Sequence-selective recognition and cleavage of double-helical DNA

Molecular dynamics simulation studies for DNA sequence recognition by reactive metabolites of anticancer compounds

The discovery of novel anticancer molecules 5F-203 (NSC703786) and 5-aminoflavone (5-AMF, NSC686288) has addressed the issues of toxicity and reduced efficacy by targeting over expressed Cytochrome P450 1A1 (CYP1A1) in cancer cells. CYP1A1 metabolizes these compounds into their reactive metabolites, which are proven to mediate their anticancer effect through DNA adduct formation. However, the drug metabolite-DNA binding has not been explored so far. Hence, understanding the binding characteristics and molecular recognition for drug metabolites with DNA is of practical and fundamental interest. The present study is aimed to model binding preference shown by reactive metabolites of 5F-203 and 5-AMF with DNA in forming DNA adducts. To perform this, three different DNA crystal structures covering sequence diversity were selected, and 12 DNA-reactive metabolite complexes were generated. Molecular dynamics simulations for all complexes were performed using AMBER 11 software after development of protocol for DNA-reactive metabolite system. Furthermore, the MM-PBSA/GBSA energy calculation, per-nucleotide energy decomposition, and Molecular Electrostatic Surface Potential analysis were performed. The results obtained from present study clearly indicate that minor groove in DNA is preferable for binding of reactive metabolites of anticancer compounds. The binding preferences shown by reactive metabolites were also governed by specific nucleotide sequence and distribution of electrostatic charges in major and minor groove of DNA structure. Overall, our study provides useful insights into the initial step of mechanism of reactive metabolite binding to the DNA and the guidelines for designing of sequence specific DNA interacting anticancer agents.

Artificial DNA cutters for DNA manipulation and genome engineering

This tutorial review provides recent developments in artificial cutters for site-selective scission of DNA with the focus on chemistry-based DNA cutters. They are useful tools for molecular biology and biotechnology, since their site-selectivity of scission is much higher than that of naturally occurring restriction enzymes and also their scission site is freely chosen. In order to prepare these cutters, a DNA-cutting molecule is combined with a sequence-recognizing molecule in a covalent or non-covalent way. At targeted sites in single-stranded and double-stranded DNAs, the scission occurs via either oxidative cleavage of nucleotides or hydrolysis of phosphodiester linkages. Among many successful examples, an artificial restriction DNA cutter, prepared from Ce(iv)/EDTA and pseudo-complementary peptide nucleic acid, hydrolyzed double-stranded DNA at the target site. The scission site and scission specificity are determined simply in terms of the Watson-Crick rule so that even the whole genome of human beings was selectively cut at one predetermined site. Consistently, homologous recombination in human cells was successfully promoted by this tool. For the purpose of comparison, protein-based DNA cutters (e.g., zinc finger nucleases) are also briefly described. The potential applications of these cutters and their future aspects are discussed.

Interstrand photocrosslinking of DNA via p-carbamoylvinyl phenol nucleoside

We report a novel interstrand photocrosslinking of oligodeoxynucleotides (ODNs). In this system, a modified ODN containing p-carbamoylvinyl phenol nucleoside reacts by photoirradiation at 366 nm with adenine residue of a complementary template ODN to yield a crosslinked ODN in 97% yield.

Amplification of C1027-induced DNA cleavage and apoptosis by a quinacrine–netropsin hybrid molecule in tumor cell lines

We examined the effect of a newly synthesized DNA-binding ligand, quinacrine-netropsin hybrid molecule (QN), on cytotoxicity, apoptosis, and DNA strand breaks induced by an enediyne antitumor antibiotic, C1027. QN significantly enhanced C1027-induced cellular DNA strand breaks, caspase-3 activation, and DNA ladder formation, characteristic of apoptosis, in human HL-60 cells. Flow cytometry revealed that C1027-induced intracellular H(2)O(2) generation was enhanced by QN, suggesting that QN enhances C1027-induced cytotoxic effect through H(2)O(2)-mediated apoptosis. QN also significantly enhanced C1027-induced apoptosis in BJAB cells, and the inhibition of apoptosis was observed in BJAB cells transfected with Bcl-2 gene. The experiment using (32)P-labeled DNA fragments showed that the addition of QN enhanced C1027-induced double-stranded DNA cleavage at the 5'-AGG-3'/3'-TCC-5' sequence (cutting sites are underlined). These results suggest that QN enhances C1027-induced antitumor effect via DNA cleavage and apoptosis. The present study shows a novel approach to the potentially effective anticancer therapy.

Thiazole Orange−Peptide Conjugates: Sensitivity of DNA Binding to Chemical Structure

[structure: see text] Derivatives of the highly fluorescent and DNA-binding dye thiazole orange (TO) are described that feature appended peptides. Functionalization of TO can be achieved at either of the endocyclic nitrogens, and the photophysical properties and DNA-binding modes are sensitive to the position of the tethered peptide. A series of TO-peptide conjugates are described, demonstrating the utility of a solid-phase synthesis approach to their preparation and illustrating how the photophysical and DNA-binding properties of the compounds are influenced by chemical structure.

Tiny telomere DNA

We describe the design, synthesis and biophysical characterization of a novel DNA construct in which a folded quadruplex structure is joined to a standard double helix. Circular dichroism, gel electrophoresis, three-dimensional UV melting and differential scanning calorimetry were all used to characterize the structure. Rigorous molecular dynamics simulations were used to build a plausible atomic-level structural model of the DNA construct. This novel DNA construct provides a model for the duplex-quadruplex junction region at the end of chromosomal DNA and offers a system for the study of structure-selective ligand binding.

DNA and its associated processes as targets for cancer therapy

DNA is the molecular target for many of the drugs that are used in cancer therapeutics, and is viewed as a non-specific target of cytotoxic agents. Although this is true for traditional chemotherapeutics, other agents that were discovered more recently have shown enhanced efficacy. Furthermore, a new generation of agents that target DNA-associated processes are anticipated to be far more specific and effective. How have these agents evolved, and what are their molecular targets?

Double duplex invasion by peptide nucleic acid: a general principle for sequence-specific targeting of double-stranded DNA

Pseudocomplementary PNAs containing diaminopurine.thiouracil base pairs have been prepared and are shown to bind with high specificity and efficiency to complementary targets in double-stranded DNA by a mechanism termed "double duplex invasion" in which the duplex is unwound and both DNA strands are targeted simultaneously, each by one of the two pseudocomplementary peptide nucleic acids (PNAs). On the basis of our results we predict that (for decameric targets) more than 80% of all sequences can be targeted by straightforward Watson-Crick base pairing by using this approach in its present form. Targeting of pseudocomplementary PNAs to the promoter of the T7 phage RNA polymerase effectively inhibits transcription initiation. These results have important implications in the development of gene therapeutic agents as well as for genetic diagnostic and molecular biology applications.

Triple-helix formation at different positions on nucleosomal DNA

We have prepared a series of seven DNA fragments, based on the 160 base-pair tyrT sequence, which contain 12-14 base-pair oligopurine tracts at different positions, and have examined their availability for triple-helix formation after reconstituting onto nucleosome core particles. By using DNase I footprinting we find that in general, triplexes can only be formed at sites located toward the ends of nucleosomal DNA fragments. For the native fragment, bases 1-145 are in contact with the protein surface. Stable triplexes can be formed on these nucleosome-bound fragments for sites located before position 33 and beyond position 94. These are formed with both CT-containing oligonucleotides, generating parallel triplexes at pH 5.5, and GT-containing oligonucleotides forming antiparallel triplexes at pH 7.5. No antiparallel triplexes were formed at sites located between these positions. Parallel triplexes were also not formed at sites between positions 39-50 and 43-54 with oligonucleotide concentrations as high as 30 microM. However parallel triplex formation was evident at a site between positions 48 and 59, albeit with a reduced affinity compared to free DNA, suggesting that this oligopurine tract is less tightly associated with the nucleosome surface or that it has an altered translational position. The introduction of an oligopurine tract in the vicinity of the nucleosome dyad caused the fragment to adopt a different nucleosomal position, which could be targeted with parallel, but not antiparallel triplexes.

A novel palindromic triple-stranded structure formed by homopyrimidine dodecamer d-CTTCTCCTCTTC and homopurine hexamer d-GAAGAG

We have carried out NMR and molecular mechanics studies on a complex formed when a palindromic homopyrimidine dodecamer (d-CTTCTCCTCTTC) and a homopurine hexamer (d-GAAGAG) are mixed in 1:1 molar ratio in aqueous solutions. Such studies unequivocally establish that two strands of each oligomer combine to form a triple-stranded DNA structure with a palindromic symmetry and with six T.A:T and six C+. G:C hydrogen-bonded base triads. The two purine strands are placed head to head, with their 3' ends facing each other in the center of the structure. One-half of each pyrimidine strand contains protonated and the other half contains non-protonated cytosines. The two half segments containing protonated cytosines are hydrogen bonded to each of the two purine hexamers through Hoogsteen T.A and C+.G base pairing. The segments containing non-protonated cytosines are involved in Watson-Crick (A:T and G:C) base pairing. This leads to a palindromic triplex with a C2-dyad symmetry with respect to the center of the structure. The complex is less stable at neutral pH, but the cytosines involved in Hoogsteen base pairing remain protonated even under these conditions. Molecular mechanics calculations using NMR constraints have provided a detailed three-dimensional structure of the complex. The entire stretches of purine, and the pyrimidine nucleotides have a conformation close to B-DNA.

Stabilizing and destabilizing effects of intercalators on DNA triplexes

Oligonucleotide-directed triplex formation attracts much attention due to its potential usefulness in diagnostic and biotechnological applications. Among other aspects, the research embraces numerous studies probing the influence of intercalating ligands on triplex stability. The effect of the intercalator on triplex formation and stability is known to depend on nucleotide sequence, type of intercalator and solution conditions. The present work is aimed at determining the average number of intercalated ethidium bromide (EtBr) and acridine orange (AO) molecules leading to the most effective stabilization of triplexes. First, fluorescing complexes of intramolecular parallel (recombinant) triplex 5'-d(CATGCTAACT)-L-d(AGTTAGCATG)-L-d(CATGCTAACT)-3' (parARB) and classical antiparallel 5'-(dA)10-L-(dT)10-L-(dT)10-3'(antiATT) (L = -pO(CH2CH2O)3p-) with EtBr and AO were characterized, binding constants were obtained and compared to those for homologous DNA duplexes. Then the total EtBr and AO concentrations corresponding to an average of one, two or three intercalated molecules per oligonucleotide were estimated. Thermal denaturation of parARB and antiATT complexes with an average of one, two or three bound molecules was carried out, thermodynamic parameters of the triplex-to-duplex and duplex-to-open-strand transitions were evaluated using a three-state model. The ability of EtBr and AO to stabilize or destabilize both parallel (recombinant) and classical antiparallel triplexes was found to depend strongly on the concentration of bound intercalator. The triplexes were shown to be stabilized by intercalation of the first and second EtBr or AO molecules, while binding of the third intercalator molecule to 10 nucleotide long triplex resulted in significant triplex destabilization.

Extended DNA-recognition repertoire of peptide nucleic acid (PNA): PNA-dsDNA triplex formed with cytosine-rich homopyrimidine PNA

Peptide nucleic acid (PNA) is an oligonucleotide mimic in which the backbone of DNA has been replaced by a pseudopeptide. Thymine-rich homopyrimidine PNA oligomers have been found to recognize double-stranded DNA targets by displacement of the pyrimidine DNA strand and forming an internal Watson-Crick-Hoogsteen base-paired PNA(pyr)-DNA(pu)-PNA(pyr) triplex. We here show that cytosine-rich homopyrimidine PNA sequences instead add to double-stranded polynucleotide targets as Hoogsteen strands forming PNA(pyr)-DNA(pu)-DNA(pyr) triplexes. Furthermore, PNA strands with homopurine or alternating thymine-guanine sequences are shown to invade their respective DNA targets by displacing the identical DNA strands of the polynucleotides and forming new PNA-DNA duplexes. These results indicate distinct mechanistic variations as to how PNA interacts with a DNA target depending on choice of nucleobases, which could be of importance for future design of gene-specific diagnostic or therapeutic agents.

RecA-mediated Achilles' heel cleavage

The specific protection of only one of many restriction sites in a genome from inactivation by a cognate methyltransferase (MTase) creates a unique cleavage site - an Achilles' heel cleavage (AC) site. In the RecA-AC, or RARE, technique, such specific protection is provided by a synaptic complex composed of RecA protein, a gamma-S analog of ATP and a 30-60 nucleotide long oligodeoxynucleotide complementary or identical to the sequence-targeted site in which the protected restriction site is embedded. Upon methylation and the subsequent removal of the protective complex and MTase, the protected site is the only site cut by the cognate restriction enzyme. Two such targeted cuts permit the excision of a unique DNA fragment from the genome. Recent advances include the calibration of DNA clones, the mapping of gaps, and the determination of the sizes of excised fragments by pulsed-field gel electrophoresis, which allows one to measure distances between any two neighboring sequence-targeted sites, in the range of a few kilobases to 10 megabases, with the purpose of physically mapping the genome.

Hydrated water molecules of pyrimidine/purine/pyrimidine DNA triple helices as revealed by FT-IR spectroscopy: a role of cytosine methylation

Hydrated water molecules of pyrimidine/purine/pyrimidine DNA hairpin triplex was studied by a comparison of triplex (CC.AG6) formed by a host oligodeoxypyrimidine of 5'-d(TC)3T4(CT)3(CC) with a target hexadeoxypurine 5'-d(AG)3(AG6) strand and by triplexes (MM.AG6, MC.AG6, and CM.AG6) formed by oligonucleotides with the exact sequences as above except 5-methylcytosine replaced all (MM), 5' end half (MC), and 3' end half (CM) cytosine bases in CC via FT-IR spectroscopy in hydrated film. Results revealed that: (i) all these triplexes have a similar hydration pattern, in which water molecules probably bound in the N7 sites of adenines and guanines in the Crick-Hoogsteen groove, and to the methyl group of thymidines in the Watson-Hoogsteen groove. There are also some bound water molecules found at the O2 sites of thymines in both Watson-Crick and Crick-Hoogsteen grooves. (ii) In the CC.AG6 triplex the S-type sugars are always dominant in all hydrated states, whereas in MM.AG6 triplex the relative population of the N-type sugars is very close to that of the S-type between 86% and 66% of humidity. Furthermore, the sugar conformation in two partially modified triplexes (CM.AG6, and MC.AG6) are dominant by the N-type at lower humidity. This phenomenon might reflect that the degree of bound water varies among the binding sites of bases. (iii) The effect of introducing a methyl group on cytosine is to generate a spine of hydrophobic region in MM (MC and MC). The enlarging hydrophobic area not only increase the stability in solution, and also the stability in sodium hydrated films of the pyrimidine/purine/pyrimidine hairpin triplexes.

Specificity of antiparallel DNA triple helix formation

We have used DNase I footprinting to examine the formation of antiparallel DNA triple helices on DNA fragments containing the homopurine target sites (GGA)2GGX(GGA)2GG.(CCT)2CCZ(CCT)2CC (where X.Z is each base pair in turn), with the GA- and GT-rich oligonucleotides, (GGA)2GGN(GGA)2GG and (GGT)2GGN(GGT)2GG (N = each base in turn). These were designed to form G.GC and A.AT or T.AT triplets with a central N.XZ mismatch, which should bind in an antiparallel orientation. We find that almost all combinations generate DNase I footprints at low micromolar concentrations. At each target site, the relative binding of the GA- and GT-containing oligonucleotides was not the same, suggesting that these two triplexes adopt different conformations. For a central GC base pair, the most stable complex is observed with a third strand generating a G.GC triplet as expected. A.GC is also stable, especially in the GT oligonucleotides. For a central AT base pair, all four bases form stable complexes though T.AT is favored for the GA-rich thirds strands and A.AT for the GT-rich strands. For a central CG base pair, the stable complexes are seen with third strands generating T.CG triplets, though A.CG and C.CG are stable with GT- and GA-containing oligonucleotides, respectively. C.TA is the best triplet at a central TA base pair. The third strands with central guanines avoided the formation of G.YR triplets on the fragments containing central pyrimidines, producing DNase I footprints which had slipped relative to the target site. These oligonucleotides bound at a different location, generating complexes containing 11 contiguous stable triplets at the 3'-end of the third strand. The results suggest rules for designing the best third strand oligonucleotides for targeting sequences in which homopurine tracts are interrupted by pyrimidines.

Mutagenesis in mammalian cells induced by triple helix formation and transcription-coupled repair

When mammalian cells were treated with triplex-forming oligonucleotides of sufficient binding affinity, mutations were specifically induced in a simian virus 40 vector contained within the cells. Triplex-induced mutagenesis was not detected in xeroderma pigmentosum group A cells nor in Cockayne's syndrome group B cells, indicating a requirement for excision repair and for transcription-coupled repair, respectively, in the process. Triplex formation was also found to stimulate DNA repair synthesis in human cell extracts, in a pattern correlating with the inhibition of transcription in such extracts. These findings may have implications for therapeutic applications of triplex DNA and raise the possibility that naturally occurring triple helices are a source of genetic instability.

Transcription-mediated binding of peptide nucleic acid (PNA) to double-stranded DNA: sequence-specific suicide transcription

Peptide nucleic acid (PNA) forms sequence-specific (PNA)2/DNA triplexes with one strand of double-stranded DNA by strand invasion. When formed with the template strand of DNA such a (PNA)2/DNA triplex can arrest transcription elongation in vitro and can thus act as an anti-gene agent. One of the major obstacles to applying PNA as an anti-gene agent in vivo is that PNA strand invasion occurs at a very slow rate under moderate salt conditions. In the present study we show that transcription can increase the rate of sequence-specific PNA binding dramatically. Such transcription-mediated PNA binding occurs three times as efficiently when the PNA target is situated on the non- template strand as compared with the template strand. Since transcription can mediate template strand-associated (PNA)2/DNA complexes which arrest further elongation, the action of RNA polymerase results in repression of its own activity, i.e. suicide transcription. These findings are highly relevant for the possible future use of PNA as an anti-gene agent.

Effect of selective cytosine methylation and hydration on the conformations of DNA triple helices containing a TTTT loop structure by FT-IR spectroscopy

5-Methylcytosines have been introduced into triplex-forming-oligonucleotides and shown to extend the pH range over which a triplex forms with a homopurine-homopyrimidine tract of duplex DNA. As a host strand, an oligodeoxypyrimidine with a base sequence of 5'-d(TC)3T4(CT)3 ([CC]) was designed to form a hairpin triplex with a 5'-d-A(GA)2G ([AG6]) purine strand at acidic pH (Tsay, et al., (1995) J. Biomol. Str. Dyn., 13, 1235-1245). We here present results obtained by FT-IR spectroscopy concerning the conformation of the hairpin triplex as a function of the selective substitution of cytosines by 5-methylcytosines in the host strand. Namely, cytosines are substituted by 5-methylcytosines in either the 3'-pyrimidine portion ([CM]) or the 5'-pyrimidine portion ([MC]) or in both ([MM]) of the host strand. The acidic-induced transitions of the equimolar mixtures of the purine target with either of the four pyrimidine oligomers gives rise to different apparent pK values, i.e., [MM].[AG6] (6.2) > [MC].[AG6] (6.0) > [CM].[AG6] (5.7) > [CC].[AG6] (5.2) > single-stranded oligopyrimidines (4.6 +/- 0.2), indicating that cytosine methylation expands the pH range compatible with the hairpin triplex formation regardless of whether the substitution is in the 5'-pyrimidine (Hoogsteen) portion or in the 3'-pyrimidine (Watson-Crick) portion. Thermal denaturation profiles indicated that all the triplexes denatured in a monophasic manner in the pH range of 4.0 to 7.0, and that cytosine methylations in any position of the 16-base pyrimidine oligomer increase the stability of the hairpin triplex DNA. IR spectra recorded in D2O and H2O solutions revealed that cytosine methylation does not significantly influence the conformation of triplex DNA in solution, i.e., all the four triplexes accept a similar sugar conformation, and predominately take on a S-type sugar pucker with a relative proportion of two S-type sugars for one N-type. Furthermore, we also investigated the effect of relative humidity (RH) on the conformation of triplex MC.AG6 in hydrated films, and found that the conformational change induced by the decrease of RH, from predominant S-type to primary N-type sugar pucker, might first occur in the purine strand at 86% RH.

Sequence-specific transcription arrest by peptide nucleic acid bound to the DNA template strand

The effects of PNA (peptide nucleic acid) bound to double-stranded (ds) DNA targets positioned downstream from phage T3 or T7 promoters in pBluescriptKS+ derived plasmids on transcription by RNA polymerases T3 or T7 have been studied. The dsDNA targets A10, 5'-A5GA4 or 5'-A2GA2GA4, and the corresponding PNAs T10, T5CT4 and T2CT2CT4 were used and the target-PNA strand displacement complexes were performed in low-salt buffer, since PNA does not bind efficiently to ds DNA in higher salt than 50 mM. It is shown that transcription elongation is arrested at the target site with PNA bound to the template strand, whereas only a marginal effect is observed with PNA bound to the non-template strand. With PNA T10, transcription arrest occurs at the first base of the PNA-binding site, while the arrest with the PNA T5CT4 takes place 2-3 nt inside the PNA binding site. In the case of PNA T2CT2CT4 the arrest is less efficient and occurs at the last 1-3 nt of the binding site. Transcription arrest was also shown for PNAs T6 and T8, although with a much lower efficiency. These results show that efficient transcription elongation arrest can be obtained by PNA targeting of the template DNA strand.

Comparison of antiparallel A.AT and T.AT triplets within an alternate strand DNA triple helix

We have examined the formation of alternate strand triple-helices at the target sequence A11(TC)6.(GA)6T11 using the oligonucleotides T11(AG)6 and T11(TG)6, by DNase I footprinting. These third strands were designed so as to form parallel T.AT triplets together with antiparallel G.GC and A.AT or T.AT triplets. We find that, although both oligonucleotides yield clear footprints at similar concentrations (0.3 microM) in the presence of manganese, only T11(TG)6 forms a stable complex in magnesium-containing buffers, albeit at a higher concentration (10-30 microM). Examination of the interaction of (AG)6 and (TG)6 with half the target site confirmed that the complex containing A.AT triplets was only stable in the presence of manganese. In contrast no binding of (TG)6 was detected in the presence of either metal ion, suggesting that the reverse-Hoogsteen T.AT triplet is less stable that G.GC. We suggest that, within the context of G.GC triplets, the rank order of antiparallel triplet stability is A.AT (Mn2+) > T.AT (Mn2+) > T.AT (Mg2+) > A.AT (Mg2+). Third strands containing a single base substitution in the centre of either the parallel or antiparallel portion showed a (10-fold) weaker interaction in manganese-containing buffers, and no interaction in the presence of magnesium.

Site-specific targeting of psoralen photoadducts with a triple helix-forming oligonucleotide: characterization of psoralen monoadduct and crosslink formation

A polypurine tract in the supF gene of bacteriophage lambda (base pairs 167-176) was selected as the target for triple helix formation and targeted mutagenesis by an oligopurine (5'-AGGAAGGGGG-3') containing a chemically linked psoralen derivative (4'-hydroxymethyl-4,5',8-trimethylpsoralen) at its 5' terminus (psoAG10). The thymines at base pairs 166 and 167, a 5'ApT site, were targeted for photomodification. Exposure of the triple helical complex to long wavelength ultraviolet radiation led to the covalent binding of psoAG10 to the targeted region in the supF gene and to the induction of site-specific mutations. We report here experiments to characterize the photomodification of the targeted region of the supF gene in the context of triple helix formation. An electrophoretic mobility-shift assay showed that, at low radiation doses, monoadducts at base pair 166 were the major photoadducts. At higher doses the monoadducts were converted to crosslinks between base pairs 166 and 167. HPLC analysis of enzymatically hydrolyzed photoreaction mixtures was used to confirm the electrophoresis results. A strong strand preference for specific photoadduct formation was also detected.

Stability of triple helices containing RNA and DNA strands: experimental and molecular modeling studies

UV-absorption spectrophotometry and molecular modeling have been used to study the influence of the chemical nature of sugars (ribose or deoxyribose) on triple helix stability. For the Pyrimidine.purine* Pyrimidine motif, all eight combinations were tested with each of the three strands composed of either DNA or RNA. The chemical nature of sugars has a dramatic influence on triple helix stability. For each double helix composition, a more stable triple helix was formed when the third strand was RNA rather than DNA. No stable triple helix was detected when the polypurine sequence was made of RNA with a third strand made of DNA. Energy minimization studies using the JUMNA program suggested that interactions between the 2'-hydroxyl group of the third strand and the phosphates of the polypurine strand play an important role in determining the relative stabilities of triple-helical structures in which the polypyrimidine third strand is oriented parallel to the polypurine sequence. These interactions are not allowed when the third strand adopts an antiparallel orientation with respect to the target polypurine sequence, as observed when the third strand contains G and A or G and T/U. We show by footprinting and gel retardation experiments that an oligoribonucleotide containing G and A or G and U fails to bind double helical DNA, while the corresponding DNA oligomers form stable triple-helical complexes.