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.

A 5 ns unrestrained molecular dynamics (MD) simulation of the DNA duplex d(GCCG*G*ATCGC)-d(GCGATCCGGC), bearing a cis-Pt(NH(3))(2)(2+) unit crosslinking the two G* guanine bases, is reported. The MD trajectory was a posteriori correlated with NMR data determined for the same adduct, and it is shown that interproton distances and the characteristic chemical shifts are accounted for by the simulation. The simulation and its confrontation with the NMR data have confirmed the finding derived early from static models that the cytosine complementary to the 5' G*, C17, is mobile with respect to its adjacent bases. However, in contrast to our previous description of this mobility, which included rupture of the Watson-Crick hydrogen bonds and formation of non-Watson-Crick hydrogen bonds, the MD simulation indicated that the G*4-C17 pair moves continuously along a trajectory roughly perpendicular to the local helix axis, with retention of all three Watson-Crick hydrogen bonds. The simulation indicated the reversible formation of a hydrogen bond between the 5' oriented NH(3) ligand of platinum and the C3pG*4 phosphate group, in accord with our former prediction. Furthermore, the simulation has disclosed previously undetected BI <=> BII transitions at the G*5pA6 and A6pT7 steps, connected to formation/rupture of a hydrogen bond between the 3' oriented NH(3) ligand of platinum and the N7 atom of A6. All these conformational equilibria affect the form of the minor groove and increase the conformational flexibility at the platination site, and are thus likely to facilitate recognition by cellular proteins.

A complete kinetic study of GG versus AG plantination suggests that the doubly aquated derivatives of cisplatin are the actual DNA binding species.

The hairpin-stabilized double-stranded oligonucleotides d(TATGGTATT4ATACCATA) (I) and d(TATAGTATT4ATACTATA) (II) were allowed to react with the three aquated forms of the antitumor drug cisplatin (cis-[PtCl2(NH3)2], 1) which are likely candidates for DNA binding, that is, cis-[PtC1(NH3)2(H2O)]+ (2), cis-[Pt(NH3)2(H2O)2]2+ (3), and its conjugate base cis-[Pt(OH)(NH3)2(H2O)]+ (4). The reaction between I and [Pt(NH3)3(H2O)]2+ (5) was also studied for comparison. All reactions were monitored by HPLC. The platination reactions of I and II were carried out in NaClO4 (0.1M) at 293 K and at a constant pH of 4.5 +/- 0.1 for 2, 3, and 5. The data relative to the platination by 4 were obtained from measurements in unbuffered NaClO4 solutions (0.1M) at a starting pH close to neutrality, where 3 and 4 are present in equilibrium. In this case, a fit function describing the pH-time curve allowed the determination of the actual concentrations of 3, 4, and the dihydroxo complex. The platination rate constants characterizing the bimolecular reactions between either I or II and 2, 3, and 4 were individually determined along with the rate constants for hydrolysis of the chloro-monoadducts and for the chelation reactions of the aqua-monoadducts. The reactivity of compounds 2-5, which have the general formula cis-[Pt(NH3)2(H2O)(Y)]2+/-, decreases in the order 3>4>5>>2, that is, Y= H2O > OH- >NH3 >> Cl-, which is the order of decreasing hydrogen-bond donating ability of Y. Deprotonation of 3 to 4 reduces the reactivity of the platinum complex only by a factor of approximately equals 2, and both complexes discriminate between the different purines of I and II in the same manner. Whereas 3 and 4 react approximately three times faster with the GG sequence of I than with the AG sequence of II, 2 shows a similar reactivity towards both sequences. In view of the well-established preferential binding of cisplatin to GG sequences of DNA in vivo and in vitro, this result suggests that the actual DNA platination species are derived from double hydrolysis of cisplatin.

Circular dichroism spectra of the individual rotamers of [Pt(N,N-dmen)(GpG)]+ (N,N-dmen=N,N-dimethylethylenediamine) indicate that the base-base oscillator coupling is not the main source of ellipticity in cis-PtG2L2 head-to-head complexes.

The asymmetrical platinum complex [PtCl2(N,N-dmen)] (N,N-dmen = N,N-dimethylethylenediamine) reacts with the dinucleotide GpG to form two isomeric chelates of the formula [Pt(N,N-dmen)(GpG)]+ [9]. One of the isomers forms two stable rotamers separable by HPLC, whereas the other apparently prefers one single rotameric form. The favored conformations of these three forms were elucidated by means of molecular mechanics and dynamics techniques. In parallel, we have prepared the adduct, isolated the three rotamers, and recorded their solution circular dichroism (CD) spectra. For the first time we were thus able to correlate the CD features of individual rotamers of a cis-Pt(GpG) chelate with their structures. We show here that the two forms labeled in Inagaki's paper 1'e and 2e have the same right-handed helicoidal arrangement of the guanine bases but display different CD spectra in which the prominent bands have inverted signs. Thus, base-base interactions cannot be the principal cause of the CD of these compounds.

O-H small middle dot small middle dot small middle dotPt(II): Hydrogen Bond with a Strong Dispersion Component.

An MP2 ab initio study of the interaction between a H(2)O molecule and trans-[Pt(OH)(2)(NH(3))(2)] revealed a HO-H small middle dot small middle dot small middle dotPt(II) hydrogen bond (see picture) with a strong dispersion component (ca. 4 kcal mol(-1)). This dispersion interaction is independent of the charge on the complex and is likely to be ubiquitous in aqueous solutions of Pt(II) complexes.

Probing the mechanism of an Mn2+-dependent ribozyme by means of platinum complexes.

The smallest ribozyme system known is the pentanucleotide GAAACp, which is specifically cleaved by Mn2+, in the presence of poly(U), generating guanosine 2',3'-cyclic phosphate and AAACp. A plausible mechanism has been proposed, involving the participation of two Mn2+ with structural and catalytic roles, the first one cross-linking the two N7 atoms of G1 and A4, and the other binding to the N7 atom of A2 and activating the 2'-OH group of G1 [Kazakov, S. & Altman, S. (1992) Proc. Natl Acad. Sci. USA 89, 7939-7943]. In the present work, we have utilized the high affinity of Pt(II) complexes for N7 atoms of purines in an attempt to form a stable active ribozyme by replacing the structural Mn2+ by Pt2+. We thus replaced the proposed kinetically labile G1N7-Mn2+-A4N7 cross-link by an inert N7-trans-Pt(NH3)(2)(2+)-N7 cross-link. In a complementary investigation, the N7 atoms of the individual purines of GAAACp were selectively blocked by a Pt(NH3)(3)(2+) residue to determine which of the N7 sites participate in the Mn2+-mediated cleavage. Other N7-Pt(II)-N7 crosslinks were also investigated. Accordingly, we have prepared four monoadducts, each bearing the Pt(NH3)(3)(2+) residue on one of the purines and a series of chelates of trans-Pt(NH3)(2)(2+) and cis-Pt(NH3)(2)(2+) and have tested them for Mn2+-induced cleavage. Binding of Pt(NH3)(3)(2+) to G1 or A4 did not alter the efficiency of the specific cleavage between G1 and A2 catalyzed by Mn2+/poly(U), whereas cross-linking of G1 and A4 by trans-Pt(NH3)(2)(2+) inhibited it completely. Hence, a cross-link between G1 and A4 is not required for the site-specific cleavage. Binding of Pt(NH3)(3)(2+) to A2 or A3 strongly inhibits the G1/A2 cleavage, suggesting that these bases are likely to be involved in manganese coordination in the cleaving complex. A site-specific Mn2+-dependent cleavage between A4 and C5 was observed for the G1-A4 and G1-A3 adducts cross-linked by trans-Pt(NH3)(2)(2+), the G1-A2 adduct cross-linked by cis-Pt(NH3)(2)(2+), and the three monoadducts bearing the Pt(NH3)(3)(2+) residue on G1, A2 or A3; poly(U) did not exert any influence on this cleavage.

Asymmetrically substituted ethylenediamine platinum(II) complexes as antitumor agents: synthesis and structure-activity relationships.

A series of platinum dichloroethylenediamine complexes [PtCl2(R-en)] bearing a side chain on one carbon atom of the ethylenediamine ligand, with or without a functional group on the side chain, have been prepared and investigated for antitumor activity against L1210 leukemia. They were tested both in vitro, with cisplatin-sensitive and resistant cell lines, and in vivo, with cisplatin-sensitive and resistant tumors grafted i.p. in B6D2F1 mice. The rationale for this study was to test how charge, polarity and shape of the R side chain influence antitumor activity. Complexes carrying one or more ammonium groups on the side chain were all inactive. Derivatives with a carbamate function attached by the nitrogen atom, via a methylene group, to the ethylenediamine moiety ('N-bound' carbamate) were highly active in vitro and in vivo. The best results were obtained with these carbamates bearing hydrophobic substituents of intermediate size. Replacement of N-bound by O-bound carbamate or by urea groups led to decreased in vivo activity. Sulfonamide derivatives were all inactive. Good to excellent activities were also recorded for complexes bearing bulky bicycloalkyl substituents, without any functional group, attached to one ethylenediamine carbon atom. Thus, it is the steric features of the side chain rather than its polarity that appear to favor the antitumor activity of the complex. Compared to cisplatin and oxaliplatin, the present complexes do not exhibit advantages in terms of experimental antitumor activities in solid tumor models.

Cytotoxic activity of platinum (II) complexes with tri-n-butylphosphine. Crystal structure of the dinuclear hydrazine-bridged complex, cis,cis-[PtCl(PBu3n)2(mu-N2H4)PtCl(PBu3n) 2] (ClO4)2.2CHCl3.

A series of platinum(II) tri-n-butylphosphine complexes having the formulas cis-[PtCl2L2], NEt4[PtCl3L], [PtCl(en)L]Cl, [Pt(en)L2](ClO4)2, sym-trans-[Pt2Cl4L2], [Pt2Cl2L4](ClO4)2, trans,trans-[PtCl2L(mu-N2H4)PtCl2L] trans,trans-[PtCl2L(mu-en)PtCl2L], and cis,cis-[PtClL2(mu-N2H4)PtClL2](ClO4)2 (L = tri-n-butylphosphine; en = ethylenediamine) have been synthesized and their cytotoxic activity in vitro and in vivo has been studied. The solution behavior of the novel dinuclear diamine-bridged platinum(II) complexes has been investigated by means of UV and 31P NMR spectroscopy. For the ionic hydrazine compound cis,cis-[PtClL2(mu-N2H4)PtClL2](ClO4)2, an x-ray structure determination is reported. Crystal data: space group P2(1)/a, a = 17.803(1), b = 18.888(3), c = 12.506(3) A, beta = 107.97(2) degrees, Z = 2, R = 0.052, RW = 0.058. The platinum coordination is approximately square-planar, with the bond lengths Pt-Cl = 2.358(5), Pt-N = 2.15(1), Pt-P(trans to Cl) = 2.260(5), and Pt-P(trans to N) = 2.262(6) A. All investigated compounds were cytotoxic in vitro against L1210 cells and showed no cross-resistance to cisplatin. On the other hand, no antitumor activity was observed vs L1210 leucemia in DBA2 mice.

H8 chemical shifts in oligonucleotide cross-linked at a GpG sequence by cis-Pt(NH3)2(2+): a clue to the adduct structure.

The origin of the anomalous H8 chemical shifts observed in 1H-NMR spectra of oligonucleotides cross-linked at a GpG sequence with cis-[Pt(NH3)2]2+ has been investigated and clarified. The main contributions that distinguish the H8 resonances of the two platinum-ligating guanines from other GH8 signals and from each other are: (a) the inductive effect of platinum binding which we have recently quantified as a downfield shift of 0.48 +/- 0.07 ppm (M. H. Fouchet, D. Lemaire, J. Kozelka and J.-C. Chottard, unpublished results); (b) the ring-current effect of one GpG guanine on the H8 resonance of the other guanine, which is negative (shielding) for the 5'-H8 and positive (deshielding) for the 3'-H8 in single-stranded adducts, but has the opposite sign in double-stranded adducts; (c) a deshielding polarization effect of the phosphate 5' to the GpG unit. The different signs of the ring-current effects in single-stranded and double-stranded oligonucleotides originate from the orientation of the guanines in the cis-[Pt(NH3)2(Gua)2]2+ moiety (Gua, guanine), which is left-handed helicoidal in single strands and right-handed helicoidal in double strands. In the platinated dinucleotides (cis-[Pt(NH3)2(GpG)]+, cis-[Pt(NH3)2(d(GpG))]+ and cis-[Pt(NH3)2(d(GpG))]), the guanines assume either the left-handed or the right-handed arrangement, depending on the sugar moiety (ribose or deoxyribose), protonation state at N1 and, in the solid state, on crystal forces. This work shows that chemical shifts contain valuable structural information which is complementary to that extracted from correlated spectroscopy and nuclear Overhauser spectroscopy data.

A d(GpG)-platinated decanucleotide duplex is kinked. An extended NMR and molecular mechanics study.

A conformational study of the double-stranded decanucleotide d(GCCG*G*ATCGC).d(GCGATCCGGC), with the G* guanines chelating a cis-Pt(NH3)2 moiety, has been accomplished using 1H and 31P NMR, and molecular mechanics. Correlation of the NMR data with molecular models has disclosed an equilibrium between several kinked conformations and has ruled out an unkinked structure. The deformation is localized at the CG*G*.CCG trinucleotide where the helix is kinked by approximately 60 degrees towards the major groove and unwound by 12-19 degrees. The models revealed an unexpected mobility of the cytosine complementary to the 5'-G*. This cytosine can stack on either branch of the kinked complementary strand. The energy barrier between the two positions has been calculated to be less than or equal to 12 kJ/mol. The NMR data are in support of rapid flip-flopping of this cytosine. An explanation for the strong downfield shift observed in the 31P resonance of the G*pG* phosphate is given.

How does cisplatin alter DNA structure? A molecular mechanics study on double-stranded oligonucleotides.

Molecular models for two double-stranded decanucleotides, d(GCCG*G*ATCGC)-d(GCGATCCGGC) (1) and d(GCTG*G*ATCGC)-d(GCGATCCAGC) (2), with the G* guanines cross-linked by a cis-Pt(NH3)2 moiety, were calculated using molecular mechanics. Nine models for 1 and eight models for 2 are reported; in all of them, the double helix is kinked by approx. 60 degrees towards the major groove and slightly unwound. The model building has been guided by comparison with the NMR data available for duplex 1. The influence of the base at the 5'-side of the coordinated G*G* dinucleotide is discussed.

Ascending pathways mediating somatoautonomic reflexes in exercising dogs.

The ascending spinal pathways mediating somatocardiovascular reflexes during exercise were studied in unanesthetized dogs by placing lesions in the lumbar spinal cord. After training to run on a treadmill with hindlimbs only, 20 dogs were anesthetized and instrumented using sterile surgical techniques. To chronically record heart rate and arterial blood pressure, the aorta was cannulated via the omocervical artery. To test the intactness of descending spinal sympathetic pathways, reflex pressor responses to baroreceptor hypotension were produced by bilateral carotid arterial occlusion using pneumatic vessel occluders placed around the common carotid arteries. To generate transient ischemic exercise (120 s), a pneumatic occluder was placed around the left iliac artery. Eight to 10 days after instrumentation, blood pressure and heart rate were monitored at rest and during hindlimb running with and without simultaneous iliac arterial occlusion. The modest pressor response and tachycardia elicited by hindlimb exercise were markedly augmented by simultaneous hindlimb ischemia (i.e., iliac arterial occlusion). Lesion placement in the dorsolateral sulcus area and the dorsolateral funiculus at L2 significantly reduced the blood pressure and heart rate responses to simultaneous exercise occlusion. The cardiovascular responses to nonischemic exercise and bilateral carotid arterial occlusion were not altered by such spinal sections. It is concluded that in the dog the ascending spinal pathways mediating cardiovascular responses to ischemic exercise are located in the lateral funiculus, including the dorsolateral sulcus area and dorsolateral funiculus.

Ascending spinal pathways for somatoautonomic reflexes in the anesthetized dog.

The ascending spinal pathways mediating pressor and heart rate responses to somatic afferent stimulation and induced exercise were studied in pentobarbital- and alpha-chloralose-anesthetized dogs. Bilateral sciatic nerve stimulation and induced exercise, via lumbosacral ventral root stimulation, produced pressor and heart rate responses that were blocked by bilateral dorsolateral sulcus (DLS) lesions of the lumbar spinal cord (L1-L3). Baroreceptor-mediated bradycardia were attenuated by sciatic stimulation but not by induced exercise. This attenuation was blocked by combined dorsolateral funiculus (DLF) and DLS lesions. Induced exercise with vascular occlusion to the muscle augmented the pressor and heart rate responses to exercise. These responses were blocked by combined DLS and DLF lesions. Such lesions did not influence responses to bilateral carotid occlusion, indicating that descending autonomic pathways were intact. Therefore, ascending spinal pathways mediating somatocardiovascular reflexes in anesthetized dogs are located in the lateral funiculus, extending from the dorsal root entry zone to a position somewhat ventral to the dentate ligament.

Somato-autonomic reflexes in anesthetized and unanesthetized dogs.

Blood pressure and heart rate alterations were induced in anesthetized and unanesthetized mongrel dogs. Pressor responses were brought about in the anesthetized group by bilateral, high intensity stimulation of the sciatic nerves. In these animals, bilateral section of the dorsolateral sulcus area (DLS) in the lumbar spinal cord completely eliminated both blood pressure and heart rate responses to this stimulation. Baroreceptor-mediated bradycardia induced by pressor doses of phenylephrine was attenuated by sciatic nerve stimulation. Obliteration of this baroreceptor-somatic afferent interaction required ventrolateral extension of the bilateral spinal lesions to include both the DLS and the dorsolateral funiculus (DLF). Blood pressure and heart rate increments in unanesthetized dogs were induced by treadmill running. The pressor response to exercise was markedly increased by transient hind limb arterial occlusion during the course of the run. Surgical interruption of the ascending limb of the somato-autonomic reflex in the spinal cord of these animals (L1-L2 combined DLS and DLF lesion) significantly reduced the blood pressure response to simultaneous exercise-occlusion. These spinal lesions also reduced the heart rate response to treadmill running without occlusion. The descending pathways involved in autonomic reflexes appeared intact as the spinal lesions did not alter the blood pressure or heart rate response to bilateral carotid artery occlusion.

Ascending spinal pathways mediating somato-cardiovascular reflexes.

The influence of spinal somatic afferent pathways upon cardiovascular parameters were studied in anesthetized and conscious dogs. Previous studies on anesthetized cats indicated that activation of somatic A afferent fibres results in a depressor response mediated by ascending spinal pathways in the dorsolateral funiculus (DLF). Additional activation of somatic C afferent fibers results in a pressor response mediated by pathways in the region of the dorsolateral sulcus (DLS). In anesthetized dogs sympathetic-mediated cardiovascular responses to somatic afferent stimulation are found to be conducted in the DLS. To test the possibility that these afferent pathways may mediate cardiovascular responses to exercise, dogs instrumented for blood pressure and heart rate monitoring were trained to run on a treadmill on all 4 legs and on their hindlegs only. Lesions on the DLS and a portion of the DLF significantly altered heart rate responses to treadmill running on all 4 legs. These data suggest that somato-autonomic reflexes participate in the cardiovascular adjustments which occur during exercise.