Site-specific cleavage of human chromosome 4 mediated by triple-helix formation

Chimeric restriction endonuclease

Fok I restriction endonuclease recognizes the nonpalindromic pentadeoxyribonucleotide 5'-GGATG-3'.5'-CATCC-3' in duplex DNA and cleaves 9 and 13 nt away from the recognition site. Recently, we reported the presence of two distinct and separable domains within this enzyme: one for the sequence-specific recognition of DNA (the DNA-binding domain) and the other for the endonuclease activity (the cleavage domain). Here, we report the construction of a chimeric restriction endonuclease by linking the Drosophila Ultrabithorax homeodomain to the cleavage domain (FN) of Fok I restriction endonuclease. The hybrid enzyme, Ubx-FN, was purified, and its cleavage properties were characterized. The hybrid enzyme shows the same DNA sequence-binding preference as that of Ubx; as expected, it cleaves the DNA away from the recognition site. On the 5'-TTAATGGTT-3' strand the hybrid enzyme cleaves 3 nt away from the recognition site, whereas it cuts the complementary 5'-AACCATTAA-3' strand 8, 9, or 10 nt away from the binding site. Similarly engineered hybrid enzymes could be valuable tools in physical mapping and sequencing of large eukaryotic genomes.

Solution structure of a pyrimidine.purine.pyrimidine DNA triplex containing T.AT, C+.GC and G.TA triples

BACKGROUND

Under certain conditions, homopyrimidine oligonucleotides can bind to complementary homopurine sequences in homopurine-homopyrimidine segments of duplex DNA to form triple helical structures. Besides having biological implications in vivo, this property has been exploited in molecular biology applications. This approach is limited by a lack of knowledge about the recognition by the third strand of pyrimidine residues in Watson-Crick base pairs.

RESULTS

We have therefore determined the solution structure of a pyrimidine.purine.pyrimidine (Y.RY) DNA triple helix containing a guanine residue in the third strand which was postulated to specifically recognize a thymine residue in a Watson-Crick TA base pair. The structure was solved by combining NMR-derived restraints with molecular dynamics simulations conducted in the presence of explicit solvent and counter ions. The guanine of the G-TA triple is tilted out of the plane of its target TA base pair towards the 3'-direction, to avoid a steric clash with the thymine methyl group. This allows the guanine amino protons to participate in hydrogen bonds with separate carbonyls, forming one strong bond within the G-TA triple and a weak bond to an adjacent T.AT triple. Dramatic variations in helical twist around the guanine residue lead to a novel stacking interaction. At the global level, the Y.RY DNA triplex shares several structural features with the recently solved solution structure of the R.RY DNA triplex.

CONCLUSIONS

The formation of a G.TA triple within an otherwise pyrimidine.purine.pyrimidine DNA triplex causes conformational realignments in and around the G.TA triple. These highlight new aspects of molecular recognition that could be useful in triplex-based approaches to inhibition of gene expression and site-specific cleavage of genomic DNA.

Rapid isolation of cosmid insert DNA by triple-helix-mediated affinity capture

A simple and rapid method for the isolation of cosmid insert DNA was developed based on triple-helix-mediated affinity capture (TAC). A modified cosmid was constructed from the SuperCos 1 cosmid vector by flanking the cloning site with two homopurine-homopyrimidine triple-helix-forming sequences. The cosmid DNA is digested with NotI restriction enzyme to release the insert DNA. The NotI-digested cosmid DNA is then combined with a biotinylated homopyrimidine oligonucleotide in an acidic buffer solution to form a triple-helix complex. The triple-helix complex is captured with streptavidin-coated magnetic beads. Insert DNA is eluted by adding a pH 9 buffered solution to the captured complex. The purified insert DNA is recovered with a yield of up to 95% and a purity of at least 95%. The isolated insert DNA was directly digested with CviJI restriction endonuclease to generate random fragments for shotgun sequencing.

Targeted mutagenesis of simian virus 40 DNA mediated by a triple helix-forming oligonucleotide

Triple-helical DNA can be formed by oligonucleotides that bind as third strands of DNA in a sequence-specific manner in the major groove in homopurine/homopyrimidine stretches in duplex DNA. Such triple helix-forming oligonucleotides have been used to inhibit gene expression by blocking transcription factor access to promoter sites in transient expression assays. In an alternative approach to genetic manipulation using triplex DNA, we show that triplex-forming oligonucleotides can be used to produce site-specific, targeted mutations in a viral genome in order to achieve a permanent, heritable effect on gene function and expression. We use a triplex-forming oligonucleotide linked to a psoralen derivative at its 5' end to achieve targeted mutagenesis in a simian virus 40 (SV40) vector genome. Site-specific triplex formation delivers the psoralen to the targeted site in the SV40 DNA. Photoactivation of the psoralen yields adducts and thereby mutations at that site. Mutations were produced in the target gene in over 6% of the viral genomes. DNA sequence analysis of the mutations in the target gene showed that all were in the targeted region, and 55% were found to be the same T:A-to-A:T transversion precisely at the targeted base pair. In control experiments, no mutagenesis above the background frequency in the assay was produced by a non-triplex-forming, psoralen-linked oligonucleotide unless a vast excess of this oligonucleotide was used, demonstrating the specificity of the targeted mutagenesis. This frequency of targeted mutagenesis of SV40 in monkey cells represents a 30-fold increase relative to similar experiments using lambda phage in bacteria, suggesting that fixation of the triplex-directed lesion into a mutation occurs more efficiently in mammalian cells. If the ability to reproducibly and predictably target mutations to sites in viral DNA in vitro by using modified oligonucleotides can be extended to DNA in vivo, this approach may prove useful as a technique for gene therapy, as a strategy for antiviral therapeutics, and as a tool for genetic engineering.

NMR structural studies on a nonnatural deoxyribonucleoside which mediates recognition of GC base pairs in pyrimidine-purine-pyrimidine DNA triplexes

As a part of our ongoing efforts to define the structural aspects of unusual pairing alignments in DNA triplexes by nuclear magnetic resonance spectroscopy, we have examined the structural role of a nonnatural deoxyribonucleoside, P1, that has been shown to mediate the recognition of GC base pairs in pyrimidine-purine-pyrimidine DNA triplexes [Koh, J.S., & Dervan, P.B. (1992) J. Am. Chem Soc. 114, 1470]. A qualitative interpretation of the NMR data indicates that this analog of protonated cytosine is readily accommodated in the third strand segment of an intramolecular triplex system. Furthermore, the observed NOE patterns position the imino and amino protons of P1 opposite the N7 and O6 atoms of guanine, respectively, consistent with the previously proposed pairing scheme.

Solution structure of a purine.purine.pyrimidine DNA triplex containing G.GC and T.AT triples

BACKGROUND

Oligonucleotide-directed triple helix formation allows sequence specific recognition of double helical DNA. This powerful approach has been used to inhibit gene transcription in vitro and to mediate single site specific cleavage of a human chromosome.

RESULTS

Using a combined NMR and molecular dynamics approach (including relaxation matrix refinement), we have determined the solution structure of an intramolecular purine.purine.pyrimidine (R.RY) DNA triplex containing guanines and thymines in the third strand to high resolution. Our studies define the G.GC and T.AT base triple pairing alignments in the R.RY triplex and identify the structural discontinuities in the third strand associated with the non-isomorphism of the base triples. The 5'-d(TpG)-3' base steps exhibit a pronounced increase in axial rise and reduction in helical twist, while the reverse is observed, to a lesser extent at 5'-d(GpT)-3' steps. A third groove is formed between the purine-rich third strand and the pyrimidine strand. It is wider and deeper than the other two grooves.

CONCLUSIONS

Our structure of the R.RY DNA triplex will be important in the design of oligonucleotide probes with enhanced specificity and affinity for targeting in the genome. The third groove presents a potential target for binding additional ligands.

Thermodynamic and kinetic studies of DNA triplex formation of an oligohomopyrimidine and a matched duplex by filter binding assay

The filter binding method was found to be a powerful method for studying the formation of triplexes composed of a single-stranded homopyrimidine and a duplex with a matched purine-pyrimidine tract. With this technique, we were able to determine thermodynamic and kinetic parameters for triplex formation between a homopyrimidine 19-mer (5'-TCCTCTTCTTTTCTTTCTT-3') and a duplex with sequence 5'-GCAGGAGAAGAAAAGAAAGAACG-3' for the purine strand. The experiments were performed over a wide pH range (3.8-7.4) and a temperature range of 0-35 degrees C. pH and temperature dependencies of the thermodynamic parameters were best explained in terms of a three-state model for triplex formation at low temperatures relative to the melting point. The main results were as follows: (1) pH dependence of the dissociation constants of the triplex is a result of the rapid acid-base equilibrium of pyrimidine single strands; (2) the association rate for triplex formation decreases with increasing pH in accordance with the dissociation constants; (3) the dissociation constant is virtually temperature-independent at low pH, while it becomes strongly temperature-dependent with increasing pH (these results can be explained in terms of a negative, non-zero delta Cp for triplex formation at low pH); (4) the association rate decreases with increasing temperature, and the resulting negative activation energy indicates that the triplex formation process involves a quasi-stable intermediate; (5) the triplex formation is a second-order reaction at low pH, whereas it can be interpreted as a third-order reaction at neutral pH, suggesting that different triplex formation pathways are observed depending on the pH.

Targeted mutagenesis of DNA using triple helix-forming oligonucleotides linked to psoralen

Oligonucleotides can bind as third strands of DNA in a sequence-specific manner in the major groove in homopurine/homopyrimidine stretches in duplex DNA. Here we use a 10-base triplex-forming oligonucleotide linked to a psoralen derivative at its 5' end to achieve site-specific, targeted mutagenesis in an intact, double-stranded lambda phage genome. Site-specific triplex formation delivers the psoralen to the targeted site in the lambda DNA, and photoactivation of the psoralen produces adducts and thereby mutations at that site. Mutations in the targeted gene were at least 100-fold more frequent than those in a nontargeted gene, and sequence analysis of mutations in the targeted gene showed that 96% were in the targeted region and 56% were found to be the same T.A to A.T transversion precisely at the targeted base pair. The ability to reproducibly and predictably target mutations to sites in intact duplex DNA by using modified oligonucleotides may prove useful as a technique for gene therapy, as an approach to antiviral therapeutics, and as a tool for genetic engineering.

In vivo transcription of a progesterone-responsive gene is specifically inhibited by a triplex-forming oligonucleotide

Oligonucleotides provide novel reagents for inhibition of gene expression because of their high affinity binding to specific nucleotide sequences. We describe a 38 base, single-stranded DNA that forms a triple helix or 'triplex' on progesterone response elements of a target gene. This triplex-forming oligonucleotide binds with a Kd = 100 nM at 37 degrees C and physiological pH, and blocks binding of progesterone receptors to the target. Furthermore, it completely inhibited progesterone receptor-dependent transcription in vitro. To approach in vivo conditions, triplex-forming oligonucleotides were tested in cell transfection studies. The derivation of the oligonucleotides with cholesterol enhanced their cellular uptake and nuclear concentration by at least four-fold. The cholesterol-derivatized triplex-forming oligonucleotide specifically inhibited transcription of the PRE-containing reporter gene in cells when applied to the medium at micromolar concentrations. This is the first demonstration of steroid-responsive gene inhibition by triplex formation and joins the growing body of evidence indicating that oligonucleotides have therapeutic potential.

Inhibition of gene transcription by purine rich triplex forming oligodeoxyribonucleotides

Several oligodeoxynucleotides (ODNs) were designed in order to interact with the purine rich element of the IRE (Interferon Responsive Element) of the 6-16 gene by triplex formation. An ODN of 21 bases, the sequence being identical to that of the purine strand of the IRE (48% G), but in reverse orientation, was able to interact with the IRE (KD: 20 nM). The binding was Mg2+ dependent. The two purine strands of the triplex were oriented antiparallel as confirmed by DNAase I and copper-phenanthroline footprinting experiments. An ODN in which A were replaced by T, also interacted with the same target, but with a lower affinity. Exonuclease III action indicated that the two IRE repeats of the 6-16 promoter interacted with each other through Hoogsteen base pairing, the third strand being parallel to the paired Watson-Crick strand. This led to a potential H-DNA structure which could be destabilized by adding ODNs able to form a triplex structure. 6-16 IRE driven-reporter gene constructs lost their interferon stimulability when co-transfected with triplex forming ODNs. The range of effective ODN concentrations was compatible with the affinity determined when measuring their direct interactions with the DNA.

A new method for specific cleavage of megabase-size chromosomal DNA by lambda-terminase

The development of methods for cleavage of DNA at specific site(s) that are widely spaced would facilitate physical mapping of large genomes. Several methods for rare and specific cleavage of chromosomal DNAs require a nearly complete methylation of a given type of restriction site except the one that is specifically protected. It is expected that as the target DNA increases in length, it will become less likely to achieve nearly complete methylation. The intron-encoded endonucleases may also provide a capability to cleave megabase-sized DNA segments due to their very large recognition sequences. However, there are endogenous cleavage sites in the chromosomes of most organisms. We present here a new method to specifically cleave intact chromosomal DNA using lambda-terminase. A plasmid containing two specific cleavage sites (cohesive-end sites) for lambda-terminase was specifically introduced into the E.coli genome and into chromosome V of S.cerevisiae. Chromosomal DNA was prepared from the resulting strains, and then cleaved with lambda-terminase. The results showed that the 4.7-megabase pair (Mb) circular E.coli chromosome and the 0.58-Mb linear yeast chromosome V were specifically cleaved at the desired sites with very high efficiencies. The approach of using the lambda-terminase cleavage reaction is a simple one-step procedure with a high specificity which is particularly suitable for mapping very large genomes of eucaryotes.

Sequence-specific recognition of double helical RNA and RNA.DNA by triple helix formation

The stabilities of eight triple helical pyrimidine.purine.pyrimidine structures comprised of identical sequence but different RNA (R) or DNA (D) strand combinations were measured by quantitative affinity cleavage titration. The differences in equilibrium binding affinities reveal the importance of strand composition. For the sequences studied here, the stabilities of complexes containing a pyrimidine third strand D or R and purine.pyrimidine double helical DD, DR, RD, and RR decrease in order: D + DD, R + DD, R + DR, D + DR > R + RD, R + RR >> D + RR, D + RD (pH 7.0, 25 degrees C, 100 mM NaCl/1 mM spermine). These findings suggest that RNA and DNA oligonucleotides will be useful for targeting (i) double helical DNA and (ii) RNA.DNA hybrids if the purine Watson-Crick strand is DNA. However, RNA, but not DNA, oligonucleotides will be useful for sequence-specific binding of (i) double helical RNA and (ii) RNA.DNA hybrids if the purine Watson-Crick strand is RNA. This has implications for the design of artificial ligands targeted to specific sequences of double helical RNA and RNA.DNA hybrids.

7,8-Dihydro-8-oxoadenine as a replacement for cytosine in the third strand of triple helices. Triplex formation without hypochromicity

Oligonucleotides containing thymine and cytosine (or 5-methylcytosine) bases are known to bind to specific homopurine sequences in double-stranded DNA by means of T.AT and C+.GC base triplets. Cytosine in the third strand of such triple helices can be completely replaced by 7,8-dihydro-8-oxoadenine a base which should not require protonation to form base triplets. Experiments using native PAGE and inhibition of triplex-directed photo-cross-linking demonstrate that triplexes with 7,8-dihydro-8-oxoadenine in the third strand are as stable at pH 6.0 as triplexes with 5-methylcytosine. The stability of triplexes with 7,8-dihydro-8-oxoadenine, unlike those with 5-methylcytosine, is not substantially diminished upon raising the pH to 7.4. Surprisingly, triplex formation with an oligonucleotide containing only thymine and 7,8-dihydro-8-oxoadenine was not associated with significant hypochromicity and could not be detected in conventional thermal denaturation experiments.

Targeting in linear DNA duplexes with two complementary probe strands for hybrid stability

A new in vitro hybridization reaction targets two short complementary RecA protein-coated DNA probes to homologous sequences at any position in a linear duplex DNA molecule. Stable hybrids are obtained after RecA protein removal when both complementary probe strands are present in a four-stranded hybrid, but not when one probe strand is present in a three-stranded hybrid. In four-stranded hybrids with one probe strand biotinylated and the other radiolabelled, the deproteinized hybrids can be isolated and detected by affinity capture on streptavidin-coated magnetic beads. RecA-mediated targeting of complementary biotinylated DNA probe strands allows the affinity capture of 48.5-kilobase duplex lambda genomic DNA. These reactions provide a means of isolating any desired duplex gene or chromosomal DNA fragment.

Triplex formation by oligodeoxyribonucleotides involving the formation of X.U.A triads

The stabilities of oligodeoxyribonucleotide triplexes containing a single pyrimidine-purine base pair, which interrupts an otherwise purine-pyrimidine base pair motif, were studied by UV melting experiments. The oligomer systems consisted of an oligodeoxyribonucleotide target duplex d-GAAGAAAAAAYAAAA/d-TTTTZTTTTTTCTTC, I.II(Y.Z), or d-GAAGAAAAAGUGAAA/d-TTTCACTTTTTCTTC, IV.V(U.A), where Y.Z is C.G, T.A, or U.A and U is deoxyuridine. The third strand oligodeoxyribopyrimidine was d-CTTCTTTTTTXTTTT, III(X), or d-CTTCTTTTTCXCTTT, VI(X), where C is 5-methyldeoxycytidine. Triplexes were observed in the system III.I.II(X.C.G) when X was T or U. This may involve formation of T. or U.C.G triads in which the 4-carbonyl of T or U serves as a hydrogen bond acceptor for the N4-amino group of C. Triplex formation between III(X) and I.II(T.A) was only observed when X was G. In contrast to T.A or C.G, it appears a U.A base pair in the duplex target is a much more versatile participant in triad formation. Thus, stable triplexes were observed in III.I.II(X.U.A) and in VI.IV.V(X.U.A) when X was C, C, T, or U. The formation of a T.U.A or U.U.A triad can occur if the T or U of III translates approximately 1.4 A into the major groove, thereby allowing the 3-NH of T or U to donate a hydrogen bond to the 4-carbonyl oxygen of U in the duplex. Formation of C. or C.U.A base triads could involve formation of a single hydrogen bond between the third strand N4-amino group of C or C and the 4-carbonyl group of U of the target.(ABSTRACT TRUNCATED AT 250 WORDS)

Sequence-selective recognition and cleavage of double-helical DNA

Single sites within long double-helical DNA molecules can be recognized by a variety of mechanisms. Different strategies have been used to adapt sequence-specific recognition to sequence-specific cleavage of duplex DNA. Any nucleic acid can be converted into an artificial nuclease by the attachment of a cleaving reagent. Alternatively, a sequence-specific ligand can be used to protect a methylase recognition site from methylation. The protected site may then be cleaved selectively by a restriction endonuclease (the so-called 'Achilles heel' cleavage technique). Recent developments in this area have shown that it is possible to cleave chromosomal DNA at single sites within bacterial and eukaryotic genomes.

Structure and applications of intermolecular DNA triplexes

Current DNA binding drugs are not sequence specific. Triplex-forming oligonucleotides will bind targeted duplex DNA sites in a sequence-specific manner. A new class of DNA binding molecules based on triple-helical DNA formation promises a sequence-specific method of targeting discrete regions of DNA. DNA modifying molecules linked to third strands have been shown to modify only regions of DNA to which they were targeted. Current research will increase the understanding of triplex DNA structure and will lead to improved DNA binding drugs.

RecA-AC: single-site cleavage of plasmids and chromosomes at any predetermined restriction site

We have developed a novel version of the Achilles' Cleavage (AC) reaction in which virtually any restriction site on DNA of any size can be converted to a unique cleavage site. We first polymerized RecA protein on a synthetic oligodeoxyribonucleotide (oligo) in the presence of a nonhydrolyzable ATP analogue to generate oligo:RecA nucleoprotein filaments. These filament were then incubated with plasmid or intact chromosomal DNA from Saccharomyces cerevisiae to form stable complexes in the yeast LEU2 gene at the target sequence identical (or complementary) to that of the oligo. When HhaII (HinfI) methyltransferase (M.HhaII) was added, all of the recognition sites for HhaII with the exception of the one protected by the RecA filament were methylated and thus no longer cleaved by the cognate restriction endonuclease (HinfI). After inactivation of the RecA and the M.HhaII, HinfI was used to efficiently cleave the plasmid or chromosome specifically at the targeted restriction site. Since oligos specific for any sequence can be easily synthesized and the other reagents necessary to perform RecA-mediated AC (RecA-AC) reactions on both plasmids and intact chromosomes are readily available, this procedure can be applied immediately to the precise dissection and analysis of genomic DNA from any source and to any other research problem requiring efficient, highly specific cleavage of DNA at predetermined sites.

Oligonucleotide-directed triple helix formation at adjacent oligopurine and oligopyrimidine DNA tracts by alternate strand recognition

A significant limitation to the practical application of triplex DNA is its requirement for oligopurine tracts in target DNA sequences. The repertoire of triplex-forming sequences can potentially be expanded to adjacent blocks of purines and pyrimidines by allowing the third strand to pair with purines on alternate strands, while maintaining the required strand polarities by combining the two major classes of base triplets, Py.PuPy and Pu.PuPy. The formation of triplex DNA in this fashion requires no unusual bases or backbone linkages on the third strand. This approach has previously been demonstrated for target sequences of the type 5'-(Pu)n(Py)n-3' in intramolecular complexes. Using affinity cleaving and DNase I footprinting, we show here that intermolecular triplexes can also be formed at both 5'-(Pu)n(Py)n-3' and 5'-(Py)n(Pu)n-3' target sequences. However, triplex formation at a 5'-(Py)n(Pu)n-3' sequence occurs with lower yield. Triplex formation is disfavored, even at acid pH, when a number of contiguous C+.GC base triplets are required. These results suggest that triplex formation via alternate strand recognition at sequences made up of blocks of purines and pyrimidines may be generally feasible.

Reagents for the site-specific cleavage of megabase DNA

The physical mapping of chromosomes will be facilitated by methods of breaking large DNA into manageable fragments, or cutting uniquely at genetic markers of interest. Key issues in the design of sequence-specific DNA cleaving reagents are the specificity of binding, the number of different sequences that can be targeted and the cleavage yield.

Integrating maps of chromosome 21

The past year has seen major progress in the construction of various types of maps of human chromosome 21. Perhaps more significantly, the chromosome 21 research community is making very significant progress on integration of these maps through the use of common resources and increased collaboration and communication.

Stabilities of double- and triple-strand helical nucleic acids

In this selected literature survey, we have seen that the stabilities of duplexes and triplexes are governed by the vertical base stacking, the horizontal specific base-paired H-bonding and the environmental parameters. The entropic contribution in the solvation/desolvation process is important in driving the aggregation of NA strands and duplex formation, but base stacking and specific H-bonding maintain the helical order. Triplex formation shares most of the physical environmental prerequisites with those of duplex NAs. However, some additional environmental conditions are often needed. Only in low pH solution is the polycytidylic strand protonated and, thus, it is possible for the strand to bind to a G.C duplex sequence to give the C+(G.C) triplex. High ionic strength is often necessary for the screening of inter-phosphate repulsion due to the high linear charge density in triplexes. The presence of specific counterions is important for complexation. In the absence of negative supercoiling, existence of an intramolecular triplex is rare except under very acidic conditions for the formation of C+(G.C)-type intramolecular triplex. As expected, the stabilities of both inter- and intramolecular triplexes increase with sequence length. The thermodynamic principles of helix-coil transition of oligo-duplex may be described by the van't Hoff relationship, which assumes a two-state cooperative melting profile. Thus, the enthalpy, entropy and free energy of transition can be evaluated from the experimental melting curves (e.g. OD, DSC). For polynucleotides, because of the non-two-state nature of transition, the simple van't Hoff relationship is no longer valid, and direct calorimetry is needed to obtain reliable thermodynamic parameters. The pH and salt concentration dependence of duplex stability can be formulated and derived from a van't Hoff equation. Base-stacking patterns are simple in duplexes but not so in triplexes due to the diversity in triplet schemes. The sequence dependence of base stacking for duplexes has been characterized and employed to predict the stability of an arbitrary sequence. In conclusion, the stability of duplex is relatively well-characterized by thermodynamic data in terms of both base stacking and specific H-bonding. Thermodynamic studies of triplexes have been far fewer in number. Oligonucleotides have found application in the detection and localization of a mRNA or its gene, the detection of bacterial or viral sequences, and the inhibition of the translation of mRNA and the transcription and replication of DNA (Englisch and Gauss, 1991). In a different approach, oligonucleotides have been targeted directly to a DNA duplex motif of a gene in order to inhibit the expression at the beginning of the transcriptional process.(ABSTRACT TRUNCATED AT 400 WORDS)