Comparison of binding sites in DNA for berenil, netropsin and distamycin. A footprinting study

Techniques of DNase I and micrococcal nuclease footprinting have been used to compare the binding sites for berenil, netropsin and distamycin on two different DNA fragments. Each ligand binds to the A + T-rich zones which contain clusters of at least four A.T base pairs. Neither guanosine nor cytidine nucleotides appear to be allowed within the A + T-rich runs which constitute the preferred binding sites, although they are sometimes protected from DNase I cleavage in neighbouring regions. Berenil and netropsin share with distamycin the property of causing enhanced rates of cleavage at certain sequences flanking their binding sites. There are significant differences in the concentrations of each ligand required to produce defined patterns of protection, seemingly dependent upon the nature (and possibly the gross base composition) of the piece of DNA being used in the experiment.     

Echinomycin binding to alternating AT.

We have studied the binding of echinomycin to DNA fragments containing GC-rich regions flanked by blocks of alternating AT by DNase I footprinting and diethylpyrocarbonate modification. Regions of alternating AT flanking the sequences CCCG, CCGC, CGGC and GG show a four base pair DNase I cleavage pattern and reaction of alternate adenines with diethylpyrocarbonate. This pattern is strongest when the AT-block is immediately adjacent to the CpG ligand binding site. We explain these phenomena by suggesting that echinomycin binds to the dinucleotide step ApT in a cooperative fashion. The cooperative effects can be transmitted through the dinucleotide step GC but not CC or AA. No such repetitive patterns are seen with surrounding regions of (ATT).(AAT). Evidence is presented for secondary drug binding sites at CpC and TpG with weaker interaction at the CpG site within the hexanucleotide TTCGAA.     

Effects of sequence selective drugs on the gel mobility of a bent DNA fragment.

The effects of various drugs on the structure of a bent DNA fragment have been investigated by studying DNA mobility in polyacrylamide gels. This DNA fragment has an anomalously slow rate of migration on account of its phased runs of adenines. Nogalamycin and echinomycin increase the gel mobility of kinetoplast DNA suggesting that the bending has been removed. Mithramycin, actinomycin, distamycin and ethidium have either no effect or cause a further reduction in mobility. These results are compared with other, non-bent DNA species which always show a decrease in gel mobility in the presence of DNA binding drugs.     

Wrapping of genomic polydA.polydT tracts around nucleosome core particles.

Five human clones containing genomic regions of polydA have been isolated by their ability to form intermolecular triple helices with agarose cross-linked polyU. All of these clones contain Alu repetitive DNA sequences. End-labelled DNA fragments containing these sequences have been successfully reconstituted onto nucleosome core particles by salt exchange. The structure of these has been examined by digesting with DNase I, hydroxyl radicals or diethylpyrocarbonate. DNase I cleavage of the polydA tracts is poor in the free DNA but is markedly enhanced at certain positions when complexed with nucleosome cores. Phased digestion patterns are observed which continue through the (A)n blocks and reveal an average helical periodicity of about 10 base pairs. The distance between adjacent maxima varies between 8-12 base pairs, suggesting that the exact helical repeat is not necessarily constant. One fragment containing the sequence (TA)11T34 reveals a 12 base pair repeat within the (AT)n region. A pUC19 polylinker fragment containing a block of A69.T69 cloned into the Smal site could also be reconstituted onto nucleosome cores and reveals the same phased DNaseI digestion pattern. The DNase I cleavage pattern is not identical at each of the maxima, suggesting that the structural distortions imposed by the core particles are not constant along the DNA.     

Specific protection of DNA by distamycin A, netropsin and bis-netropsins against the action of DNAse I

Interaction of netropsin, distamycin A and a number of bis-netropsins with DNA fragments of definite nucleotide sequence was studied by footprinting technique. The nuclease protection experiments were made at fixed DNA concentration and varying ligand concentrations. The affinity of ligand for a DNA site was estimated from measurements of ligand concentration that causes 50% protection of the DNA site. Distribution pattern of the protected and unprotected regions along the DNA fragment was compared with the theoretically expected arrangement of the ligand along the same DNA. The comparison led us to the following conclusions: 1. Footprinting experiments show that at high levels of binding the arrangement of netropsin molecules along the DNA corresponds closely to the distribution pattern expected from theoretical calculations based on the known geometry of netropsin--DNA complex. However, the observed differences in the affinity of netropsin for various DNA sequences is markedly greater than that expected from theoretical calculations. 2. Netropsin exhibits a greater selectivity of binding than that expected for a ligand with three specific reaction centers associated with the antibiotic amide groups. It binds preferentially to DNA regions containing four or more successive AT pairs. Among 13 putative binding sites for netropsin with four or more successive AT pairs there are 11 strong binding sites and two weaker sites which are occupied at 2 D/P less than or equal to 1/9 and 2 D/P = 1/4, respectively. 3. The extent of specificity manifested by distamycin A is comparable to that shown by netropsin although the molecule of distamycin A contains four rather than three amide groups. At high levels of binding distamycin A occupies the same binding sites on DNA as netropsin does. 4. The binding specificity of bis-netropsins is greater than that of netropsin. Bis-netropsins can bind to DNA in such a way that the two netropsin-like fragments are implicated in specific interaction with DNA base pairs. However, the apparent affinity of bis-netropsins estimated from footprinting experiments is comparable with that of netropsin for the same DNA region. 5. At high levels of binding bis-netropsins and distamycin A (but not netropsin) can occupy any potential site on DNA irrespectively of the DNA sequence. 6. Complex formation with netropsin increases sensitivity to DNase I at certain DNA sites along with the protection effect observed at neighboring sites.     

Formation, stability and core histone positioning of nucleosomes reassembled on bent and other nucleosome-derived DNA

DNA originating from chicken erythrocyte mononucleosomes was cloned and sequenced. The properties of nucleosome reconstruction were compared for two cloned inserts, selected on account of their interesting sequence organization, length and difference in DNA bending. Cloned fragment 223 (182 base-pairs) carries alternatively (A)3-4 and (T)4-5 runs approximately every ten base-pairs and is bent; cloned fragment 213 (182 base-pairs) contains a repeated C4-5ATAAGG consensus sequence and is apparently not bent. Our experiments indicate the preference of the bent DNA fragment 223 over fragment 213 to associate in vitro with an octamer of histones under stringent conditions. We provide evidence that the in vitro nucleosome formation is hampered in the case of fragment 213, whereas the reconstituted nucleosomes were equally stable once formed. For the correct determination of the positioning of the histone octamer with regard to the two nucleosome-derived cloned DNA sequences, the complementary use of micrococcal nuclease, exonuclease III and DNase I is a prerequisite. No unique, but rotationally related, positions of the histone octamer were found on these nucleosome-derived DNA fragments. The sequence-dependent anisotropic flexibility, as well as intrinsic bending of the DNA, resulting in a rotational setting of the DNA fragments on the histone core, seems to be a strong determinant for the allowed octamer positions, Exonuclease III digestion indicates a different histone-DNA association when oligo(d(C.G)n) stretches are involved. The apparent stagger near oligo(d(A.T)n) stretches generated by DNase I digestion on reconstituted nucleosome 223 was found to be inverted from the normal two-base 3' overhang to a two-base 5' overhang. Two possibilities of the oligo(d(A.T)n) minor groove location relative to the histone core are envisaged to explain this anomaly in stagger.     

DNA sequence dependent binding modes of 4',6-diamidino-2-phenylindole (DAPI)

The interactions of DAPI with natural DNA and synthetic polymers have been investigated by hydrodynamic, DNase I footprinting, spectroscopic, binding, and kinetic methods. Footprinting results at low ratios (compound to base pair) are similar for DAPI and distamycin. At high ratios, however, GC regions are blocked from enzyme cleavage by DAPI but not by distamycin. Both poly[d(G-C)]2 and poly[d(A-T)]2 induce hypochromism and shifts of the DAPI absorption band to longer wavelengths, but the effects are larger with the GC polymer. NMR shifts of DAPI protons in the presence of excess AT and GC polymers are significantly different, upfield for GC and mixed small shifts for AT. The dissociation rate constants and effects of salt concentration on the rate constants are also quite different for the AT and the GC polymer complexes. The DAPI dissociation rate constant is larger with the GC polymer but is less sensitive to changes in salt concentration than with the AT complex. Binding of DAPI to the GC polymer and to poly[d(A-C)].poly[d(G-T)] exhibits slight negative cooperativity, characteristic of a neighbor-exclusion binding mode. DAPI binding to the AT polymer is unusually strong and exhibits significant positive cooperativity. DAPI has very different effects on the bleomycin-catalyzed cleavage of the AT and GC polymers, a strong inhibition with the AT polymer but enhanced cleavage with the GC polymer. All of these results are consistent with two totally different DNA binding modes for DAPI in regions containing consecutive AT base pairs versus regions containing GC or mixed GC and AT base pair sequences. The binding mode at AT sites has characteristics which are similar to those of the distamycin-AT complex, and all results are consistent with a cooperative, very strong minor groove binding mode. In GC and mixed-sequence regions the results are very similar to those observed with classical intercalators such as ethidium and indicate that DAPI intercalates in DNA sequences which do not contain at least three consecutive AT base pairs.     

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

We have prepared DNA fragments containing the sequences A15CGT15, T15CGA15 and T(AT)8CG(AT)15 cloned within the SmaI site of the pUC19 polylinker. These have been used as substrates in footprinting experiments with DNase I and diethylpyrocarbonate probing the effects of echinomycin, binding to the central CG, on the structure of the surrounding sequences. No clear DNase I footprints are seen with T15CGA15 though alterations in the nuclease susceptibility of surrounding regions suggest that the ligand is binding, albeit weakly at this site. All the other fragments show the expected footprints around the CG site. Regions of An and Tn are rendered much more reactive to DNase I and adenines on the 3'-side of the CG become hyperreactive to diethylpyrocarbonate. Regions of alternating AT show unusual changes in the presence of the ligand. At low concentrations (5 microM) cleavage of TpA is enhanced, whereas at higher concentrations a cleavage pattern with a four base pair repeat is evident. A similar pattern is seen with micrococcal nuclease. Modification by diethylpyrocarbonate is strongest at alternate adenines which are staggered in the 5'-direction across the two strands. We interpret these changes by suggesting secondary drug binding within regions of alternating AT, possibly to the dinucleotide ApT. DNase I footprinting experiments performed at 4 degrees C revealed neither enhancements nor footprints for flanking regions of homopolymeric A and T suggesting that the conformational changes are necessary consequence of drug binding.     

Demethylation of thymine residues affects DNA cleavage by endonucleases but not sequence recognition by drugs.

The 5-methyl group of thymidine residues protrudes into the major groove of double helical DNA. The structural influence of this exocyclic substituent has been examined using a PCR-made 160 bp fragment in which thymidine residues were replaced with uridine residues. We show that the dT-->dU substitution and the consequent deletion of the methyl group affects the cleavage of DNA by deoxyribonuclease I and micrococcal nuclease. Analysis of the DNase I cleavage sites, in terms of di and trinucleotides, indicates that homopolymeric tracts of d(AT) become significantly more susceptible to DNase I cleavage when uridine is substituted for thymidine residues. The results indicate that removal of the thymidine methyl groups from the major groove at AT tracts induces structural perturbations that transmit into the opposite minor groove, where they can be detected by endonuclease probing. In contrast, DNase I footprinting experiments with different mono and bis-intercalating drugs reveal that dT-->dU substitution does not markedly affect sequence-specific drug-DNA recognition in the minor or major groove of the double helix. The consequences of demethylation of thymidine residues are discussed in terms of changes in the minor groove width connected to variations in the flexibility of DNA and the intrinsic curvature associated with AT tracts. The study identifies the methyl group of thymine as an important molecular determinant controlling the width of the minor groove and/or the flexibility of the DNA.     

Interaction of minor groove binding ligands with long AT tracts.

We have used quantitative DNase I footprinting to examine the ability of distamycin and Hoechst 33258 to discriminate between different arrangements of AT residues, using synthetic DNA fragments containing multiple blocks of (A/T)6or (A/T)10in identical sequence environments. Previous studies have shown that these ligands bind less well to (A/T)4sites containing TpA steps. We find that in (A/T)6tracts distamycin shows little discrimination between the various sites, binding approximately 2-fold stronger to TAATTA than (TA)3, T3A3and GAATTC. In contrast, Hoechst 33258 binds approximately 20-fold more tightly to GAATTC and TAATTA than T3A3and (TA)3. Hydroxyl radical footprinting reveals that both ligands bind in similar locations at the centre of each AT tract. At (A/T)10sites distamycin binds with similar affinity to T5A5, (TA)5and AATT, though bands in the centre of (TA)5are protected at approximately 50-fold lower concentration than those towards the edges. Hoechst 33258 shows a similar pattern of preference, with strong binding to AATT, T5A5and the centre of (TA)5. Hydroxyl radical footprinting reveals that at low concentrations both ligands bind at the centre of (TA)5and A5T5, while at higher concentrations ligand molecules bind to each end of the (A/T)10tracts. At T5A5two ligand molecules bind at either end of the site, even at the lowest ligand concentration, consistent with the suggestion that these compounds avoid the TpA step. Similar DNase I footprinting experiments with a DNA fragment containing T n (n = 3-6) tracts reveals that both ligands bind in the order T3< T4 << T5 = T6.     

Methidiumpropyl-EDTA.Fe(II) and DNase I footprinting report different small molecule binding site sizes on DNA.

DNase I and MPE.Fe (II) footprinting both employ partial cleavage of ligand-protected DNA restriction fragments and Maxam-Gilbert sequencing gel methods of analysis. One method utilizes the enzyme, DNase I, as the DNA cleaving agent while the other employs the synthetic molecule, methidium-propyl-EDTA (MPE). For actinomycin D, chromomycin A3 and distamycin A, DNase I footprinting reports larger binding site sizes than MPE.Fe (II). DNase I footprinting appears more sensitive for weakly bound sites. MPE.Fe (II) footprinting appears more accurate in determining the actual size and location of the binding sites for small molecules on DNA, especially in cases where several small molecules are closely spaced on the DNA. MPE.Fe (II) and DNase I report the same sequence and binding site size for lac repressor protein on operator DNA.     

Minor groove to major groove, an unusual DNA sequence-dependent change in bend directionality by a distamycin dimer

DNA sequence-dependent conformational changes induced by the minor groove binder, distamycin, have been evaluated by polyacrylamide gel electrophoresis. The distamycin binding affinity, cooperativity, and stoichiometry with three target DNA sequences that have different sizes of alternating AT sites, ATAT, ATATA, and ATATAT, have been determined by mass spectrometry and surface plasmon resonance to help explain the conformational changes. The results show that distamycin binds strongly to and bends five or six AT base pair minor groove sites as a dimer with positive cooperativity, while it binds to ATAT as a weak, slightly anticooperative dimer. The bending direction was evaluated with an in phase A-tract reference sequence. Unlike other similar monomer minor groove binding compounds, such as netropsin, the distamycin dimer changes the directionality of the overall curvature away from the minor groove to the major groove. This distinct structural effect may allow designed distamycin derivatives to have selective therapeutic effects.     

Rate enhancements in the DNase I footprinting experiment.

Footprinting experiments for DNase I digests of a 139-base-pair segment of pBR-322 DNA in the presence of either netropsin or actinomycin D were carried out. Plots of oligonucleotide concentration as a function of drug concentration were analyzed to study the enhancement in cleavage rates at approximately 30 sites, accompanying drug binding at other sites. The pattern of enhancements is not consistent with drug-induced DNA structural changes, but agrees with a redistribution mechanism involving DNase I. Since the total number of enzyme molecules per fragment remains unchanged, drug binding at some sites increases the enzyme concentration at other sites, giving rise to increased cleavage. The consequences of the redistribution mechanism for analysis of footprinting experiments are indicated.     

Determination of equilibrium binding affinity of distamycin and netropsin to the synthetic deoxyoligonucleotide sequence d(GGTATACC)2 by quantitative DNase I footprinting

A new method for determining the equilibrium binding constant of antitumor drugs to specific DNA sequences by quantitative DNase I footprinting is presented. The use of a short synthetic DNA oligomer to define a homogeneous population of DNA binding sites enables the calculation of the free drug concentration and the fraction of DNA sites complexed with drug in solution and is described for the first time. Since a 1:1 stoichiometry is observed for each drug-oligomer DNA complex, it becomes possible to calculate equilibrium binding constants in solution. By use of this technique, the binding affinities of the nonintercalating drugs netropsin and distamycin to the synthetic oligonucleotide d(GGTATACC)2 are determined to be Ka (25 degrees C) = 1.0 X 10(5) and 2.0 X 10(5) M-1, respectively. Quantitation of the temperature dependence associated with complex formation results in a determination of standard enthalpies of -3.75 and -8.48 kcal mol-1 for the binding of netropsin and distamycin, respectively. Calculation of other thermodynamic parameters are found to be in agreement with previous studies and indicate that the DNA binding process for these compounds is predominantly enthalpy driven. This method of quantitative DNase I footprinting is demonstrated to be a useful technique for the measurement of drug affinities to specific binding sites on DNA oligomers which are designed and synthesized expressly for this purpose. Applications of the technique to the determination of drug binding affinities at specific sites within native DNA sequences are discussed.     

Effects of a hairpin polyamide on DNA melting: comparison with distamycin and Hoechst 33258

We have used DNase I footprinting and fluorescence melting studies to study the interaction of the hairpin polyamide Im-Py-Py-Py-(R)H2Ngamma-Im-Py-Py-Py-beta-Dp with its preferred binding sites (5'-WGWWCW; W=A or T) and other sequences. DNase I footprinting confirmed that the ligand binds to the sequence AGAACA at nanomolar concentrations and that changing the terminal A to G causes a dramatic decrease in affinity, while there was no interaction with the reverse sequence WCWWGW. Fluorescence melting studies with 11-mer duplexes showed that the polyamide had very different effects on the forward (TGWWCT) and reverse (TCTAGT) sequences. At low concentrations, the polyamide produced biphasic melting curves with TGATCT, TGTACT and TGAACT, suggesting a strong interaction. In contrast, the melting profiles with TCTAGT were always monophasic and showed much smaller concentration dependent changes in Tm. The polyamide also showed weak binding to the sequence TGATCT when one of the central AT pairs was replaced with an AC mismatch. These melting profiles were compared with those produced by the AT-selective minor groove binding agents distamycin and Hoechst 33258 at the same sites and at similar sequences containing A5 and (AT)3, which are expected to bind distamycin in the 1:1 and 2:1 modes, respectively. These ligands produced simple monophasic melting curves in which the Tm steadily increased as the ligand concentration was raised.     

Heterogeneity in the actions of drugs that bind in the DNA minor groove.

Distamycin and Hoechst 33258 have long served as the model compounds for biochemical, biophysical, and clinical studies of the drugs that bind in the DNA minor groove. However, the results presented in this investigation clearly show that 4,6-diamidino-2 phenylindole (DAPI) is superior to both of these drugs at negating the effects of intrinsic DNA curvature and anisotropic bendability as measured by electrophoretic and ligation analysis. In addition, DAPI was more effective than distamycin and Hoechst 33258 at inhibiting the assembly of nucleosomes onto synthetic and natural sequences that have multiple closely spaced oligo-AT sequences that serve as drug binding sites. Since these effects may be related to the biological action of the drugs, it was of interest to determine the mechanism that was responsible for the enhanced action of DAPI. The possibility that the differential drug potencies resulted from differential overall affinities of the ligands for A-tract molecules was considered, but drug binding studies suggested that this was not the case. It is also unlikely that the differential drug effects resulted from the binding of the drugs to different DNA sites since the oligo A/T binding sites for DAPI and Hoechst were centered on the same nucleotide positions as revealed by footprinting studies using exonuclease III, DNase I, and hydroxyl radical. However, the footprinting studies with DNase I did uncover a potentially important difference between the drugs. DAPI protected only the AT bp in the binding sites, while distamycin and Hoechst protected these bp as well as flanking Gs and Cs. These results permitted us to advance a preliminary model for the enhanced action DAPI. According to the model, the short length of DAPI and its absolute specificity for A/T bps with narrow minor grooves ensures that only particularly minor grooves that give rise to curvature and anisotropic bendability are occupied by the drug. Consequently, each helical deflection induced by an A-tract in the absence of the drug is countered by an opposite deflection induced by DAPI binding, thus effectively neutralizing intrinsic curvature and bending into the minor groove.     

DNA recognition by DNase I

Bovine pancreatic DNase I shows a strong preference for double-stranded substrates and cleaves DNA with strongly varying cutting rates suggesting that the enzyme recognises sequence-dependent structural variations of the DNA double helix. The complicated cleavage pattern indicates that several local as well as global helix parameters influences the cutting frequency of DNase I at a given bond. The high resolution crystal structures of two DNase I-DNA complexes showed that the enzyme binds tightly in the minor groove, and to the sugar-phosphate backbones of both strands, and thereby induces a widening of the minor groove and a bending towards the major grooves. In agreement with biochemical data this suggests that flexibility and minor groove geometry are major parameters determining the cutting rate of DNase I. Experimental observations showing that the sequence environmental of a dinucleotide step strongly affects its cleavage efficiency can be rationalized by that fact that six base pair are in contact with the enzyme. Mutational analysis based on the structural results has identified critical residues for DNA binding and cleavage and has lead to a proposal for the catalytic mechanism.