The solution structure of the hairpin formed by d(TCTCTC-TTT-GAGAGA).

The 15-residue oligonucleotide d(TCTCTC-TTT-GAGAGA) forms a hairpin structure with a loop of three thymidine residues at neutral pH or above. The three-dimensional solution structure of this oligonucleotide has been determined by means of two-dimensional nuclear magnetic resonance methods. Interproton distance constraints derived from NOEs, in combination with torsion angle constraints obtained from J-coupling constants were used in the variable target function program DIANA to derive the hairpin structure. It was found that hairpins with two different loop conformations fit the NMR data, i.e. an equilibrium between these two conformational states can only fully explain the NOE data available. In one state, loop residue T7 is turned into the minor groove, while in the second state residue T8 is in the minor groove. In both conformations the phosphate backbone changes its direction by 180 degrees between residues T9 and G10. Concomitantly, torsion angles zeta of T9 and alpha of G10 both adopt a gauche(+) conformation and gamma of residue G10 adopts a trans conformation to induce this complete change in the direction of the backbone.

Yeast plasmids have fewer superhelical turns than predicted by the nucleosome repeat of yeast DNA.

The superhelical density of three Saccharomyces cerevisiae plasmids was determined with respect to a defined reference state during vegetative growth and stationary phase. The levels of supercoiling determined were approximately 20% lower than predicted by comparisons with SV40 DNA and reconstituted minichromosomes using histones from higher eukaryotes. In two different plasmids with the ARS1 origin of replication, the level of supercoiling changed substantially as the host cells entered stationary phase. Supercoiling of the endogenous 2-microns plasmid during vegetative growth was lower than in the ARS1-containing plasmids but did not change significantly upon entry of the cells into stationary phase.

Structural characterization of separated H DNA conformers.

Polypyrimidine/polypurine DNA sequences in plasmids can adopt protonated triplex-containing structures (H DNA) in response to negative superhelical stress and low pH. A d(TC)17-d(GA)17 insert adopts two isomeric protonated structures, which differ in degree of helical unwinding. The variant forms of individual topoisomers were separated by agarose gel electrophoresis and their reactivities to permanganate and acid-induced depurination were compared. Depurination patterns of the individual conformers indicate that in the more mobile form (H-y5) the 5'-half of the d(GA)n strand participates in a triplex while in the other (H-y3) the 3'-half forms the triplex. The H-y5 form is more stable than the H-y3 form at low negative superhelix densities. Because of the difference in helical unwinding, the H-y5 form becomes relatively less stable as the superhelix density increases. Topological models of the two forms show that providing there is no linkage at the tips of the triple helical segments one more positive twist is localized in the H-y5 form than in the H-y3 form. The foldback in the pyrimidine strand of the H-y5 form is less accessible to solvent than that of the H-y3 form as assessed by its lower reactivity to permanganate. Consideration of a pyrimidine loop model (Harvey, S. C., Luo, J., & Lavery, R. (1988) Nucleic Acids Res. 16, 11795-11809) suggests that the unique stability of the H-y5 form results from Watson-Crick base pairs between residues of the d(TC)n loop and the d(GA)n strand as it exits the triplex.

Polypurine/polypyrimidine hairpins form a triple helix structure at low pH.

1D and 2D NMR investigations of the 15 residue deoxynucleotide sequence d(TCTCTC-TTT-GAGAGA) show that above pH = 6.5 the molecule adopts a B-form hairpin conformation. As the pH is lowered below 6.5 molecules progressively associate in pairs to form a partially triple helical, partially single stranded structure in which the bases of the oligopyrimidine d(TC)3 tract from one molecule form Hoogsteen pairs with the d(G-A)3 tract of the other. Imino protons of protonated cytosines can be observed at very low field (approximately 15 ppm). The enthalpy of triplex formation was estimated by NMR techniques to be -16 kcal mol-1. Intense H6 to H3' cross peaks from residues in all three strands suggest the presence of N-type sugars at some but not at all possible sites. Surprisingly strong cross peaks between H5' or H5" and non-exchangeable base protons are also observed. These suggest that certain of the O5'-C5'-C4'-C3' phosphate backbone torsion angles (gamma) are unusual.

Protonated polypurine/polypyrimidine DNA tracts that appear to lack the single-stranded pyrimidine loop predicted by the "H" model.

Three synthetic oligomers: 5'd(AG)8.dA.d(CT)(8)3'(A), 5'd(TC)7.d(TTA).d(GA)(8)3'(B) and d(GA)17(C) were cloned into the plasmid vector p915 in order to study the effects of sequence symmetry on pH-dependent structural transitions in polypurine/polypyrimidine DNA. When present in linear molecules all three sequences undergo transitions to protonated states. These are kinked to different degrees as determined by a non-denaturing gel mobility shift assay. Chemical probe analysis shows that the protonated states adopted by the linear forms of A and C exhibit certain features which have been regarded as indicating partially triple stranded "H" transition structures. The chemical reactivities of the transition structure adopted by linear molecule B and certain features of those exhibited by the transition structures of linear molecules A and C do not conform to the predictions of the "H" model.

H3 Cys-110 is in close proximity to the C-terminal regions of H2B and H4 in a nucleosome core with an altered internal arrangement of histones.

A particle obtained by nuclease digestion of nucleohistone complexes prepared by direct mixing of histones with DNA in 0.15 M NaCl was indistinguishable by composition and physical properties from nucleosome cores prepared under the same conditions from nucleohistone preannealed in 0.6 M NaCl. We show here that different photo-cross-links form when these particles are prepared from H3 labeled with photoaffinity reagents on the unique histone H3 cysteine. H3-H3 histone dimers were dominant when the particles were prepared by dilution of the nucleohistone from 0.6 M NaCl while H3-H2B and H3-H4 histone dimers were prominent if the nucleohistone complex was prepared directly in 0.15 M NaCl. Peptide mapping of the novel H3-H4 and H3-H2B dimers showed that Cys-110 of histone H3 is cross-linked to the 18 amino acid C-terminal end of H4 or to the 66 amino acid C-terminal half of H2B.

Triplex DNA in plasmids and chromosomes.

Circular plasmids containing pyrimidine purine tracts can form both inter-and intramolecular triplexes. Addition of poly(dTC) to plasmid pTC45, which contains a (TC)45.(GA)45 insert, results in intermolecular triplex formation. Agarose-gel electrophoresis gives rise to many well-resolved bands, which correspond to 1, 2, 3, 4... plasmid molecules attached to the added pyrimidine strand. In the electron microscope these complexes appear as a rosette of petals. The mobility of these triplex-containing complexes can be retarded by the addition of a triplex-specific monoclonal antibody, Jel318. Intramolecular triplex formation can be demonstrated at pH 5 in pTC45 and also in pT463-I, a plasmid containing a segment of a crab satellite DNA with both (G)n.(C)n and (TCC)n.(GGA)n inserts. However, although the intermolecular triplex remains stable for some time at pH 8, intramolecular triplex formation only occurs at low pH. Triplexes can also be detected by an immunoblotting procedure with Jel318. This unfamiliar structure is readily demonstrated in eukaryotic extracts, but not in cell extracts from Escherichia coli. Triplexes may thus be an inherent feature of eukaryotic chromosome structure.

Intermediate range effects in DNA. I: Low pH/stress induced conformational changes in the vicinity of an extruded d(AT)n.d(AT) in cruciform.

A family of plasmids which contain d(AT)n cruciforms are sensitive to "single strand specific" (SS) endonucleases and a variety of chemical probes in a "random sequence" region centered 10-30 residues away from the cruciform junction. The SS nuclease sensitive structures are dependent on the presence of the extruded cruciform and exhibit a degree of sequence independence. Their appearance depends upon the combined effects of slightly lower than neutral pH and superhelical coiling above the minimum required to drive the extrusion of the d(AT)n cruciform arms. The nuclease sensitive structure is therefore underwound with respect to the B conformation and contains protonated bases.

Bovine elastin cDNA clones: evidence for the occurrence of a new elastin-related protein in fetal calf ligamentum nuchae.

Poly (A+) mRNA was isolated from fetal calf ligamentum nuchae and used for the construction of cDNA libraries. A fraction highly enriched in elastin mRNA was used to prepare the cDNA probes for screening the libraries. A 2 kb clone, pRE1, gave the most positive signal in colony hybridization. It hybridized to a mRNA of the same size as reported for elastin mRNAs from chick and sheep. Hybrid-arrested translation showed that translation of mRNAs for proteins other than elastin doublet was not inhibited by pRE1. Southern blot analysis showed that pRE1 has sequence homology with pVE6 and pVE10, which were tentatively identified as elastin-related cDNA clones representing two distinct mRNAs. DNA sequence data from the 5' end of pRE1 show that the translated amino acid sequence is not typical of known elastin sequences but contains some elastin-like sequences. All of this evidence strongly suggests the occurrence in fetal calf nuchal ligament of a mRNA which codes for a previously unknown elastin-related protein.

A structural basis for S1 nuclease sensitivity of double-stranded DNA.

A protonated form of a cloned simple repeating DNA sequence d(TC)n X d(GA) is detectable in equilibrium with the usual Watson-Crick base-paired form at pHs up to 7. This form is anomalously sensitive to a variety of single-strand-specific endonucleases. The observed pH dependent protection of N-7 of dG residues within the insert suggests that these residues are either Hoogsteen or reverse Hoogsteen base-paired to protonated dC residues of the polypyrimidine strand. A structure in which dA:dT Watson-Crick base pairs alternate with Hoogsteen syndG:dCH+ pairs appears to be the most stereochemically acceptable structure consistent with the chemical properties of this protonated DNA. Protonated d(TC)n X d(GA)n interacts with an anti-Z DNA antibody raised against brominated d(GC)n X d(GC)n.

Transition of a cloned d(AT)n-d(AT)n tract to a cruciform in vivo.

A 34 base pair tract of the simple repeating dinucleotide d(AT)n-d(AT)n cloned into a 2.4 kb polylinker plasmid vector undergoes a structural transition in response to negative superhelical coiling. The transition has been characterized by 2 dimensional gel electrophoresis, mapping of S1, P1 and T7 endonuclease 1 sensitive sites, and mapping of sites that are sensitive to modification by bromoacetaldehyde. After S1 nuclease treatment it is possible to trap supercoiled species that are nicked on one or both strands near the center of the palindrome. These data show that the alternate state adopted by the d(AT)n-dAT)n insert is a cruciform rather than a Z conformation. Unlike other B-cruciform transitions the transition in d(AT)n-d(AT)n has a low activation energy and the transition is facilitated by the presence of magnesium ions. Evidence from in-vivo topoisomer distributions is presented which shows that under conditions of blocked protein synthesis the d(AT)n-d(AT)n insert will spontaneously adopt the cruciform state in-vivo in E. coli.

The in-vivo occurrence of Z DNA.

The energetics of the B-Z transition of two different types of cloned alternating purine/pyrimidine DNA sequences have been analysed by a two dimensional electrophoretic technique. Since the transition between right handed and left handed forms of these polymers is detected by alterations of electrophoretic mobilities of topoisomers of the plasmid DNA molecules, the method is not dependent on Z-DNA binding ligands. The measurements reflect intrinsic properties of the DNA unperturbed by the free energy of binding such a ligand. Direct evidence from the analysis of topoisomer distributions is presented which shows that d(GC)n.d(GC)n sequence elements within an E. coli plasmid will adopt a Z conformation in-vivo under conditions of blocked protein synthesis. Evidence for the in-vivo occurrence of Z-DNA was not detected in plasmid DNA isolated from bacterial cells growing in the absence of protein synthesis inhibitors. A model is proposed for a function for the B-Z transition in ensuring the correct pairing of homologous chromosomes during meiosis.

Internal structure of discrete nucleohistone complexes which form in vitro under conditions of physiological ionic strength.

Three discrete histone-DNA complexes assemble spontaneously when the four core histones are mixed with DNA under conditions which are close to physiological (0.15 M NaCl, pH 8). These species include the (H2A,H2B) dimeric complex (P1), the (H2A2,H2B2,H3,H4) hexameric complex (P2) and the nucleosome core complex (P3). This report compares several properties of these complexes with the properties of nucleosome cores assembled at high ionic strength (0.6 M NaCl). Based on histone-histone cross-linking studies, CD spectra, and thermal denaturation experiments, P1 is structurally similar to the subnucleosomal (H2A,H2B) fragment isolated from nuclease-digested chromatin. P1, P2, P3, and high salt-assembled nucleosome cores can all incorporate (H2A,H2B) pairs which have been previously cross-linked. Although the CD spectra and thermal denaturation profiles of P2 and P3 are closely related to those of nucleosome cores assembled in 0.6 M NaCl, cross-linking studies indicate that the arrangement of the histones in P2, and in a proportion of the P3 particles assembled in 0.15 M NaCl, are significantly different from their arrangement in nucleosome cores assembled in 0.6 M NaCl. The single cysteine residue on the H3 of P2 is accessible to the solvent. The two fluorescently labeled cysteine residues in a large proportion of the P3 particles assembled in 0.15 M NaCl are in a different orientation with respect to each other than the same residues in nucleosome cores assembled at high ionic strength.

Pathways of assembly of nucleohistone complexes formed in vitro under physiological conditions. Implications for the structure of the nucleosome.

The composition and structure of three discrete nucleohistone particles produced by micrococcal nuclease digestion of the complex formed when the four core histones are mixed directly with DNA under conditions that are close to physiological have been reported in accompanying articles (Ellison, M. J., and Pulleyblank, D. E. (1983) J. Biol. Chem. 258, 13307-13313; 13314-13320). These include a peak containing a compact hexameric complex composed of one pair of (H3,H4) and two pairs of (H2A,H2B) (P2) and a peak containing complexes with the composition and sedimentation properties of nucleosome cores (P3). We have examined the pathways of assembly of P2 and P3 by determining the effect of order of histone addition on the yields of these species. The assembly of P2 is initiated by the binding of an (H2A,H2B) pair to the DNA which then directs the placement of an (H3,H4) pair. The incorporation of the final (H2A,H2B) pair to complete the particle is probably cooperative. The assembly of P3 is less dependent than the assembly of P2 upon the order in which the histone pairs are added to the DNA. More than one pathway has been implicated in the formation of these particles. P3 contains a component which has many but not all of the properties of nucleosome cores assembled at elevated ionic strength. Alternative pathways of assembly of histone pairs onto DNA are discussed which could lead to the formation of the discrete histone DNA structures that have been isolated here. A model is proposed in which the nucleosome unfolds and refolds into an alternative configuration.

The assembly of an H2A2,H2B2,H3,H4 hexamer onto DNA under conditions of physiological ionic strength.

A novel nucleohistone particle is generated in high yield when a complex of DNA with the four core histones formed under conditions that are close to physiological (0.15 M NaCl, pH 8) is treated with micrococcal nuclease. The particle was found to contain 102 base pairs of DNA in association with six molecules of histones in the ratio 2H2A:2H2B:1H3:1H4 after relatively brief nuclease treatment. Prolonged nuclease digestion resulted in a reduction in the DNA length to a sharply defined 92-base pair fragment that was resistant to further degradation. Apparently normal nucleosome core particles containing two molecules each of the four core histones in association with 145 base pairs of DNA and a particle containing one molecule each of histones H2A and H2B in association with approximately 40 base pairs of DNA were also generated during nuclease treatment of the histone-DNA complexes formed under physiological ionic strength conditions. Kinetic studies have shown that the hexamer particle is not a subnucleosomal fragment produced by the degradation of nucleosome core particles. Furthermore, the hexamer particle was not found among the products of nuclease digestion when histones and DNA were previously assembled in 0.6 M NaCl. The high sedimentation coefficient of the hexameric complex (8 S) suggests that the DNA component of the particle has a folded conformation.

Facile transition of poly[d(TG) x d(CA)] into a left-handed helix in physiological conditions.

The DNA polymer d(GC)n . d(GC)n can undergo a transition from the usual right-handed 10.4 base pairs (bp) per turn B form to a novel left-handed 12 bp per turn Z form in response to altered environmental conditions. Several other alternating purine-pyrimidine DNA polymers with modified bases have been shown to undergo transitions from B to Z conformations, with varying degrees of difficulty. We report here that the unmodified DNA polymer d(TG)n . d(CA)n readily undergoes a transition to a Z conformation when subjected to unwinding torsional stress in ionic conditions that are close to physiological. By using a two-dimensional gel electrophoresis system, we have determined both the critical free energy of supercoiling that is required to initiate the transition and the free energy of supercoiling that is required to maintain this polymer in the Z form.

A topoisomerase from chicken erythrocyte nuclei which does not assemble nucleosome core particles in vitro.

We have examined the ability of a topoisomerase purified from chicken erythrocyte nuclei to mediate nucleosome core assembly in vitro at physiological ionic strength (0.15 M NaCl). Although we have detected limited amounts of spontaneously assembled nucleosome cores at this salt concentration, the addition of this topoisomerase does not increase the amount of assembly observed. Nucleosome assembly was assayed by quantitating the amount of core particle length DNA accumulated with time upon the nuclease digestion of histone-DNA complexes. In addition, the amount of negative supercoils introduced into relaxed closed circular DNA upon nucleosome core particle assembly was determined. Correctly assembled complexes do not protect more DNA from nuclease digestion than random histone-DNA complexes but shift the heterogeneous size distribution of protected fragments to a more homogeneous distribution centred around 145 base pairs. Under our conditions of nucleosome assembly, a second histone-DNA complex which is distinct from the nucleosome core can be detected under physiological ionic strength conditions. This particle does not form in high salt assembly experiments. Similarly, the assembly of this particle is unaffected by the presence or absence of topoisomerase.

Purification and properties of type 1 topoisomerase from chicken erythrocytes: mechanism of eukaryotic topoisomerase action.

A simple method for the purification of the major topoisomerase (topoisomerase 1) from chicken erythrocytes is described. Because of the generally repressed state of the chromatin from these nuclei, the heterogeneity of the non-histone proteins is reduced, and it is possible to purify this enzyme from a nuclear extract by a single chromatographic step. The chicken erythrocyte topoisomerase appears to be similar to previously described eukaryotic type I topoisomerases with respect to its physical and enzymological properties. The pattern of intermediate products generated during the action of chicken erythrocyte topoisomerase on a supercoiled closed circular DNA substrate has been examined quantitatively and has been shown to be consistent with a mechanism in which the enzyme closes its substrate DNA molecular after the removal of each superhelical turn and in which dissociation of the enzyme substrate complex may, but does not necessarily, occur after each cycle of the reaction.