Binding of cis- and trans-diamminedichloroplatinum(II) to deoxyribonucleic acid exposes nucleosides as measured immunochemically with anti-nucleoside antibodies

Slipped (CTG).(CAG) repeats of the myotonic dystrophy locus: surface probing with anti-DNA antibodies

At least 15 human diseases have been associated with the length-dependent expansion of gene-specific (CTG).(CAG) repeats, including myotonic dystrophy (DM1) and spinocerebellar ataxia type 1 (SCA1). Repeat expansion is likely to involve unusual DNA structures. We have structurally characterized such DNA, with (CTG)(n).(CAG)(n) repeats of varying length (n=17-79), by high-resolution gel electrophoresis, and have probed their surfaces with anti-DNA antibodies of known specificities. We prepared homoduplex S-DNAs, which are (CTG)x.(CAG)y where x=y, and heteroduplex SI-DNAs, which are hybrids where x>y or x<y. S-DNAs formed many different species of slipped isomers, as indicated by its multiple electrophoretic species. In contrast, SI-DNAs formed distinct structures, as indicated by the limited electrophoretic species for all possible repeat length pairings. Sister SI-DNAs with an excess of CAG repeats always migrated slower than their sister SI-DNAs with an excess of CTG repeats. Strikingly, both the propensity to form slipped structures and the pattern of S-DNAs, but not SI-DNAs, varied for similar lengths of CTG/CAG repeats between the DM1 and SCA1 loci, highlighting a role for flanking cis-elements in S-DNA but not SI-DNA formation. Slipped structures bound structure and nucleotide-specific anti-DNA antibodies. Binding of anti-B-DNA antibodies was reduced for both S-DNAs and SI-DNAs relative to their linear forms. SI-DNAs bound anti-Z-DNA antibodies, while both S and SI-DNAs bound anti-cruciform antibodies, revealing shared characteristics between the corresponding DNA structures and slipped DNAs. Such features of the repeats may be recognized by cellular proteins known to bind such structures.

Cisplatin DNA adduct detection and depurination measured by 32P DNA radiolabeling and two-dimensional thin-layer chromatography: a time and concentration study

Platinum-based chemotherapies cause the formation of DNA adducts and have profound effects on DNA. This study measured cis-diamminedichloroplatinum II (cisplatin) DNA adducts by 32P-radiolabeling DNA, enzymatically digesting radiolabeled DNA, separating the formed adducts on two-dimensional thin-layer chromatography, and quantitating the adducts with autoradiography and densitometry. HeLa DNA was incubated with cisplatin at varying concentrations (6.25-325 nM) and times (0 min to 72 hr). Cisplatin rapidly depurinated dGMP and dAMP (90%, 15-min incubation with 325 nM cisplatin). Partial depurination of dGMP (15%) and dAMP (25%) occurred with lower cisplatin concentrations at equal incubation times. A minimum of four new adducts, with relatively rapid migratory patterns, were detected at high cisplatin concentrations with short incubation times. These results indicate that the depurination of DNA correlates with DNA adduct formation and that the quantification of these adducts may be applicable to monitoring tumor and host cell response to cisplatin chemotherapy.

Mutagenic and carcinogenic properties of platinum-based anticancer drugs

Cisplatin (DDP) is currently one of the most effective drugs for the treatment of cancer. It causes primarily intrastrand DNA-DNA cross-links, and is highly mutagenic and carcinogenic in both in vitro and in vivo experimental models. There is, however, considerable variability between the response seen in different cellular systems, probably at least partly because of the different cellular DNA repair capacities. A number of analogues of cisplatin have been developed and one of these, carboplatin (CDDCA), is also in widespread clinical use. Although it is somewhat less toxic, there is no evidence that its mode of action differs from that of cisplatin. A limited amount of mutagenicity data suggests that it has similar mutagenic and carcinogenic consequences as the parent drug. Many further analogues of cisplatin are now in clinical trials, and some of these appear to have different DNA repair responses (and therefore possibly the development of clinical resistance). Although some (e.g., iproplatin and spiroplatin) are less mutagenic than either cisplatin or carboplatin, these appear to be the ones least likely to achieve wide use. There are insufficient data on several of the most promising clinical analogues (e.g., DWA2114R and ACDDP) to judge their relative mutagenic and carcinogenic potential. Detailed studies on the DNA repair and mutagenicity characteristics of these compounds will not only provide clinically relevant data, but may also aid in the selection of further useful antitumour agents in this series.

Conversion of DNA adducts of antitumour cis-diamminedichloroplatinum(II). Immunochemical analysis

Polyclonal antibodies that bind selectively to DNA modified by antitumour cisplatin and its analogues were isolated. The reactivity of the antibodies with the epitope was enhanced by thermal denaturation of DNA that had been modified by cisplatin before its denaturation. On the other hand, denaturation of DNA before its modification resulted in considerably less reaction of the antibodies. The conversion of monofunctional cisplatin-DNA adducts to bifunctional lesions increased the capability of the modified DNA to competitively inhibit the antibodies. The double-helical oligonucleotides containing a unique bifunctional adduct formed by cisplatin at the d(GG) site cross-reacted with the antibodies in contrast to the oligonucleotide containing a single monofunctional adduct formed at the d(G) site. In addition, poly(dG-dC) . poly(dG-dC) modified by cisplatin did not react with the antibodies. It was concluded that the antibodies recognized monodentate lesions, intrastrand cross-links between two purine nucleosides separated by one or more nucleosides and interstrand cross-links negligibly. The antibodies apparently recognized a chemical nature of the bifunctional adduct formed between two adjacent purines and not an unusual conformational feature of DNA resulting from the formation of this adduct. The antibodies were used to analyse the adducts formed by cisplatin on DNA of cultured cells exposed to this drug. During the subsequent incubation of the already exposed cells in the drug-free medium, a part of the bifunctional adducts of cisplatin was completely removed from DNA or transformed to the adducts not recognized by the antibodies.

Changes in the cytoskeleton pattern of tumor cells by cisplatin in vitro

The influence of the antitumor drug cis-diamminedichloroplatinum(II) ('cisplatin') upon the structural pattern of the main cytoskeletal components, i.e. microtubules, intermediate filaments and microfilaments, was investigated in squamous carcinoma cells derived from the mouse stomach (G 22) or the human lung (L 266) and growing in vitro as monolayer cultures. The studies were performed by the indirect immunofluorescence technique using monoclonal antibodies against alpha-tubulin, type 19 cytokeratin and actin at the end of a 90-min exposure to 2.5 x 10(-6), 5 x 10(-6) or 10(-5) mol cisplatin/l and a subsequent 24-h recovery period. Under the influence of cisplatin, the cytoskeletal tubules and filaments, which were distributed in untreated cells as a finely organized network spreading through the whole cytoplasm like a spider's web, collapsed and aggregated to dense and circularly arranged bands of bright immunofluorescence around the nucleus or to cap-like structures apposing the nucleus. These phenomena developed in clear dependence upon the dose of cisplatin applied and were observable in a modified manner and to a different degree with the three structural elements of the cytoskeleton. During the subsequent 24-h interval, during which the cells were allowed to recover in drug-free growth medium, the before-mentioned collapse of the cytoskeletal network was only partially reversible following previous treatment with the medium (5 x 10(-6) mol/l) and the high (10(-5) mol/l) dose of cisplatin and restored totally to the normal structural pattern of untreated control cells when the low dose of 2.5 x 10(-6) mol cisplatin/l had been administered before. These results give evidence that the DNA cannot be the only cellular target for the antitumor drug cisplatin, but that it also effects other intracellular lesions which cause structural alterations of cellular organelles independently of the primary molecular attack at nuclear DNA strands. Probably, these additional interactions fortify the antiproliferative effect and contribute to the achievement of important biological and cytological effects of cisplatin such as growth inhibition or giant cell formation.

Probing of DNA structure with osmium tetroxide,2,2'-bipyridine. Adduct-specific antibodies

Antibodies against DNA modified with a single-strand selective probe, OsO4 in complex with 2,2'-bipyridine (Os,bipy), were raised in rabbits. These antibodies were fractionated using affinity column chromatography and fractions S89-II and S89-III characterized as highly specific for DNA-Os,bipy adduct with no cross reactivity to at least 1000-fold excess of unmodified DNA, RNA and Os,bipy-modified and unmodified proteins. Cross-reactivity to Os,bipy-modified RNA was very small. S89-II showed no cross-reactivity to DNA modified with OsO4 complexed with tetramethylethylenediamine or with bathophenanthroline disulphonic acid and to DNA oxidized with KMnO4. It cross-reacted, however, with DNA modified with OsO4,1,10-phenanthroline complex. The limit of detection of immunodot-blot analysis of extensively Os,bipy-modified DNA was below 0.5 pg. Small extent of Os,bipy-modification of supercoiled and linearized plasmids can be detected by DNA gel retardation and immunoblotting techniques. E. coli cells contain DNA regions in which bases are accessible to the single-strand selective probe.

Polyclonal antibodies elicited against trans-diamminedichloroplatinum(II)-modified DNA and their application in nonradioactive hybridization experiments

Polyclonal antibodies that bind selectively to adducts formed with DNA by chemotherapeutically inactive trans-diamminedichloroplatinum (II) [trans-[Pt(NH3)2Cl2]) were produced by immunization with calf-thymus double-helical DNA modified by trans-[Pt(NH3)2Cl2] at a ratio of bound platinum/nucleotide (rb) of 0.1. High selectivity was obtained by separation of the antibodies from the antiserum with the aid of affinity chromatography on a Sepharose column. The antibodies were competitively inhibited in an ELISA assay by 25 pM trans-[Pt(NH3)2Cl2] bound to double-helical DNA and 2.5 pM trans-[Pt(NH3)2Cl2] bound to denatured DNA (rb = 0.1). The conversion of monofunctional adducts, formed on DNA at the early stage of its interaction with trans-[Pt(NH3)2Cl2], to bifunctional lesions, decreased the ability of the modified DNA to competitively inhibit these antibodies. They did not cross-react with unmodified, denatured DNA, but they reacted with diethylenetriamine-chloroplatinum(II)-chloride-modified double-helical DNA and with double-helical DNA treated with cis-diamminedichloroplatinum(II) for a short time (10 min). The results of this work best fit a model in which one of the major antigenic determinants of double-helical DNA modified by trans-[Pt(NH3)2Cl2] is the platinum atom coordinated in a monodentate or bidentate manner with non-paired nucleotide residues or perhaps a short segment of single-stranded DNA which occurs around the platination site. Nucleic acids modified by trans-[Pt(NH3)2Cl2] can be used as immuno-probes in hybridization experiments.

Decreased stability of DNA in cells treated with alkylating agents

A modified highly sensitive procedure for the evaluation of DNA damage in individual cells treated with alkylating agents is reported. The new methodology is based on the amplification of single-strandedness in alkylated DNA by heating in the presence of Mg2+. Human ovarian carcinoma cells A2780 were treated with nitrogen mustard (HN2), fixed in methanol, and stained with monoclonal antibody (MOAB) F7-26 generated against HN2-treated DNA. Binding of MOAB was measured by flow cytometry with indirect immunofluorescence. The maximal difference in fluorescence between untreated and HN2-treated cells was observed after heating at 100 degrees C for 5 min in PBS containing 1.25 mM MgCl2. Higher concentrations of MgCl2 inhibited MOAB binding to HN2-treated cells and heating at lower concentrations induced binding to control cells. Intensive binding of MOAB to control and drug-treated cells was observed after heating in Tris buffer supplemented with MgCl2. Thus, the presence of phosphates and MgCl2 during heating was necessary for the detection of HN2-induced changes in DNA stability. Fluorescence of HN2-treated cells decreased to background levels after treatment with single-strand-specific S1 nuclease. MOAB F7-26 interacted with single-stranded regions in DNA and did not bind to dsDNA or other cellular antigens. Specific reactivity of MOAB F7-26 with deoxycytidine was established by avidin-biotin ELISA. Single-stranded conformation was necessary for the binding of MOAB to deoxycytidine on the DNA molecule. It is suggested that alkylation of guanines decreased the stability of the DNA molecule and increased the access of MOAB F7-26 to deoxycytidines on the opposite DNA strand.

Base-specific binding of adenosine antibodies to double-stranded DNA

Antibodies raised against adenosine have been reported to react with single-stranded DNA but not with double-stranded DNA. Using a highly sensitive avidin-biotin microELISA we report that these antibodies also react with double-stranded DNA. The binding was specific as it was completely inhibited by the homologous hapten. The results indicate that the antibody populations binding to ssDNA and dsDNA are not cross-reactive. The antibodies were shown to react with the topoisomers of plasmid DNA as assessed by gel retardation assay. The antibodies showed differential binding to restriction fragments of DNA indicating that some of the A residues in dsDNA are accessible to the antibodies.

Antibodies against (1R,2R)-cyclohexanediamineplatinum(II)-DNA adduct recognize the conformational differences of isomeric analogues of cyclohexanediamine

Antibodies reactive to (1R,2R)-cyclohexanediamineplatinum(II)-DNA ((1R,2R)-cyclohexanediamine: 1R,2R-dach) adducts were elicited by immunization of rabbit with calf thymus DNA modified by Pt(1R,2R-dach)Cl2 at a ratio of bound platinum per nucleotide ((D/N)b) of 0.0335. In an enzyme-linked immunosorbent assay (ELISA), the binding of specific antibodies to Pt(1R,2R-dach)-DNA adduct (60 microliters of 1.235 x 10(-7) M Pt in each wells) on the assay plate was competitively inhibited by Pt(1R,2R-dach)-DNA adduct ((D/N)b = 0.0653) in the solution. Almost equal inhibition was observed with Pt(1S,2S-dach)-DNA ((D/N)b = 0.0412), an optical isomer of 1R,2R-dach. Pt(1R,2S-dach)-DNA ((D/N)b = 0.0371) and Pt(1R,3S-dach)-DNA ((D/N)b = 0.0281) in which the cyclohexane ring is stereochemically perpendicular to the platinum chelate plane, also inhibited antibody binding, but these adducts gave only incomplete inhibition at higher Pt-DNA adduct concentrations. Although Pt(1R,2R-dach)-d(GpG) and Pt(1R,2R-dach)(NH3)2 inhibited antibody binding, the affinity of the antibody for Pt(1R,2R-dach)(NH3)2 was lower than with Pt(1R,2R-dach)-DNA, and the inhibition behavior of Pt(1R,2R-dach)-d(GpG) was biphasic, i.e., at the lower concentration the inhibition curve was consistent with that of Pt(1R,2R-dach)-DNA, but at the higher concentration it shifted to that of Pt(1R,2R-dach)(NH3)2. The affinity of the antibody for cis-DDP was markedly lower than with Pt(1R,2R-dach)(NH3)2. These facts suggest that the antibodies may bind to the substituents (the platinum and its surroundings) of the various Pt complexes rather than the DNA structure altered by platinum binding.

Biophysical studies of the modification of DNA by antitumour platinum coordination complexes

Cisplatin (cis-diamminedichloroplatinum(II] is widely used in the treatment of various human tumours. A large body of experimental evidence indicates that the reaction of cisplatin with DNA is responsible for the cytostatic action of this drug. Several platinum-DNA adducts have been identified and their effect on the conformation of DNA has been investigated. Structural studies of platinum-DNA adducts now permit a reasonably good explanation of the biophysical properties of platinated DNA. Antitumouractive platinum compounds induce in DNA, at low levels of binding, local conformational alterations which have the character of non-denaturing distortions. It is likely that these changes occur in DNA due to the formation of intrastrand cross-links between two adjacent purine residues. On the other hand, the modification of DNA by antitumour-inactive complexes results in the formation of more severe local denaturation changes. Conformational alterations induced in DNA by antitumour-active platinum compounds may be reparable with greater difficulty than those induced by the inactive complexes. Alternatively, non-denaturation change induced in DNA by antitumour platinum drugs could represent more significant steric hindrance against DNA replication as compared with inactive complexes.

Detection of cells resistant to alkylating agents by flow cytometric analysis of DNA damage

DNA damage was measured by flow cytometric analysis of cells sensitive and resistant to alkylating agents. Human ovarian carcinoma cell line A2780 and a subline which is 7 times more resistant to L-phenylalanine mustard (L-PAM) were treated with the drug, fixed, and stained with monoclonal antibody (MOAB) F7-26 which detects single-stranded regions in alkylated DNA. Mean fluorescent intensity was measured on a flow cytometer. Cells were heated before staining to amplify single-strandedness in alkylated DNA. Significantly larger amount of MOAB was bound to DNA in sensitive than in resistant cells. Fluorescence increased by 80 channels per micrograms L-PAM insensitive cells and only by 17 channels in resistant cells. Sensitive and resistant cells were treated with L-PAM, mixed in different proportions, and stained with MOAB. Populations of sensitive and resistant cells were clearly separated on fluorescence histograms by more than a decade difference in fluorescence intensity. Presence of 2-5% resistant cells was detected among sensitive cells as a separate cell subset. We conclude that staining with MOAB F7-26 can be used as an indicator of cell sensitivity or resistance to alkylating agents. Detection of minor subsets of resistant cells in heterogeneous populations by FCM analysis may be useful for monitoring emerging drug resistance.

Monoclonal antibodies to DNA modified with cis- or trans-diamminedichloroplatinum(II)

Murine monoclonal antibodies that bind selectively to adducts formed on DNA by the antitumor drug cis-diamminedichloroplatinum(II), cis-DDP, or to the chemotherapeutically inactive trans isomer trans-DDP were elicited by immunization with calf thymus DNA modified with either cis- or trans-DDP at ratios of bound platinum per nucleotide, (D/N)b, of 0.06-0.08. The binding of two monoclonal antibodies to cis-DDP-modified DNA was competitively inhibited (50% control) in an enzyme-linked immunosorbent assay (ELISA) by 4-6 nM concentrations (600-900 fmol) of cis-DDP bound to DNA, (D/N)b = 0.031. Similar concentrations (4-6 nM) of cis-DDP-modified poly(dG).poly(dC) also inhibited antibody binding, whereas higher concentrations (17-36 nM) of cis-DDP-modified poly[d(AG)].poly[d(TC)] were required for inhibition. Adducts formed by cis-DDP on other synthetic DNA polymers did not inhibit antibody binding to cis-DDP-DNA. The biologically active compounds [Pt(en)Cl2], [Pt(dach)Cl2], and [Pt(NH3)2(cbdca)] (carboplatin) (where en is ethylenediamine, dach is 1,2-diaminocyclohexane, and cbdca is cyclobutane-1,1-dicarboxylate) all formed antibody-detectable adducts on DNA, whereas the inactive platinum complexes trans-DDP and [Pt(dien)Cl]Cl (dien, diethylenetriamine) did not. The monoclonal antibodies therefore recognize a bifunctional Pt-DNA adduct with cis stereochemistry in which platinum is coordinated by two adjacent guanines or, to a lesser degree, by adjacent adenine and guanine. A monoclonal antibody raised against trans-DDP-DNA was competitively inhibited in an ELISA by 40 nM trans-DDP bound to DNA, (D/N)b = 0.022. This antibody crossreacted with unmodified, denatured DNA. Its binding to trans-DDP-DNA was selectively inhibited by trans-DDP-modified poly[d(GT)].poly[d(CA)] (50% inhibition at 1 nM bound trans-DDP). The recognition of cis- or trans-DDP-modified DNAs by monoclonal antibodies thus parallels the known modes of DNA binding of these compounds and may correlate with their biological activities.

Synthesis and 1H NMR spectroscopic characterization of trans-[Pt(NH3)2[d(ApGpGpCpCpT)-N7-A(1),N7-G(3)]]

The reaction of trans-[Pt(NH3)2Cl2] with the sodium salt of [d(ApGpGpCpCpT)]2 in aqueous solution at 37 degrees C was monitored by reversed-phase high-performance liquid chromatography and UV spectroscopy. Two intermediates, most likely monofunctional adducts, were observed, which subsequently formed one predominant single-stranded product, as well as several polymeric species proposed to be interstrand cross-linked products. The single-stranded adduct was structurally characterized by 1H NMR spectroscopy. From the pH dependence of the chemical shifts, two-dimensional homonuclear chemical shift correlation (COSY) spectroscopy, and one- and two-dimensional nuclear Overhauser effect (NOESY) experiments, the platinum(II) moiety was found to be coordinated to the N7 positions of adenine(1) and guanine(3), with the intervening guanine(2) base destacked from its neighboring residues. This intrastrand 1,3 adduct induces changes in the backbone torsion angles and causes the deoxyribose ring of adenine(1) to switch from a C2'-endo to a predominantly C3'-endo conformation. The other deoxyribose rings retain B DNA type conformations. The structure of trans-[Pt(NH3)2[d(ApGpGpCpCpT)-N7-A(1),N7-G(3)]] differs from those previously reported for cis-DDP 1,2- and 1,3-intrastrand oligonucleotide adducts but is consistent with the structures of trans-DDP 1,3-intrastrand adducts of two previously reported trinucleotides.