Interaction of echinomycin with An.Tn. and (AT)n regions flanking its CG binding site

Thioester Bonds of Thiocoraline Can Be Replaced with NMe-Amide Bridges without Affecting Its DNA-Binding Properties

In the search for new drug candidates for DNA recognition, affinity and sequence selectivity are two of the most important features. NMe-azathiocoraline, a synthetic antitumor bisintercalator derived from the natural marine product thiocoraline, shows similar potency to the parent compound, as well as possessing enhanced stability. Analysis of the DNA-binding selectivity of NMe-azathiocoraline by DNase I footprinting using universal substrates with all 136 tetranucleotides and all possible symmetrical hexanucleotide sequences revealed that, although this ligand binds to all CpG steps with lower affinities than thiocoraline, it displays additional binding to AT-rich sites. Moreover, fluorescence melting studies showed a strong interaction of the synthetic molecule with CACGTG and weaker binding to ACATGT and AGATCT. These findings demonstrate that NMe-azathiocoraline has the same mode of action as thiocoraline, mimicking its DNA-binding selectivity despite the substitution of its thioester bonds by NMe-amide bridges.

Interaction of mithramycin with isolated GC and CG sites

We have studied the interaction of the GC-specific, minor groove-binding ligand, mithramycin, with cloned DNA inserts containing isolated GC and CG sites flanked by regions of (AT)n and An.Tn using DNase I and hydroxyl radical footprinting. We find that mithramycin binds to GC better than CG and that AGCT is a better site than TGCA. Sites flanked by (AT)n appear to be bound better than those surrounded by An.Tn. Although no footprints are produced at T9GCA9 and T15CGA15, DNase I cleavage is enhanced within the GC sites suggesting that there is some interaction with the ligand. Mithramycin also alters the DNase I cleavage of (GA)n.(CT)n.

DNase I footprinting of triple helix formation at polypurine tracts by acridine-linked oligopyrimidines: stringency, structural changes and interaction with minor groove binding ligands

We have investigated the binding of short (10 base) acridine-linked triplex-forming oligonucleotides to the target sequence A6G6.C6T6 by DNase I footprinting. Specific binding is detected at low pH (< 6.0) for 5'-Acr-T5C5 and 5'-Acr-5BrU5(5Me)C5. The sequence T5C5, lacking the acridine modification, binds less strongly, though specific binding is still evident. 5'-Acr-T5C5 produces footprints at slightly lower concentrations than 5'-Acr-5BrU5(5Me)C5. All three oligonucleotides produce enhanced DNase I digestion at the 3'-end of the target purine strand, suggesting that there is a DNA structural change at the triplex-duplex boundary. Target sequences AnG4A and TAC3Tn, containing one and two triplex mismatches, show no interaction with the acridine-free oligonucleotide, but bind the acridine-linked oligonucleotides. In these secondary binding modes the third strand is positioned so that the mismatches are located at the 3'-end of the oligonucleotide. Mithramycin and distamycin, binding in the minor groove to GC- and AT-rich sequences respectively, abolish triple helix formation.

Localized chemical reactivity in DNA associated with the sequence-specific bisintercalation of echinomycin

Four complementary footprinting and probing techniques utilizing DNAse I, methidiumpropyl EDTA (MPE).FeII, diethyl pyrocarbonate (DEPC) and KMnO4 as DNA-cleaving or DNA-modifying agents have been applied to investigate the sequence-specific binding to DNA of the antitumour antibiotic echinomycin. A 265 bp EcoRI-PvuII DNA restriction fragment excised from plasmid pBS was used as a substrate. Six regions of protection against DNAase I cleavage were located on the 265-mer: three sites encompass the sequences 5'-TCGA or 5'-GCGT and the three others contain 5'-GpG (CpC) dinucleotide sequences where the inhibition of DNAase I cutting by echinomycin is less pronounced. In contrast, MPE.FeII cleavage allows identification of only three echinomycin-binding sites on the 265-mer: two sites contain the sequence 5'-TCGA and one encompasses the sequence 5'-ACCA. Cleavage of DNA by MPE.FeII in the presence of echinomycin remains practically unaffected at the sequence 5'-GCGT, despite its identification by DNAase I as a strong site for binding the antibiotic, as well as at the two other sequences containing GpG steps. With both DNAase I and MPE.FeII, enhanced DNA cleavage is evident at AT-rich sequences in the presence of echinomycin. Enhanced reactivity towards KMnO4 and DEPC provides clear evidence for sequence-dependent conformational changes in DNA induced by the antibiotic. The experiments reveal that KMnO4 reacts most strongly with thymines located around, but not necessarily adjacent to, an echinomycin-binding site, whereas the carbethoxylation reactions caused by DEPC occur primarily at the adenine residues lying immediately 5' or 3' to the dinucleotide that denotes an echinomycin-binding site. The results reported here demonstrate that DEPC and KMnO4 serve as sensitive probes for different states of the DNA helix. It seems that the reaction with KMnO4 involves transient unstacking events, whereas the carbethoxylation reaction of DEPC requires larger-scale helix opening.

Light-activated cleavage of DNA by cobalt-bleomycin

We have studied the light-activated cleavage of DNA by cobalt-bleomycin using a series of synthetic DNA fragments containing (AT)n and (GC)n. This cleavage reaction requires high concentrations of the antibiotic and appears to be a stoichiometric process rather than a catalytic process. We find that, in common with the iron-complex, cobalt-bleomycin can cleave at ApT steps within regions of alternating AT residues; ApT steps within other sequences including (AAT)n. (ATP)n are not good substrates for cobalt-bleomycin cleavage. Some repetitive regions display an alternating pattern of cleavage products, revealing the preferred arrangement of ligand molecules along a saturated DNA lattice. A similar repetitive pattern is found for diethylpyrocarbonate modification and hydroxyl-radical cleavage. Although cleavage of ApT and GpC proceeds at equivalent rates, the data suggest that bleomycin binds more tightly to the latter. Adenine residues on the 3' side of both GpC-cleavage and ApT-cleavage sites are rendered more reactive to diethylpyrocarbonate, consistent with a ligand-induced alteration in local DNA structure. The cobalt-bleomycin-binding site consists of not more than four base pairs, and may be as small as three base pairs.

DNA structure influences sequence specific cleavage by bleomycin

We have examined the cleavage of several synthetic DNA sequences by iron(II)-bleomycin. We find that, although bleomycin cuts mixed sequence DNAs with a preference for GC = GT > GA >> GG, it efficiently cleaves regions of (AT)n cutting exclusively at ApT, not TpA. Isolated ApT steps show very little cleavage while blocks of three or more contiguous ATs are cut as efficiently as GpT. This cleavage is specific for (AT)n, since sequences of the type (TAA)n.(TTA)n and (ATT)n.(AAT)n are hardly cut at all. No cleavage is observed at ApC or CpA within sequences of the type (AC)n.(GT)n; regions of An.Tn are also not cut. Although the cobalt-bleomycin complex (which binds to but does not cleave DNA) yields good DNase I footprints at GT and GC sites, no footprints are observed within (AT)n, suggesting that although the cleavage reaction is efficient, the binding affinity is relatively weak. We propose a model in which bleomycin cleavage is determined by local DNA structure, while strong binding requires the presence of a guanine residue.

Sequence-specific binding of [N-MeCys3,N-MeCys7]TANDEM to TpA

The sequence selective binding of [N-MeCys3,N-MeCys7]TANDEM to DNA has been studied by footprinting experiments on DNA fragments containing the self-complementary sequences CGCGATATCGCG, CGCGTATACGCG, CGCGTTAACGCG and CGCGAATTCGCG. DNAase I and micrococcal nuclease reveal drug-induced footprints with the central sequences ATAT, TATA and TTAA, but not AATT, suggesting that the ligand binds to the dinucleotide TpA. The ligand renders certain adenines hyper-reactive to diethyl pyrocarbonate. These are observed with ATAT, TATA and TTAA, but not AATT, and are located both within, and distal to, the TpA-binding sites.

The 2-amino group of guanine is absolutely required for specific binding of the anti-cancer antibiotic echinomycin to DNA

The 2-amino group of guanine is believed to be a critical determinant of potential DNA binding sites for echinomycin and related quinoxaline antibiotics. In order to probe its importance directly we have studied the interaction between echinomycin and DNA species in which guanine N(2) is deleted by virtue of substitution of inosine for guanosine residues. The polymerase chain reaction was used to prepare inosine-substituted DNA. Binding of echinomycin, assessed by DNAse I footprinting, was practically abolished by incorporation of inosine into one or both strands of DNA. We conclude that both the purines in the preferred CpG binding site need to bear a 2-amino group to interact with echinomycin.

Footprinting studies of DNA-sequence recognition by nogalamycin

We have studied the DNA sequence binding preference of the antitumour antibiotic nogalamycin by DNase-I footprinting using a variety of DNA fragments. The DNA fragments were obtained by cloning synthetic oligonucleotides into longer DNA fragments and were designed to contain isolated ligand-binding sites surrounded by repetitive sequences such as (A)n.(T)n and (AT)n. Within regions of (A)n.(T)n, clear footprints are observed with low concentrations of nogalamycin (< 5 microM), with apparent binding affinities for tetranucleotide sequences which decrease in the order TGCA > AGCT = ACGT > TCGA. In contrast, within regions of (AT)n, the ligand binds best to AGCT; binding to TCGA and TGCA is no stronger than to alternating AT. Within (ATT)n, the preference is for ACGT > TCGA. Although each of these binding sites contains all four base pairs, there is no apparent consensus sequence, suggesting that the selectivity is affected by local DNA dynamic and structural effects. At higher drug concentrations (> 25 microM), nogalamycin prevents DNAse-I cleavage of (AT)n but shows no interaction with regions of (AC)n.(GT)n. Regions of (A)n.(T)n, which are poorly cut by DNase I, show enhanced rates of cleavage in the presence of low concentrations of nogalamycin, but are protected from cleavage at higher concentrations. We suggest that this arises because drug binding to adjacent regions distorts the DNA to a structure which is more readily cut by the enzyme and which is better able to bind further ligand molecules.

Secondary (non-GpC) binding sites for actinomycin on DNA

Actinomycin D has long been known to bind selectively to the dinucleotide step GpC. We have investigated its ability to bind to other non-canonical sequences using a series of synthetic DNA fragments. DNase I footprinting experiments reveal that actinomycin can also bind well to GG (CC). Binding to this sequence and the canonical GC site is potentiated by flanking regions of (GT)n.(AC)n. Weaker but specific binding to GT and AC is also evident and appears to be cooperative.