Regulation of malignant phenotype and bioenergetics by a π-electron donor-inducible mitochondrial MgATPase.

The recognition of poly ADP-ribose transferase-1 (PARP-1) as an ATP sensor receiving this energy source by way of a specific adenylate kinase ATP wire (AK) from mitochondrial ATP synthase (F0F1), and directly regulating cellular mRNA and DNA synthesis, was the first step towards the identification of an effect by PARP-1 that is of fundamental significance. The molecular target of AK-ATP is Arg 34 of the Zn finger I of PARP-1, which is also a site for cation-π interactions as a target of π-electron donors. We now identify this π-electron receptor site as the second active center of PARP-1 which by interaction with a π-electron donor-inducible MgATPase reversibly controls a malignant vs. non-malignant phenotype through energizing the NADH➝NADP+ transhydrogenase, a reaction which is the metabolic connection of PARP-1 to cell function. The specific enzyme-inducing action of the π-electrons is executed by the PARP-1 -topoisomerase I - DNA complex of the nuclei regulating both the nature and the quantity of cellular enzymes that constitute cell-specific physiology.

Enzymatic mechanism of the tumoricidal action of 4-iodo-3-nitrobenzamide.

Activation of the prodrug 4-iodo-3-nitrobenzamide critically depends on the cellular reducing system specific to cancer cells. In non-malignant cells, reduction of this prodrug to the non-toxic amine occurs by the flavoprotein of complex?I of mitochondria receiving Mg2+-ATP-dependent reducing equivalents from NADH to NADPH via pyridine nucleotide transhydrogenation. This hydride transfer is deficient in malignant cells; therefore, the lethal synthesis of 4-iodo-3-nitrosobenzamide takes place selectively. Enzymatic evidence for this mechanism has been provided by cellular studies with lysolecithin-permeabilized cells and cell fractions, which have identified the defect in transhydrogenation in neoplastic cells to be located at the energy transfer site. Confirming previous results, the present study demonstrates the validity of the selective tumoricidal action of the prodrug in cell cultures.

Dependence of trans-ADP-ribosylation and nuclear glycolysis on the Arg 34-ATP complex of Zn2+ finger I of poly-ADP-ribose polymerase-1.

The H-bonded complex of ATP with Arg 34 of Zn2+ finger I of poly-ADP-ribose polymerase-1 (PARP-1) determines trans-oligo-ADP-ribosylation from NAD+ to proteins other than PARP-1. This mechanism was tested in lysolecithin fractions of non-malignant and cancer cells separately and after their recombination. Cellular PARP-1 activity was recovered when the centrifugal sediment was recombined with the supernatant fraction containing cellular ADP-ribose oligomer acceptor proteins. Combination of the matrix fraction (Mx) of cancer cells (lacking OXPHOS) with its supernatant had the same PARP-1 activity as the Mx alone. The supernatant of non-malignant cells was replaced by glycolytic enzymes as ADP-ribose acceptor. The hexokinase activity of the supernatant increased when OXPHOS of intact cells was uncoupled by carbonyl cyanide 4-(trifluoro methoxy) phenylhydrazone. trans-ADP-ribosylation was demonstrated by polyacrylamide gel electrophoresis.

Identification of poly(ADP-ribose) polymerase-1 as the OXPHOS-generated ATP sensor of nuclei of animal cells.

Our results show that in the intact normal animal cell mitochondrial ATP is directly connected to nuclear PARP-1 by way of a specific adenylate kinase enzymatic path. This mechanism is demonstrated in two models: (a) by its inhibition with a specific inhibitor of adenylate kinase, and (b) by disruption of ATP synthesis through uncoupling of OXPHOS. In each instance the de-inhibited PARP-1 is quantitatively determined by enzyme kinetics. The nuclear binding site of PARP-1 is Topo I, and is identified as a critical "switchpoint" indicating the nuclear element that connects OXPHOS with mRNA synthesis in real time. The mitochondrial-nuclear PARP-1 pathway is not operative in cancer cells.

Quantitative correlation between cellular proliferation and nuclear poly (ADP-ribose) polymerase (PARP-1).

Treatment of cells with lysophosphatidyl choline and centrifugal extraction can separate poly (ADP-ribose) synthetase (PARP-1) and DNA synthetase activities, permitting the experimental analysis and comparison of both multienzyme systems. Only PARP-1 is being assayed by our system. Ca(2+) and Mg(2+) have minor activating effects, and added histones are without activating action. Short end-blocked dsDNAs at nM concentrations and spermine at mM concentrations are maximally activating coenzymes of poly (ADP-ribose) synthesis. Comparison of non-proliferating non-malignant cells with rapidly growing cancer cells demonstrates that rates of poly (ADP-ribose) synthesis and DNA synthesis are highest in pre-confluent non-malignant cells and in proliferating cancer cells, and lowest in contact-inhibited non-malignant cells. Rates of poly (ADP-ribose) synthesis correlate with the number of enzymatically activable PARP-1 molecules per cell, determined under Vmax conditions where activity is linearly proportional to enzyme protein. Contact-inhibited non-malignant cells exhibit only trans-ADP-ribosylation that is not affected by ATP, while rapid growth, especially in cancer cells, demonstrates extensive auto-poly (ADP)-ribosylation that is strongly inhibited by ATP at concentrations present in cells exhibiting normal bioenergetics. Rates of mRNA synthesis in non-proliferating non-malignant cells and in cancer cells were indistinguishable, indicating that the differences observed between cellular phenotypes are most likely due to reassembly of PARP-1 molecules in nuclei to homo-dimers (in cancer cells) and hetero-dimers (in non-cancer cells). A specific inhibitor and an inactivator of PARP-1 each inhibit DNA synthesis when intact cancer cells are pretreated with these drugs. Direct addition of these drugs to permeabilized cells performing DNA synthesis has no effect on DNA synthesis. The most striking diagnostic signal for cancer cells is activation of PARP-1 and of DNA synthesis.

The influence of ATP on poly(ADP-ribose) metabolism.

ATP affects poly(ADP-ribose) metabolism at two distinct sites: it inhibits poly(ADP-ribose) polymerase-1 and activates the glycohydrolase directly. The inhibitory site of ATP on poly(ADP-ribose) polymerase-1 was identified by amino acid exchange mutation to be at the arginine 34 residue in the first Zn2+ finger. Mutation of 138 arginine residue of Zn2+ finger 2 had negligible influence on the inhibitory action of ATP, pinpointing arginine 34 of the first Zn2+ finger as the specific ATP site. The glycohydrolase protein was activated by ATP when the substrate was a long-chain ADP-ribose polymer, but not with a short-chain substrate. Isolated cell nuclei also responded to both inhibition of poly(ADP-ribose) polymerase by ATP and to poly(ADP-ribose) glycohydrolase activation by ATP, demonstrating that enzymological results can be extrapolated to cellular systems. The activation of poly(ADP-ribose) polymerase in nuclei by an alkylating drug was completely suppressed by ATP, demonstrating that the bioenergetic competence of cells can regulate the cytocidal action of DNA alkylating drugs. The potential significance of bioenergetic regulation of poly(ADP-ribose) metabolism is proposed.

Mechanisms of antitumor action of methyl-3,5-diiodo-4-(4'-methoxyphenoxy)benzoate: drug-induced protein dephosphorylations and inhibition of the permissive action of ceramide on growth factor induced cell proliferation.

The tumoricidal mechanism of methyl-3,5-diiodo-4-(4'-methoxypropoxy)benzoate (DIME), or DIPE, has been analyzed beyond its first recognized cellular site, which is the inhibition of tubulin polymerization. DIME (or DIPE) pretreatment of Eras cells for 3 days abolished ceramide basic fibroblast growth factor (bFGF)-induced glycolysis, coinciding with a block produced by the phosphoprotein dephosphorylation of cdc 25 by protein phosphatase 2A (PP2A). Protein dephosphorylation is directly activated by DIME (or DIPE), and enzyme activities that are dependent on P-proteins are significantly down-regulated (e.g. Topo I and II, MAP-kinase, and cdc-cyclin kinase). Purified PP2A is one target of activation by DIME (or DIPE), and an alkaline phosphatase isoenzyme is also induced by the drug. It is proposed that the pleiotropic effects of DIME (or DIPE) on cancer cells involve the activation of protein dephosphorylations, as well as inhibition of tubulin polymerization.

Activity assays for poly-ADP ribose polymerase.

Poly(ADP-ribose) polymerase (PARP-1) is a nuclear enzyme that has traditionally been thought to require discontinuous or "damaged" DNA (dcDNA) as a coenzyme, a preconception that has limited research mainly to its role in cell pathology, i.e., DNA repair and apoptosis. Recent evidence has shown that this enzyme is broadly involved in normal cell physiological functions including chromatin modeling and gene regulation when DNA strand breaks are absent. We have recently shown that double-stranded DNA (dsDNA) serves as a more efficient coenzyme for PARP-1 than dcDNA, providing a mechanistic basis for PARP-1 function in normal cell physiology. Here we provide a detailed outline of methods for analyzing PARP-1 enzymatic activity using dsDNA as a coenzyme compared with broken or damaged DNA. Two procedures are described, one for analysis of auto-, and the other for trans-ADP-ribosylation. These assays provide a means of investigating the physiological role(s) of PARP-1 in normal cells.

Regulation of the enzymatic catalysis of poly(ADP-ribose) polymerase by dsDNA, polyamines, Mg2+, Ca2+, histones H1 and H3, and ATP.

The enzymatic mechanism of poly(ADP-ribose) polymerase (PARP-1) has been analyzed in two in vitro systems: (a) in solution and (b) when the acceptor histones were attached to a solid surface. In system (a), it was established that the coenzymatic function of dsDNAs was sequence-independent. However, it is apparent from the calculated specificity constants that the AT homopolymer is by far the most effective coenzyme and randomly damaged DNA is the poorest. Rates of auto(poly-ADP-ribosylation) with dsDNAs as coenzymes were nearly linear for 20 min, in contrast to rates with dcDNA, which showed product [(ADPR)n] inhibition. An allosteric activation of auto(poly-ADP-ribosylation) by physiologic cellular components, Mg2+, Ca2+, and polyamines, was demonstrated, with spermine as the most powerful activator. On a molar basis, histones H(1) and H(3) were the most effective PARP-1 activators, and their action was abolished by acetylation of lysine end groups. It was shown in system (b) that oligo(ADP-ribosyl) transfer to histone H(1) is 1% of that of auto(poly-ADP-ribosylation) of PARP-1, and this trans(ADP-ribosylation) is selectively regulated by putrescine (activator). Physiologic cellular concentrations of ATP inhibit PARP-1 auto(poly-ADP-ribosylation) but less so the transfer of oligo(ADP-ribose) to histones, indicating that PARP-1 auto(ADP-ribosylation) activity is dormant in bioenergetically intact cells, allowing only trans(ADP-ribosylation) to take place. The inhibitory mechanism of ATP on PARP-1 consists of a noncompetitive interaction with the NAD site and competition with the coenzymic DNA binding site. A novel regulation of PARP-1 activity and its chromatin-related functions by cellular bioenergetics is proposed that occurs in functional cells not exposed to catastrophic DNA damage.

Analysis of nucleotide sequence-dependent DNA binding of poly(ADP-ribose) polymerase in a purified system.

The enzymatic transfer of ADP-ribose from NAD to histone H(1) [defined as trans(oligo-ADP-ribosylation)] or to PARP-1 [defined as auto(poly-ADP-ribosylation)] requires binding of coenzymic DNA. The preceding paper [Kun, E., et al. (2004) Biochemistry 43, 210-216] shows that oligonucleotides of dsDNA can serve as coenzymic DNA for PARP-1 trans- or auto-modification activity. Results of DNA-protein binding (EMSA) experiments reported here demonstrate that short DNA oligonucleotides containing the 5'-TGTTG-3' nucleotide sequence motif preferentially bind to cloned PARP-1 in vitro. The same nucleotide sequence motif is responsible for striated myocyte-selective transcription of a contractile protein gene encoding cardiac troponin T (cTnT). Results of experiments reported here demonstrate that mutation of this motif also abolishes the differentiation-dependent activation of the transfected cTnT promoter in myoblasts cultured in vitro, indicating that nucleotide sequence-dependent binding of PARP-1 to promoter DNA of the cTnT gene is also necessary for differentiation-dependent activation. Thus, PARP-1 has two types of dsDNA binding activity: (1) nucleotide sequence-dependent binding, analyzed here with EMSA experiments, and (2) coenzymic binding, measured catalytically, which does not depend on the nucleotide sequence of the dsDNA. We hypothesize that the well-known association of PARP-1 with chromatin can be attributed to its stable binding to chromosomal dsDNA, some portion of which is likely to be nucleotide sequence-dependent binding. According to this hypothesis, the distribution of this protein-modifying enzyme in chromatin may be targeted to specific genomic loci and vary according to cell type and developmental stage.

Synergistic anticancer action of reversibly and irreversibly acting ligands of poly (ADP-ribose) polymerase.

The synergistic interaction of two ligands (INH2BP and the prodrug INO2BA) of PARP I has been demonstrated for two human leukemia cell lines (855-2 and HL-60), for a human lung cancer cell (A549) and for Eras 20 cancer cells. Synergism was calculated using kinetic combination constants based on cell multiplication rates. Reducing cellular GSH content by BSO strongly augmented synergism, an effect partly explained by the removal of C-NO scavenging (by GSH). However, INH2BP action was augmented by BSO, an effect most probably explained by the sensitization of the cell to apoptosis by GSH removal.

Coenzymatic activity of randomly broken or intact double-stranded DNAs in auto and histone H1 trans-poly(ADP-ribosylation), catalyzed by poly(ADP-ribose) polymerase (PARP I).

The enzymatic transfer of ADP-ribose from NAD to histone H1 (defined as trans-poly(ADP-ribosylation)) or to PARP I (defined as auto-poly(ADP-ribosylation)) was studied with respect to the nature of the DNA required as a coenzyme. Linear double-stranded DNA (dsDNA) containing the MCAT core motif was compared with DNA containing random nicks (discontinuous or dcDNA). The dsDNAs activated trans-poly(ADP-ribosylation) about 5 times more effectively than dcDNA as measured by V(max). Activation of auto-poly(ADP-ribosylation) by dcDNA was 10 times greater than by dsDNA. The affinity of PARP I toward dcDNA or dsDNA in the auto-poly(ADP-ribosylation) was at least 100-fold lower than in trans-poly(ADP-ribosylation) (K(a) = 1400 versus 3-15, respectively). Mg2+ inhibited trans-poly(ADP-ribosylation) and so did dcDNA at concentrations required to maximally activate auto-poly(ADP-ribosylation). Mg2+ activated auto-poly(ADP-ribosylation) of PARP I. These results for the first time demonstrate that physiologically occurring dsDNAs can serve as coenzymes for PARP I and catalyze preferentially trans-poly(ADP- ribosylation), thereby opening the possibility to study the physiologic function of PARP I.

Anti-cancer action of 4-iodo-3-nitrobenzamide in combination with buthionine sulfoximine: inactivation of poly(ADP-ribose) polymerase and tumor glycolysis and the appearance of a poly(ADP-ribose) polymerase protease.

E-ras 20 tumorigenic malignant cells and CV-1 non-tumorigenic cells were treated with a drug combination of 4-iodo-3-nitrobenzamide (INO(2)BA) and buthionine sulfoximine (BSO). Growth inhibition of E-ras 20 cells by INO(2)BA was augmented 4-fold when cellular GSH content was diminished by BSO, but the growth rate of CV-1 cells was not affected by the drug combination. Analyses of the intracellular fate of the prodrug INO(2)BA revealed that in E-ras 20 cells about 50% of the intracellular reduced drug was covalently protein-bound, and this binding was dependent upon BSO, whereas in CV-1 cells BSO did not influence protein binding. Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was identified as the protein that covalently binds the reduction product of INO(2)BA, which is 4-iodo-3-nitrosobenzamide. Since only the enzymatically reduced drug INOBA bound covalently to GAPDH, the BSO-dependent covalent protein-drug association indicated an apparent nitro-reductase activity present in E-ras 20 cells, but not in CV-1 cells, explaining the selective toxicity. Covalent binding of INOBA to GAPDH inactivated this enzyme in vitro; INO(2)BA+BSO also inactivated cellular glycolysis in E-ras 20 cells because it provided the precursor to the inhibitory species: INOBA. Another event that occurred in INO(2)BA+BSO-treated E-ras 20 cells was the progressive appearance of a poly(ADP-ribose) polymerase protease. This enzyme was partially purified and characterized by the polypeptide degradation product generated from PARP I, which exhibited a 50kDa mass. This pattern of proteolysis of PARP I is consistent with a drug-induced necrotic cell killing pathway.