Effects of analogs of the DNA minor groove binder Hoechst 33258 on topoisomerase II and I mediated activities.

By contrast with other DNA minor groove binders, Hoechst 33258 inhibited topoisomerase-mediated activity in intact cells. To determine whether specific structural alterations could modify the topoisomerase reactivity of this drug, a series of analogs of Hoechst 33258 (compound 1) was examined. When the relative DNA binding affinities (Ka) of these agents were determined, compound 1 had the highest Ka while agents with substitutions in either of the benzimidazole moieties showed reduced affinity. Whether these changes in DNA binding correlated with topoisomerase inhibitory potency was next examined. In isolated nuclei, 25 microM of agents 1, 5 and 7 reduced VM-26 induced cross-links by 64, 65 and 83%, compared with 15 to 25% reductions by agents 2, 3, 4 and 6, respectively. The structural modification common to the less active compounds was the substitution of an oxygen for nitrogen at either position 1 or 2. On the basis of these results, agents 1, 2, 3 and 7, representing a range of inhibitory potency, were chosen for further analyses. Cross-link induction by m-AMSA and camptothecin in isolated nuclei, as well as by VM-26 in intact cells, was inhibited to a greater extent by agents 1 and 7 than 2 or 3. Additionally, all four drugs inhibited relaxation of pBR 322 DNA induced by both topoisomerases, although topoisomerase I was 2 to 5-fold more sensitive than topoisomerase II. A linear correlation was observed between the logarithms of the Ka value of compounds 1, 2 and 3 and their IC25 values for both topoisomerases, suggesting a strong dependence on DNA binding affinity for enzyme inhibition. Nevertheless, agent 7, despite having less affinity for calf thymus DNA than 1, was the most potent topoisomerase inhibitor tested in intact cells and in isolated enzyme systems. Thus, retention of nitrogen at positions 1 and 2 as well as the addition of nitrogen at position 16 was associated with increased topoisomerase inhibitory potency.

Netropsin and bis-netropsin analogs as inhibitors of the catalytic activity of mammalian DNA topoisomerase II and topoisomerase cleavable complexes.

This study examined the ability of netropsin and related minor groove binders to interfere with the actions of DNA topoisomerases II and I. We evaluated a series of netropsin dimers linked with flexible aliphatic chains of different lengths. These agents are potentially able to occupy longer stretches of DNA than the parental drug as a result of bidentate binding. Both netropsin and its dimers were found: (i) to inhibit the catalytic activity of isolated topoisomerase II and (ii) to interfere with the stabilization of the cleavable complexes of topoisomerase II and I in nuclei. Dimers with linkers consisting of 0-4 and 6-9 methylene groups (n) were far more inhibitory than netropsin against isolated enzyme and in the nuclear system. The compound with n = 5 was less active than netropsin in both assays while the dimer with n = 10 inhibited only the isolated enzyme. The comparison of dimers with fixed linker length (n = 2) but varying number of N-methylpyrrole residues (from 1 to 3) revealed that the inhibitory properties were enhanced with increasing number of N-methylpyrrole units. For dimers with varying linker length, drug ability to inhibit catalytic activity of isolated topoisomerase II was positively correlated with calf thymus DNA association constants. In contrast, no such correlation existed in nuclei. However, the inhibitory effects in the nuclear system were correlated with the association constants for poly(dAdT). The results indicate that bidentate binding can significantly enhance anti-topoisomerase activity of netropsin related dimeric minor groove binders. However, other factors such as the length of the linker, the number of pyrrole moieties and the nature of the target (isolated enzyme/DNA versus chromatin in nuclei) also contribute to these activities.

Role of the environment in the interaction of nonintercalators with Z-DNA.

Interaction of the DNA binding nonintercalators Netropsin, Distamycin and the mPD derivative with Z-DNA has been studied. It has been found that environmental factors like the solvent and added cations significantly modulate the interaction of these ligands with Z-DNA. However no definite Z to B transition in presence of these ligands was found in any case, in contrast to previously reported results (Ch. Zimmer, C. Marck and W. Guschlbauer, FEBS Lett. 154, 156-160 (1983)).

FTIR study of specific binding interactions between DNA minor groove binding ligands and polynucleotides.

The use of FTIR spectroscopy is made to study the interactions between polynucleotides and two series of minor groove binding compounds. The latter were developed and described previously as part of an ongoing program of rational design of modified ligands based on naturally occurring pyrrole amidine antibiotic netropsin, and varying the structure of bisbenzimidazole chromosomal stain Hoechst 33258. Characteristic IR absorptions due to the vibrations of thymidine and cytosine keto groups in polynucleotides containing AT and GC base pairs respectively are used to monitor their interaction with the added ligands. Although the two thiazole based lexitropsins based on netropsin structure differ in the relative orientation of nitrogen and sulfur atoms with respect to the concave edge of the molecules, they interact exclusively with the thymidine C2 = O carbonyl groups in the minor groove of the alternating AT polymer as evidenced by specific changes in the IR spectra. In the second series of compounds based on Hoechst 33258, the structure obtained by replacing the two benzimidazoles in the parent compound by a combination of pyridoimidazole and benzoxazole, exhibits changes in the carbonyl frequency region of poly dG.poly dC which is attributed to the ligand interaction at the minor groove of GC base pairs. In contrast, Hoechst 33258 itself interacts only with poly dA.poly dT. Weak or no interaction exists between the ligands and any of the polynucleotides at the levels of the phosphate groups or the deoxyribose units.

Amide isosteres of lexitropsins: synthesis, DNA binding characteristics and sequence selectivity of thioformyldistamycin.

The synthesis and properties of an amide isostere of the antibiotic distamycin, thioformyldistamycin 3 is described. Compound 3 exists predominantly in the E conformation of the thioamide group in freshly prepared DMSO solution but is converted into the Z form, predicted by molecular mechanics to be more stable, on standing for 24 h. The coalescence temperature in DMSO is 110 degrees C by 1H-NMR. The thioformyl moiety of 3 is resistant to both peptidase action and acid treatment. Complementary strand MPE footprinting on a EcoRI/Hind III restriction fragment of pBR322 DNA demonstrated that either E or Z forms of 3 give a single set of footprints very similar to that of the parent antibiotic with strongest protection at TAAG and TATTAT with moderately strong protection at ATTT and AAAA. The strength of binding of 3 and distamycin from delta Tm measurements to either poly.d(AT) or calf thymus DNA is comparable. Molecular modeling predicted a preferred conformation for 3 wherein the C = S bond has a torsional angle of 110 degrees with the pyrrole ring. The energy difference between this conformation and the E form is less than 1 kcal/mole. In contrast the E-form has an energy 17.3 kcal/mole greater than the Z and a value of 26.3 kcal/mole was calculated for the energy barrier between the two isomers.

Spectrophotometric determination of some phenols with sodium metaperiodate and aminophenols.

New spectrophotometric methods have been developed for the assay of some phenols by use of two pairs of reagents, m-aminophenol and periodate (for catechol, catecholamines, eugenol and guaiacol) and metol and periodate (for pyrogallol, gallic acid, propyl gallate, phloroglucinol and butylated hydroxyanisole). The methods are simple, sensitive, reproducible, accurate within +/- 1.0%, and applicable to the assay of antioxidants (gallic acid, propyl gallate and butylated hydroxyanisole) in oils and fats, catecholamines in dosage forms, and eugenol in clove oil.

DNA sequence selectivities in the covalent bonding of antibiotic saframycins Mx1, Mx3, A, and S deduced from MPE.Fe(II) footprinting and exonuclease III stop assays.

DNA sequence selectivities in the covalent binding of the antitumor antibiotic saframycins Mx1, Mx3, A, and S have been determined by complementary strand MPE.Fe(II) footprinting and exonuclease III stop assays on two different 545 and 135 base pair long HindIII/RsaI restriction fragments of pBR322 DNA. Saframycins Mx1, Mx3, A, and S recognize primarily 5'-GGG sequences. All four antibiotics also recognize 5'-GGPy sequences, however a cytosine is preferred over a thymine at the 3'-end of this recognition site in all cases. Saframycins Mx1, Mx3 and S, which possess the OH leaving group, also recognize the 5'-CCG sequence, in contrast to saframycin A, which contains the CN leaving group. In contrast, the OH-containing saframycins also recognize the 5'-CTA sequence. Saframycins Mx2, B and C, which lack the critical CN or OH leaving group, do not show any footprints on the restriction fragments examined in this study. The measured binding site size for all four antibiotics is three base pairs. The exonuclease III stop assay independently confirmed the formation of a covalent bond and the strong preference of the antibiotics for 5'-GGG and 5'-GCC sequences. The latter enzyme assay also suggests that the 5'-terminal or central G of the triad binding site is the base to which reversible covalent attachment of the antibiotic takes place.

Molecular recognition between ligands and nucleic acids: DNA binding characteristics of analogues of Hoechst 33258 designed to exhibit altered base and sequence recognition.

The DNA binding characteristics of new analogues (2-8) of Hoechst 33258 (1), containing pyridine and benzoxazole units and designed for altered base specificity, were evaluated using UV, fluorescence, and circular dichroism studies. Like Hoechst 33258 the new analogues also bind through the minor groove of B-DNA in a nonintercalative fashion. The interaction of the compounds with poly(dA-dT) is salt independent. The studies with poly(dA-dT), ct DNA, and poly(dG-dC) indicated a decrease in the relative binding strength of the new analogues to DNAs compared with the parent molecule, Hoechst 33258. Compounds 5 and 7 showed acceptance of GC bases adjacent to AT base pairs. None of the compounds studied exhibited affinity for A-DNA, double-stranded RNA, or Z-DNA. Structure-DNA binding relationships of the new analogues compared with their parent molecule, Hoechst 33258, are discussed.

Mode of action of saframycin antitumor antibiotics: sequence selectivities in the covalent binding of saframycins A and S to deoxyribonucleic acid.

The sequence-preferential reversible covalent binding of certain saframycin antitumor antibiotics from Streptomyces lavendulae has been examined by complementary-strand methidium propyl-EDTA (MPE) footprinting on an EcoRI/HindIII restriction fragment of pBR322 DNA under several experimental conditions. A buffer at pH 7.4 and in the presence of 9.5 mM dithiothreitol at 37 degrees C was found to be optimum for the interaction of these antibiotics with DNA. At r' = 0.6 both saframycins A and S exhibited footprints in the regions 4244-4257 (CAAATAGGGTTCC) and 4265-4286 (TTCCCCAAAAGTGCCACCTG) and a weak footprint in the region 4297-4302 (AACCAT). The binding locations identified that are common to saframycins A and S are (all 5'----3') GGGG (4250-4253), CCCC (4268-4271), and GCC (4279-4281), and weak interaction locations are ACC (4282-4284 and 4298-4300) (underlined bases are shared by two adjacent binding sites). Both the antibiotic saframycins A and S show preference for 5'-GGG or 5'-GGC sequences. It appears that saframycin A has no affinity for 5'-CGG while saframycin S shows a strong footprint at this sequence. Neither of the saframycins recognizes alternating CG sequences. Saframycin S also binds to 5'-CTA, which suggests that molecular recognition processes involving the parent antibiotics are also important, and not only recognition by, and covalent binding of, the common iminium species to the DNA. The protection sites at 5'-GCC and 5'-ACC suggest that saframycins A and S recognize 5'-GGPy sequences. However, between the two pyrimidine bases, C is preferred to T. Enhancement of cleavage by both saframycins is observed in the AT-rich region of 4301-4318.(ABSTRACT TRUNCATED AT 250 WORDS)

Interaction of synthetic analogues of distamycin and netropsin with nucleic acids. Does curvature of ligand play a role in distamycin-DNA interactions?

Distamycin and netropsin, a class of minor groove binding nonintercalating agents, are characterized by their B-DNA and A-T base-specific interactions. To understand the conformational and chemical basis of the above specificities, the DNA-binding characteristics of a novel synthetic analogue of distamycin have been studied. The analogue, mPD derivative, has the requisite charged end groups and a number of potential hydrogen-bonding loci equal to those of distamycin. The difference in the backbone curvatures of the ligands, distamycin, the mPD derivative, and NSC 101327 (another structurally analogous compound), is a major difference between these ligands. UV and CD spectroscopic studies reported here show the following salient features: The mPD derivative recognizes only B-DNA, to which it binds via the minor groove. On the other hand, unlike distamycin, it binds with comparable affinities to A-T and G-C base pairs in a natural DNA. These DNA-binding properties are compared with those reported earlier for distamycin and NSC 101327 [Zimmer, Ch., & Wahnert, U. (1986) Prog. Biophys. Mol. Biol. 47, 31-112]. The backbone structures of these three ligands were compared to show the progressive decrease in curvatures in the order distamycin, mPD derivative, and NSC 101327. The plausible significance of the backbone curvature vis-à-vis the characteristic B-DNA and AT-specific binding of distamycin is discussed. To our knowledge, this is the first attempt (with a model synthetic analogue) to probe the possible influence of backbone curvature upon the specificity of interactions of the distamycin class of groove-binding ligands with DNA.

Molecular recognition between ligands and nucleic acids: novel pyridine- and benzoxazole-containing agents related to Hoechst 33258 that exhibit altered DNA sequence specificity deduced from footprinting analysis and spectroscopic studies.

The syntheses of certain analogues of the DNA minor groove binding agent Hoechst 33258 designed to exhibit altered sequence recognition are described. The structural modifications include the following: substitution of pyridine for the benzene ring of the benzimidazole moiety, replacement of one benzimidazole unit by a benzoxazole in the two possible orientations with respect to the DNA receptor, and a synthesis of 2,2'-m-phenylene-bis[6-(4-methyl-1-piperazinyl)benzimidazole]. Sequence recognition of these agents on a HindIII/EcoRI fragment of pBR322 DNA was determined by MPE footprinting procedures. Some of the analogues exhibited altered DNA sequence preference compared with Hoechst 33258. In particular, a structure bearing a benzoxazole moiety with the oxygen oriented inward to the minor groove together with an inward-directed pyridine nitrogen appears to confer the property of recognition of a GC base pair within the binding sequence. The possible factors, structural, stereochemical, and electrostatic, contributing to the altered DNA sequence recognition properties are discussed.

Molecular recognition between oligopeptides and nucleic acids: DNA sequence specificity and binding properties of thiazole-lexitropsins incorporating the concepts of base site acceptance and avoidance.

The DNA binding and sequence specificity of a group of six novel thiazole containing lexitropsins related to the natural anti-tumor antibiotic distamycin have been examined by complementary strand MPE footprinting on two restriction fragments of pBR322 DNA. These lexitropsins comprise two groups in which the hetero atom of the thiazole moiety directed inwards to the floor of the minor groove is respectively nitrogen or sulfur. All of the new lexitropsins bind to DNAs in the minor groove with Kapp comparable with distamycin. The group of lexitropsins bearing nitrogen directed towards the DNA display comparable binding to poly(dA-dT) and to native DNAs, and complementary strand footprinting reveals their ability to accept and bind to mixed AT-GC sequences. The GC recognizing property plausibly arises from the hydrogen bonding between the thiazole nitrogen and G-2-NH2 based on precedents. In contrast the group of lexitropsins bearing sulfur directed towards the floor of the minor groove of DNA exhibit strict preference for AT sequences and are even more discriminating than distamycin. The latter agents, in common with the first group, bind firmly in the minor groove and with a binding site size of either 4 +/- 1 or 5 +/- 1 base pairs indicating intimate contact of all parts of the ligand. Therefore the property of GC site avoidance of these particular thiazole-lexitropsins is attributed to clash between the sterically more demanding sulfur and G-2NH2 groups.

Molecular recognition between oligopeptides and nucleic acids: DNA binding selectivity of a series of 1,2,4-triazole-containing lexitropsins.

The synthesis of a series of 1,2,4-triazole-containing oligopeptide lexitropsins related to the natural antitumor antibiotic distamycin is described. The binding properties of these new agents to both native DNAs and synthetic polydeoxyribonucleotides were determined by UV absorption and circular dichroism studies. The DNA binding and sequence selectivity of these agents have been determined by complementary strand MPE footprinting on two restriction fragments of pBR322 DNA. The lexitropsins bearing one 1,2,4-triazole moiety bind to AT sequences like distamycin but also tolerate GC sites. However, 3, which contains two contiguous 1,2,4-triazole moieties, appears not to recognize AT stretches and demands two consecutive GC bases in the middle of a four-base recognition site. The triazole lexitropsins avoid ATTT (4264-4267) and AAAA (4273-4276) to which distamycin binds strongly to the EcoRI/Hind III fragment. Determination of pKa values indicates the 1,2,4-triazole units are protonated under physiological conditions. The avoidance of AT stretches and the recognition of GC base sites by 3 suggest molecular recognition in these cases is determined, not by electrostatic effects, but rather by invoking a hydrogen bond between the triazole 4-nitrogen and G-2NH2 in the minor groove.

DNA-sequence specific recognition by a thiazole analogue of netropsin: a comparative footprinting study.

Four different footprinting techniques have been used to probe the DNA sequence selectivity of Thia-Net, a bis-cationic analogue of the minor groove binder netropsin in which the N-methylpyrrole moieties are replaced by thiazole groups. In Thia-Net the ring nitrogen atoms are directed into the minor groove where they could accept hydrogen bonds from the exocyclic 2-amino group of guanine. Three nucleases (DNAase I, DNAase II, and micrococcal nuclease) were employed to detect binding sites on the 160bp tyr T fragment obtained from plasmid pKM delta-98, and further experiments were performed with 117mer and 253mer fragments cut out of the plasmid pBS. MPE.Fe(II) was used to footprint binding sites on an EcoRI/HindIII fragment from pBR322. Thia-Net binds to sites in the minor groove containing 4 or 5 base pairs which are predominantly composed of alternating A and T residues, but with significant acceptance of intrusive GC base pairs. Unlike the parent antibiotic netropsin, Thia-Net discriminates against homooligomeric runs of A and T. The evident preference of Thia-Net for AT-rich sites, despite its containing thiazole nitrogens capable of accepting GC sites by hydrogen bonding, supports the view that the biscationic nature of the ligand imposes a bias due to the electrostatic potential differences in the receptor which favour the ligand reading alternating AT sequences.

Polarity of annealing and structural analysis of the RNase H resistant alpha-5'-d[TACACA].beta-5'-r[AUGUGU] hybrid determined by high-field 1H, 13C, and 31P NMR analysis.

The novel hybrid duplex alpha-5'-d[TACACA]-3'.beta-5'-r[AUGUGU]-3' was analyzed extensively by 1D and 2D NMR methods. Two forms of the duplex exist in about an 80:20 ratio. Analysis of the exchangeable imino protons of the major component revealed that three AU and one AT base pair are present in addition to two GC base pairs, confirming that the duplex anneals in parallel orientation. The presence of the AT base pair, which can only be accounted for by a parallel duplex, was confirmed by a selective INEPT experiment, which correlated the thymidine imino proton to its C5 carbon. The lesser antiparallel form could be detected by exchangeable and nonexchangeable proton resonances in both strands. An exchange peak was observed in the NOESY spectrum for the thymidine methyl group resonance in both the predominant and lesser conformations, indicating the lifetime of the individual structures was on the millisecond time scale. The nonexchangeable protons of the predominant duplex were assigned by standard methods. The sugar pucker of the ribonucleosides was determined to be of the "S" type by a pseudorotation analysis according to Altona, with the J-couplings measured from the multiplet components of the phase-sensitive COSY experiment. The NOE pattern observed for the alpha-deoxynucleosides also suggested an S-type sugar pucker. The adoption of an S-type sugar pucker for both strands indicates that, in contrast to RNA.DNA duplexes formed exclusively from beta-nucleotides, the alpha-DNA.beta-RNA duplex may form a B-type helix. The 31P resonances of the alpha and beta strands have very different chemical shifts in the hybrid duplex and the difference persists above the helix melting temperature, indicating an intrinsic difference in 31P chemical shift for nucleotides differing only in the configuration about the glycosidic bond.

Psoralen--lexitropsin hybrids: DNA sequence selectivity of photoinduced cross-linking from MPE footprinting and exonuclease III stop assay, and mode of binding from electric linear dichroism.

The properties of certain hybrids 3 and 5 bearing a photoactivatable psoralen group attached to DNA sequence recognizing lexitropsin carriers have been examined. The hybrids bind to poly(dA-dT) with Kapp of 2.8 and 0.9 x 10(7) M-1, i.e. greater than or equal to that of netropsin (Kapp = 1.0 x 10(7) M-1), indicating that the psoralen moiety may contribute to binding in the case of 5. Photoinduced cross-linking of DNA by 3 and 5, while efficient, is less so than that of individual psoralens and reaches a maximum at a ligand to DNA base pair ratio (r) of 0.2. Complementary strand methidium-propyl-EDTA (MPE).Fe(II) footprinting demonstrated that, in the dark, the sequence preferential recognition of hybrids 3 and 5 is dominated by the lexitropsin moiety. Examination of 360 nm photoinduced DNA cross-linking by the hybrids 3 and 5 was carried out using an exonuclease III stop assay. This revealed that > 95% of the DNA remained double stranded, indicating that 3 and 5 generate primarily biadducts at AT-rich sequences. This assay also located individual monoadduct sites, some of which are remote from the dominant cross-linked sites. When the samples were exposed to 254 nm UV light before loading onto the gel to reverse the photoproducts, the pattern of the exonuclease III stop bands was not altered significantly compared with the experiment without 254 nm irradiation. It is concluded that these termination sites include both mono- and biadducts. Electric linear dichroism examination of the DNA complexes of hybrids 3 and 5 (without light activation) provides evidence that the lexitropsin portion binds in the minor groove, while the psoralen portion intercalates in a suitably located site for subsequent photoinduced cross-linking.

Molecular recognition between oligopeptides and nucleic acids. DNA sequence specificity and binding properties of an acridine-linked netropsin hybrid ligand.

The binding to DNA of a mixed function ligand (NETGA) is described, in which a potential intercalating group, an acridine moiety, is incorporated at the carboxyl terminus of the minor groove binding oligopeptide netropsin skeleton. Scatchard analysis of absorption data provided evidence of two modes of binding to DNA with K1 = 9.1 x 10(5) M-1 at low r values (0.003-0.1), and a binding site size n = 10, indicative of binding of both moeities. At high binding ratios (greater than 0.1), K2 = 0.9 x 10(5) M-1 and n = 5 corresponding to external binding. Complementary strand MPE footprinting on a pBR322 restriction fragment showed NETGA binds to 5'-AAAT like netropsin. It causes enhanced cleavage by MPE, particularly at G-C rich sequences and remote from the preferred binding sites. Viscometry measurements provided evidence for biphasic modes of the two binding portions of NETGA. Fluorescence polarization and linear dichroism measurements were in accord with distinct modes of interaction of the acridine (intercalation) and oligopeptide (minor groove binding) portions of NETGA. LD measurements on NETGA indicate that the oligopeptide moiety (netropsin-like) has an orientation typical of minor groove binders, whereas the degree of intercalation of the acridine group is decreased by association of the oligopeptide moiety.