Johan H. van de Sande

Left-handed DNA: from synthetic polymers to chromosomes.

The interconversions between right-handed (R) and left-handed (L) helical conformations of DNA have been assessed by spectroscopic, electrophoretic, immunochemical, and enzymatic techniques. We have screened salt and solvent conditions which facilitate these transitions, as well as certain chemical modifications of the bases and backbone of defined synthetic polynucleotides. These include major and minor groove substituents as well as phosphorothioate analogues of selected phosphodiester bonds. We have established: R-L transitions in poly[d(G-C)] with iodo, bromo, methyl, and aza substitutions at the C5 position of cytosine, or phosphorothioate modification of the dGpC linkage. R-L transitions in the [d(A-C).d(G-T)]n sequence family using polymers modified as in the case of poly[d(G-C)]. The isomerizations are highly salt and temperature dependent. a possible L form of poly[d(A-T)] substituted with 2-amino adenine. the immunogenicities of constitutive and facultative Z-DNAs. the recognition specificities of different anti-Z-DNA IgGs for the spectrum of available polynucleotide probes. Some IgGs are sequence-specific. stabilization by IgG of otherwise transient left-handed conformations. anti-Z-DNA IgG binding to acid-fixed polytene chromosomes from the Diptera Drosophila, Chironomus, and Glyptotendipes. Laser scanning microscopy shows a maximal binding of 1 IgG per 3000-15,000 basepairs in acid fixed preparations. anti-Z-DNA IgG binding to negatively supercoiled plasmid, viral, phage, and recombinant closed circular DNAs. transcription from Z and Z* (associated) left-handed templates. From these and other results we propose that Z*-DNA may have important structural-functional roles in the cell.

Left-handed helical conformation of poly[d(A-m5C).d(G-T)].

Poly[d(G-C)] serves as the prototype for the right-to-left (B to Z) transition in the helical sense of DNA, both in solution1 and in the crystal form2,3. However, the question remains as to which other synthetic and natural DNAs have the potential to adopt left-handed conformations. One logical candidate is the canonical alternating purine–pyrimidine sequence d(A-C)n · d(G-T)n which seems to be widely disseminated in eukaryotic genomes4,5. Our approach to this question is based on the enzymatic synthesis of poly[d(A-C) · d(G-U)] derivatives with systematic methyl and halogen substitutions in the C-5 position of the pyrimidines C and U. Such modifications in poly[d(G-C)] have previously been shown to potentiate the B to Z transition6,7. Here we report a highly cooperative, reversible, salt- and temperature-dependent transition for poly[d(A-m5C) · d(G-T)], a repeat of the d(A-m5C) sequence which may occur in natural DNA8. Spectroscopic studies and the demonstrated ability to bind anti-Z-DNA antibodies suggest that the new helical conformation is left-handed and shares structural features with known Z-DNA. However, a novel property, not exhibited by poly[d(G-C)]1, is the profound temperature dependence of the conformational equilibrium.

Covalently linked sequencing primer linkers (splinkers) for sequence analysis of restriction fragments

A new method for direct sequence analysis of DNA restriction fragments uses synthetic covalently linked complementary oligodeoxynucleotides, as universal sequencing primer linkers (splinkers). These splinkers were designed to contain an inverted repeat sequence which forms a double-stranded hairpin structure with a known restriction site. The splinkers were characterized by their ability to be self-ligating (dimerized) and by their restriction digest product analysis of both the monomer and dimer. They can also be ligated to dephosphorylated DNA restriction fragments which contain the appropriate end. This was evidenced by mobility shifts of the splinker-ligated restriction fragments. The splinker-ligated restriction fragments, after denaturation, form a single-stranded DNA template containing an inverted repeat sequence (from splinker) at one terminus. The splinker is thus suitably oriented and serves as a primer for dideoxy nucleotide (nt) sequencing catalyzed by either Klenow fragment of Escherichia coli DNA polymerase I or avian myeloblastosis virus reverse transcriptase. As demonstrated for both pBR322 and phi X174, release of the primer extension strand by digestion at the splinker restriction site results in a ladder of labelled fragments which corresponds to a unique nt sequence.

End group labelling of RNA and double stranded DNA by phosphate exchange catalyzed by bacteriophage T4 induced polynucleotide kinase

Abstract End group labelling of sheared double-stranded DNA, and tRNA has been effected without prior dephosphorylation, utilizing the reversal of T4 polynucleotide kinase activity. Incubation of DNA with polynucleotide kinase in the presence and absence of a phosphate acceptor (ADP) allowed the determination of the relative ratio of 5′ hydroxyl and 5′ phosphoryl terminii in the polynucleotide. This method of analysis has demonstrated a high preference in the formation of 5′ vs 3′ phosphomonoesters during high pressure shearing of double stranded DNA.

THE MOUSE URACIL-DNA GLYCOSYLASE GENE : ISOLATION OF CDNA AND GENOMIC CLONES AND MAPPING UNG TO MOUSE CHROMOSOME 5

Uracil-DNA glycosylase (UDG) is the enzyme responsible for the first step in the base-excision repair pathway that specifically removes uracil from DNA. Here we report the isolation of the cDNA and genomic clones for the mouse uracil-DNA glycosylase gene (ung) homologous to the major placental uracil-DNA glycosylase gene (UNG) of humans. The complete characterization of the genomic organization of the mouse uracil-DNA glycosylase gene shows that the entire mRNA coding region for the 1.83-kb cDNA of the mouse ung gene is contained in an 8.2-kb SstI genomic fragment which includes six exons and five introns. The cDNA encodes a predicted uracil-DNA glycosylase (UDG) protein of 295 amino acids (33 kDa) that is highly similar to a group of UDGs that have been isolated from a wide variety of organisms. The mouse ung gene has been mapped to mouse chromosome 5 using fluorescence in situ hybridization (FISH).

Base compositional analysis of nanogram quantities of unlabeled nucleic acids

Abstract After conversion of unlabeled DNA and RNA to 3′-mononucleotides accurate base compositional analysis can be performed on as little as 10 ng of the hydrolysate. The 3′-mononucleotides are first quantitatively postlabeled with [γ-32P]ATP by T4 polynucleotide kinase and are then separated as mononucleoside diphosphates on Whatman DE-81 ion-exchange paper at pH 3.5 after hydrolysis of surplus [γ-32P]ATP to 32P1. The locations of the four labeled nucleoside diphosphates are determined by autoradiography and the ratio of radioactivity in the four spots gives the base ratio of the sample.

Length-Dependent Cruciform Extrusion in d(GTAC)n Sequences

pBR322-derived plasmids have been constructed carrying d(GTAC)n.d(GTAC)n inserts of different lengths, in order to investigate the effect of insert size on cruciform extrusion and/or the B-Z transition. Plasmids with n ranging from 4 to 12 are hypersensitive to cleavage by the single-strand specific nucleases, S1 nuclease and Bal31 nuclease. Hypersensitive sites associated with the smaller alternating purine-pyrimidine tracts, however, coexist with the major pBR322 sites. Site-selective cleavage of these plasmids with the resolvase, T7 endonuclease I, demonstrates that all the inserts form cruciform structures when stably integrated into negatively supercoiled plasmids. An increase in the negative superhelical density of the DNAs induces cruciform formation within the insert region, resulting in a reduction in torsional stress consistent with the size of the insert. Moreover, as n decreases, the superhelical density required to stabilise the cruciform state increases. Therefore, the cruciform geometry is the favoured conformation of these d(GTAC)n.d(GTAC)n sequences under torsional stress. The stability of these cruciforms increases as n increases, with cruciformation occurring at lower superhelical densities and to the exclusion of the other pBR322 cruciforms.

Probing for and with Left-Handed DNA: Poly[d(A-br5c)•d(G-T)], a Member of a New Family of Z-Forming DNAs

Left-handed helical conformations can be adopted by the poly[d(A-C)•d(G-U)] family of DNAs bearing at least one methyl or halogen (bromine, iodine) substitution at the pyrimidine C5 heterocyclic position. A representative polynucleotide is poly[d(A-br5C)•d(G-T)]. Elevation of salt concentration and temperature shifts the equilibrium in favor of the left-handed state. The spectral changes (u.v. absorption, c.d., NMR, Raman) accompanying the highly co-operative and reversible R(right)-L(left) transitions are similar to those observed with poly[d(G-C)]. Some, but not all, antibodies raised against the left-handed members of the poly[d(G-C)] family also recognize the new class of polymers, thus establishing the existence of sequence-specific determinants in the binding of certain anti-Z DNA immunoglobulins. The latter have been used to establish the presence of left-handed regions in natural plasmid and viral DNAs and in the giant salivary gland polytene chromosomes of Chironomus thummi thummi. The Z-DNA specific immunofluorescence in fixed polytene chromosomes has been quantitated by laser scanning and photon counting and is particularly intense in certain regions, such as the telomeres.

DNA Duplexes Containing Alpha Anomeric Nucleotides and Polarity Reversals: Coexistence of Parallel and Antiparallel DNA

Oligodeoxynucleotides that possess alpha anomeric nucleotides and polarity reversals show promise for application in the area of antisense therapy. Here we provide a survey of the spectroscopic, thermodynamic, and enzymatic techniques used in our laboratories to investigate model systems containing such unnatural features with the ultimate goal of designing a new class of more potent and effective antisense therapeutics.

Homooligomeric dA·dU and dA·dT Sequences in Parallel and Antiparallel Strand Orientation: Consequence of the 5-methyl Groups on Stability, Structure and Interaction with the Minor Groove Binding Drug HOECHST 33258

Oligodeoxyribonucleotides containing dA.dU base combinations were shown to form parallel stranded DNA. CD spectra and hyperchromicity profiles provide evidence that the structure is very similar to that of a related parallel stranded dA.dT oligomer. Thermal denaturation studies show that these parallel dA.dU sequences are significantly less stable than their dA.dT analogues in either antiparallel or parallel stranded orientations. The stabilizing effect of the 5-methyl group is similar for parallel and antiparallel sequences. The minor groove binding drug Hoechst 33258 binds with similar affinity to APS dA.dT and APS dA.dU sequences. However, binding to the PS dA.dT hairpin is significantly impaired as a consequence of the different groove dimensions and the presence of thymine methyl groups at the binding site. This results in an 8.6 kJmol-1 reduced free energy of binding for the PS dA.dT sequence. Replacement of the bulky methyl group with a hydrogen (ie. T-->U) results in significantly stronger Hoechst 33258 binding to the parallel dA.dU sequences with a penalty of only 4.1 kJmol-1. Our data demonstrate that although Hoechst 33258 detects the altered groove, it is still able to bind a PS duplex containing dA.dU base pairs with high affinity, despite the large structural differences from its regular binding site in APS DNA.