Functional overexpression of human poly(ADP-ribose) polymerase in transfected rat tumor cells.

Poly(ADP-ribose) polymerase (PARP, EC 2.4.2.30) is a nuclear enzyme possibly involved in DNA base excision repair. The presence of single- or double-strand breaks in DNA stimulates this enzyme to covalently modify acceptor proteins with poly(ADP-ribose) in a reaction that uses NAD+ as substrate. To test the hypothesis that increased PARP activity could promote resistance towards DNA-damaging agents and gamma-radiation, we established stable rat cell transfectants that constitutively express human PARP. A number of subclones that showed different levels of PARP activity were isolated from two primary transfectants of different clonal origin. PARP activity was determined in permeabilized cells after maximal stimulation with a short, double-stranded oligonucleotide. Activity in different human PARP-expressing subclones was increased 1.6- to 3.1-fold compared with non-expressing subclones. In vivo labeling of poly(ADP-ribose) was performed in one of these subclones, revealing that the level of poly(ADP-ribose) accumulation after the same treatment with N-methyl-N'-nitro-N-nitrosoguanidine (MNNG) was four times higher in the human PARP-expressing subclone compared with both non-expressing transfected control cells and parental cells. Clonal survival assays revealed a sensitization upon treatment with gamma-radiation (up to 1.4-fold) or MNNG (up to 2.7-fold) of several subclones expressing human PARP; in some others survival was not changed. Survival after cisplatin (DDP) treatment remained essentially unchanged. A protective effect against DNA-damage was never observed. We conclude that human PARP overexpression in rodent cells leads to increased poly(ADP-ribosyl)ation capacity and does not promote survival after gamma-radiation or treatment with the DNA-damaging agents MNNG or DDP.

Combination effects of poly(ADP-ribose) polymerase inhibitors and DNA-damaging agents in ovarian tumor cell lines--with special reference to cisplatin.

The effects of the poly(ADP-ribose) polymerase inhibitors 4-amino-1,8-naphthalimide (4-ANI), 6(5H)-phenanthridinone (PHD), 1,5-isoquinolinediol (IQD), 3-aminobenzamide (3-AB) or 4-hydroxyquinazoline (4-HYA) on the cytotoxicity of cisplatin were investigated. The human ovarian tumor cell lines SK-OV-3 and OAW 42 and the rat ovarian tumor cell line O-342 as well as its cisplatin (DDP)-resistant subline O-342/DDP were used. Cytotoxicity was determined with the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay. 1-Methyl-3-nitro-1-nitrosoguanidine (MNNG) plus its respective combinations with poly(ADP-ribose) polymerase inhibitors served as positive controls. In addition, the alkylating agents L-threitol-1,4-bismethanesulfonate (DHB) and 1,3-bis(2-chloroethyl)-1-nitrosourea (carmustine) as well as two other DNA-repair inhibitors caffeine and theophylline were included in the investigations. The cytotoxicity of cisplatin could not be increased by 4-ANI, PHD, IQD, 4-HYA or 3-AB in any cell line investigated, while it was increased by caffeine in lines O-342/DDP and SK-OV-3 as well as by theophylline in lines O-342/DDP, SK-OV-3 and OAW 42. The cytotoxicity of MNNG was increased by combination with 4-ANI, PHD, IQD, 4-HYA, 3-AB or theophylline for all lines except OAW42; in the latter line, only 4-ANI, PHD and IQD increased MNNG cytotoxicity. The cytotoxicity of DHB was increased by 4-ANI, PHD, 4-HYA, theophylline and caffeine in line O-342/DDP; by 4-HYA, theophylline and caffeine in line SK-OV-3; and by theophylline and caffeine in line OAW42. The cytotoxicity of carmustine was increased only by 3-AB in two lines (SK-OV-3 and OAW 42). Results are discussed with regard to different DNA-repair mechanisms.

Ring-substituted diaqua(1,2-diphenylethylenediamine)platinum(II) sulfate reacts with DNA through a dissociable complex.

Ring-substituted diaqua(1,2-diphenylethylenediamine)platinum(II) sulfate shows unusual kinetics in its reaction with salmon testis DNA. The mechanism for diaqua[meso-1,2-bis(2,6-dichloro-4- hydroxyphenyl)ethylenediamine]platinum(II) sulfate, [Pt(H2O)2(meso-6)]2+SO4(2-), a representative of this series, has been investigated and compared with that for cis-[Pt(NH3)2(H2O)2]2+. Reactions were followed by atomic absorption, analytical HPLC of Pt-DNA digests, arrest of enzymatic DNA synthesis/degradation, ultraviolet and fluorescence spectrophotometry. Except for the formation of monofunctional DNA adducts, the kinetics of the platinum(II) complexes are comparable. The pseudo-first-order rate constant for the attack of DNA by [Pt(H2O)2(meso-6)]2+ follows the concentration of DNA in a hyperbolic fashion, which is in contrast to the linear dependence for cis-[Pt(NH3)2(H2O)2]2+. The hyperbolic dependence is typical for a dissociable DNA/drug complex preceding the coordination reaction. By studying the binding of free ligand to DNA, and by correlating ligand structures and electrostatic charges with effects on adduct formation, both the phenyl residues and the positive charge of the platinum(II) complex are shown to be crucial for the stability of the dissociable complex. A non-intercalative mode of binding to the DNA backbone is suggested. At the high concentrations of DNA found in cell nuclei, the reaction of the dissociable complex can, principally, become rate-limiting in the attack of DNA and thus reduce the cytotoxic efficiency of a drug.

Monofunctional DNA-platinum(II) adducts block frequently DNA polymerases.

The question of whether monofunctional DNA platinum(II) adducts block synthesis of DNA by purified DNA polymerases of different types and origin has been investigated by comparing the time dependence of synthesis arrest and of DNA adduct formation. Activated salmon testis DNA is used as a suitable substrate for DNA synthesis allowing to probe inhibition by platinum(II) monoadducts for the variety of inherent template-primers. Reaction amplitudes are related to defined mixtures of dichloro and chloroaqua platinum(II) complexes. It is found that (i) all investigated DNA polymerases seem arrested (100% efficiency) at bifunctional DNA adducts. (ii) human DNA polymerase beta bypasses most of the monofunctional lesions of the three platinum(II) complexes investigated. (iii) Klenow fragment is blocked by monoadducts with increasing efficiency in the order cis-diamminechloroaquaplatinum(II) (0%) less than meso-[1,2-bis(2,6- dichloro-4-hydroxyphenyl)ethylenediamine] chloroaquaplatinum(II) (50%) less than trans-diamminechloro-aquaplatinum(II) (75%). (iv) Escherichia coli DNA polymerase I, Thermus aquaticus DNA polymerase, Physarum polycephalum DNA polymerase alpha, and calf thymus DNA polymerase alpha appear to be arrested by monoadducts. According to these examples, blocking efficiencies depend on the cis/trans-stereogeometry of fixation of the carrier ligands at platinum(II) residues, on the size/chemical nature of the platin(II) carrier ligand and on the type/origin of DNA polymerase.

The reaction of platinum(II) complexes with DNA. Kinetics of intrastrand crosslink formation in vitro.

The kinetics of the formation of bifunctional DNA platinum(II) adducts (DNA-crosslinks) have been investigated by endonuclease digestion and subsequent HPLC analysis of the soluble nucleotides and nucleotide platinum(II) adducts. The results indicate two waves of crosslinking [rate constants (0.2-0.3) min-1 and (0.015-0.025) min-1] that correlate with changes in ultra violet absorbance and ethidium bromide dependent fluorescence intensity, previously interpreted in terms of two consecutive, local conformational rearrangements of platinum-DNA (Schaller, W., Reisner, H., and Holler, E. (1987) Biochemistry 26, 943-950). The formation of crosslinks at sequences d(GpG) and d(GpNpG) follows identical kinetics. A minimal reaction mechanism is proposed for the binding of cis-diamminedichloroplatinum(II) to DNA under in vitro conditions. The approximately 3-fold higher rate for meso-[1,2-bis(2,6-dichloro-4- hydroxyphenyl)ethylenediamine]diaquaplatinum(II) in comparison to the rate for cis-diamminediaquaplatinum(II) indicates that crosslink formation is affected by the nature of the non-leaving platinum ligand(s).

Escherichia coli DNA polymerase I: inherent exonuclease activities differentiate between monofunctional and bifunctional adducts of DNA and cis- or trans-diamminedichloroplatinum(II). An exonuclease investigation of the kinetics of the adduct formation.

[3H]dGMP-3'-labelled, activated salmon testis DNA and [32P]dGMP-5'-labelled open circular M13 DNA were reacted with cis-diamminedichloroplatinum(II), cis-diamminechloroaquaplatinum(II), cis-diamminediaquaplatinum(II) or trans-diamminechloroaquaplatinum(II). The reaction was arrested after arbitrary times by adjustment to slightly alkaline solution conditions. The platinum-containing DNA was digested with Escherichia coli DNA polymerase I. The progress of nucleotide release was measured by acid precipitation of undigested DNA. Solubilized nucleotides and adducts were analyzed by HPLC. The 3'-5'-exonuclease activity liberated single-coordinated dGMP-platinum(II) adducts from both cis- and trans-platinum(II) treated salmon testis DNA and a small fraction of adducts of cis-platinum(II) that coordinated two molecules of dGMP. The bisadduct was derived from non-neighboring guanine residues probably located at or close to 3'-termini. This nuclease activity neither cut between nor after neighboring guanine residues crosslinked by cis-platinum(II). No bisadduct was liberated for trans-platinum(II). The 5'-3'-exonuclease activity did not liberate any nucleotide adducts from cis-platinum(II)-treated DNa. However, it removed single-coordinated guanine adducts of trans-diamminedichloroplatinum(II). From the kinetics of the appearance of dGMP monoadducts and the inhibition of digestion, a reaction scheme is formulated for the reaction of platinum(II) complexes with DNA that confirms and extends the previously published one [W. Schaller, H. Reisner & E. Holler (1987) Biochemistry 26, 943-950]. The longevity of the dGMP monoadduct intermediate is discussed in the context of the efficiency of cis-diamminedichloroplatinum(II) as an antitumor drug.

Effects of coordination of diammineplatinum(II) with DNA on the activities of Escherichia coli DNA polymerase I.

The effects of the reaction of cis- and trans-diamminedichloroplatinum(II) with DNA have been measured with regard to DNA synthesis, 3'-5' exonuclease (proofreading), and 5'-3' exonuclease (repair) activities of Escherichia coli DNA polymerase I. Both isomers inhibit DNA synthetic activity of the polymerase through an increase in Km values and a decrease in Vmax values for platinated DNA but not for the nucleoside 5'-triphosphates as the varied substrates. The inhibition is a consequence of lowered binding affinity between platinated DNA and DNA polymerase, and of a platination-induced separation of template and primer strands. Strand separation enhances initial rates of 3'-5' excision of [3H]dCMP from platinated DNA (proofreading), while total excision levels of nucleotides are decreased. In contrast to proofreading activity, the 5'-3' exonuclease activity (repair) discriminates between DNA which had reacted with cis- and with trans-diamminedichloroplatinum(II). While both initial rates and total excision are inhibited for the cis isomer, they are almost not affected for the trans isomer. This differential effect could explain why bacterial growth inhibition requires much higher concentrations of trans- than cis-diamminedichloroplatinum(II).