. In summary, azole-bridged diplatinum compounds bind to DNA inducing only small structural changes in the duplex, thereby lowering the chance of recognition by repair enzymes. The minor groove opening requires a larger energetic cost for azole-bridged Pt-DNA adducts compared to that of cisplatinated or free DNA, especially if the migration product 2′ binds to DNA forming C. This binding could significantly affect the binding of excision repair enzymes and other proteins involved in the cytotoxic activity of common Pt drugs. Subsequently, this process could lead to a different cellular response and in turn to a lower resistance and cross resistance development with respect to cisplatin. Interestingly, in vitro experiments have shown that compound 1 and 2 have similar cytotoxic activity12 except for cell lines resistant to cisplatin. In these cells, 2 proves to be more active.17 One possible mechanism of resistance development is an increased activity of excision-repair enzymes, which recognize the cisplatin-induced kink and widened minor groove in the DNA duplex. On the basis of our free-energy profiles, we formulate the hypothesis that the significantly larger energetic cost associated to minor groove opening of 2′-DNA compared to cisplatin-DNA may be, at least in part, responsible for the improved cytotoxic activity in cisplatin-resistant cells. However, we should remark that resistance development is a very complex mechanism, involving more than one pathway.5,6 Nevertheless, our results suggest that DNA flexibility is an additional factor, which should be taken into account when looking at resistance problems of DNA-binding compounds,14,17 and these considerations could help in tailoring drugs effective in cell lines characterized by cisplatin resistance. Acknowledgment. We thank Je´roˆme He´nin for his valuable help with ABF, Professor U. Ro¨thlisberger and Professor P. Ruggerone for reading the manuscript. Calculations were performed on BlueGene/L, San Diego Supercoputing Center and Cray XT3, Pittsburgh Supercomputer Center. The Netherlands

Letters Council for Chemical Research (CW) is thanked for financial support to one of the authors (J.R.). Supporting Information Available: Helical parameters (rise, roll, tilt, angle) for all base pairs for MD simulations A-E and average MD structure for each simulated window along the reaction coordinate ξ. This material is available free of charge via the Internet at http://pubs.acs.org. References and Notes (1) Reedijk, J. Proc. Natl. Acad. Sci. U.S.A. 2003, 100, 3611. (2) Jamieson, E. R.; Lippard, S. J. Chem. ReV. 1999, 99, 2467. (3) Brabec, V.; Kasparkova, J. Drug. Resist. Updates 2002, 5, 147. (4) van Zutphen, S.; Reedijk, J. Coord. Chem. ReV. 2005, 249, 2845. (5) Rabik, C. A.; Dolan, M. E. Cancer Treat. ReV. 2007, 33, 9. (6) Zorbas, H.; Keppler, B. K. ChemBioChem 2005, 6, 1157. (7) Wang, D.; Lippard, S. J. Nat. ReV. Drug DiscoVery 2005, 4, 307. (8) Ohndorf, U.; Rould, M.; He, Q.; Pabo, C.; Lippard, S. J. Nature 1999, 399, 708. (9) Cohen, S. M.; Mikata, Y.; He, Q.; Lippard, S. J. Biochemistry 2000, 39, 11771. (10) Cohen, S. M.; Lippard, S. J. Cisplatin: From DNA damage to cancer chemotherapy Prog. Nucleic Acid Res. Mol. Biol. 2001, 67, 93. (11) He, Q.; Ohndorf, U.; Lippard, S. J. Biochemistry 2000, 39, 14426. (12) Komeda, S.; Lutz, M.; Spek, A. L.; Chikuma, M.; Reedijk, J. Inorg. Chem. 2000, 39, 4230. (13) Komeda, S.; Kalayda, G. V.; Lutz, M.; Spek, A. L.; Yamanaka, Y.; Sato, T.; Chikuma, M.; Reedijk, J. J. Med. Chem. 2003, 46, 1210. (14) Komeda, S.; Bombard, S.; Perrier, S.; Reedijk, J.; Kozelka, J. J. Inorg. Biochem. 2003, 96, 357. (15) Kalayda, G. V.; Fakih, S.; Bertram, H.; Ludwig, T.; Oberleithner, H.; Krebs, B.; Reedijk, J. J. Inorg. Biochem. 2006, 100, 1332. (16) Teletchea, S.; Komeda, S.; Teuben, J. M.; Elizondo-Riojas, M. A.; Reedijk, J.; Kozelka, J. ChemsEur. J. 2006, 12, 3741. (17) Komeda, S.; Lutz, M.; Spek, A. L.; Yamanaka, Y.; Sato, T.; Chikuma, M.; Reedijk, J. J. Am. Chem. Soc. 2002, 124, 4738. (18) Magistrato, A.; Ruggerone, P.; Spiegel, K.; Carloni, P.; Reedijk, J. J. Phys. Chem. B 2006, 110, 3604. (19) Spiegel, K.; Magistrato, A. Org. Biomol. Chem. 2006, 4, 2507. (20) Spiegel, K.; Magistrato, A.; Maurer, P.; Ruggerone, P.; Carloni, P.; Rothlisberger, U.; Reedijk, J.; Klein, M. L. J. Comput. Chem., in press. (21) Zacharias, M. Biophys. J. 2006, 91, 882. (22) Harrington, R. E. Mol. Microbiol. 1992, 6, 2549. (23) Gromiha, M. M. J. Biotechnol. 2005, 117, 137. (24) Olson, W. K.; Gorin, A. A.; Lu, X. J.; Hock, L. M.; Zhurkin, V. B. Proc. Natl. Acad. Sci. U.S.A. 1998, 95, 11163. (25) Deremble, C.; Lavery, R. Curr. Opin. Struct. Biol. 2005, 15, 171. (26) Maurer, P.; Laio, A.; Hugosson, H.; Colombo, M. C.; Rothlisberger, U. J. Chem. Theory Comput. 2007, 3, 628. (27) Chipot, C.; Henin, J. J. Chem. Phys. 2005, 123, 224906. (28) Case, D. A.; Cheatham, T. E., III; Darden, T.; Gohlke, H.; Luo, R.; Merz, K. M., Jr.; Onufriev, A.; Simmerling, C.; Wang, B.; Woods, R. J. J. Comput. Chem. 2005, 26, 1668. (29) Cheatham, T. E., III; Cieplak, P.; Kollman, P. A. J. Biomol. Struct. Dyn. 1999, 16, 845. (30) Joergensen, W.; Chandrasekhar, J.; Madura, J.; Impey, R.; Klein, M. J. Chem. Phys. 1983, 79, 926. (31) Petrov, A. S.; Pack, G. R.; Lamm, G. J. Phys. Chem. B 2004, 108, 6072. (32) Roccatano, D.; Barthel, A.; Zacharias, M. Biopolymers 2007, 85, 407. (33) Darden, T.; York, D. L. P. J. Chem. Phys. 1993, 98, 10089. (34) Phillips, J. C.; Braun, R.; Wang, W.; Gumbart, J.; Tajkhorshid, E.; Villa, E.; Chipot, C.; Skeel, R. D.; Kale, L.; Schulten, K. J. Comput. Chem. 2005, 26, 1781. (35) Laxmikant, K.; Skeel, R.; Bhandarkar, M.; Brunner, R.; Gursoy, A.; Krawetz, N.; Phillips, J.; Shinozaki, A.; Varadarajan, K.; Schulten, K. J. Comput. Phys 1999, 151, 283. (36) Lavery, R.; Sklenar, H. J. Biomol. Struct. Dyn. 1989, 6, 655. (37) Spiegel, K.; Rothlisberger, U.; Carloni, P. J. Phys. Chem. B 2004, 108, 2699. (38) Cohen, G. L.; Bauer, W. R.; Barton, J. K.; Lippard, S. J. Science 1979, 203, 1014.