Specific DNA sequences associated with the nuclear matrix in synchronized mouse 3T3 cells.

Eukaryotic chromatin appears to be organized into arrays of supercoiled loops anchored to the scaffolding structure of the mitotic chromosome core or to the nuclear matrix of interphase nuclei. To reveal whether specific DNA sequences are involved in this level of chromatin organization, we isolated and cloned a population of DNA molecules [average length of 150 base pairs (bp)] closely associated with the nuclear matrix after exhaustive DNase digestion and subsequent extensive protease digestion. The nuclear matrix was obtained from murine BALB/c 3T3 cells synchronized at the G1/S border of the cell cycle. We report the structure of two sequences, designated G4 and G5, which are highly enriched in the matrix DNA. Sequence G4, of 152 bp, contains three 31-bp direct head-to-tail repeats. An 11-bp sequence at the end of each repeat is homologous to the first large tumor antigen recognition site of human papova virus. Sequence G5, of 135 bp, consists of two well-defined domains, in which the first domain is a fragment of the B1 repetitive sequence. The results suggest the possibility that the loops of histone-depleted chromatin are connected to the scaffold of the nuclear matrix, with specific DNA sequences at the anchorage sites.     

Directly repeated sequences associated with pathogenic mitochondrial DNA deletions.

We determined the nucleotide sequences of junctional regions associated with large deletions of mitochondrial DNA found in four unrelated individuals with a phenotype of chronic progressive external ophthalmoplegia. In each patient, the deletion breakpoint occurred within a directly repeated sequence of 13-18 base pairs, present in different regions of the normal mitochondrial genome-separated by 4.5-7.7 kilobases. In two patients, the deletions were identical. When all four repeated sequences are compared, a consensus sequence of 11 nucleotides emerges, similar to putative recombination signals, suggesting the involvement of a recombinational event. Partially deleted and normal mitochondrial DNAs were found in all tissues examined, but in very different proportions, indicating that these mutations originated before the primary cell layers diverged.     

Sequence-dependent energetics of the B-Z transition in supercoiled DNA containing nonalternating purine-pyrimidine sequences.

The likelihood that a given DNA sequence will adopt the Z conformation in negatively supercoiled DNA depends on the energy difference between the B form and the Z form for that sequence relative to other sequences in the same molecule. This energy can be viewed simply as a sum of energies for the nearest-neighbor interactions within the sequence plus the energy required to stabilize the B-Z boundaries. Knowledge of these energetic terms would be of value in predicting when sequences become left-handed in response to negative superhelicity. Here we present an approach that can be used to determine the free-energy changes associated with all the nearest-neighbor interactions that can occur in Z-DNA. Synthetic stretches of d(C-G)n containing one or two transversions were cloned into plasmids, and the extent of the B-Z transition as a function of negative superhelicity was determined for each insert by two-dimensional agarose gel electrophoresis. By subjecting the data to statistical mechanical analysis, it was possible to evaluate the energetic penalty resulting from each base-pair (bp) substitution. Guanine to cytosine transversions cost 2.4 kcal (1 cal = 4.18 J)/(mol X bp), whereas guanine to thymine transversions cost 3.4 kcal/(mol X bp), to stabilize in the Z conformation. We have used these numbers, along with energetic values determined by others for the B-Z transition, to predict that certain strictly nonalternating purine and pyrimidine sequences may adopt the Z form readily.     

Hybridization of mouse leukemia virus c-DNA to mouse repeated DNA sequences.

Experiments of hybridization between mouse leukemia virus synthetic 3H-DNA probe and mouse main band and satellite DNAs indicate that there is not a higher concentration of viral sequences in the satellite DNA. On the contrary, viral sequences appear to be enriched in the fast renaturing intermediate main band DNA.     

Nonrandom binding of the carcinogen N-hydroxy-2-acetylaminofluorene to repetitive sequences of rat liver DNA in vivo.

We have examined the distribution of individual adducts in repetitive DNA sequences of rat liver in vivo after a single dose of the carcinogen N-hydroxy-2-acetyl-aminofluorene. Repetitive fragments [82, 125, 179, 225, and 370 base pairs (bp)] were isolated by digestion of hepatic DNA with HindIII restriction endonuclease (EC 3.1.23.21) and gel electrophoresis. As assayed by 32P postlabeling, no qualitative differences were observed between the DNA-bound metabolites in repetitive sequences and total DNA, but preferential binding to these sequences occurred. After 1 day of treatment, the amounts of N-hydroxy-2-acetylaminofluorene-induced adducts were found to be 13.8, 2.0, and 3.0 times higher in 179-, 225-, and 370-bp repeats, respectively, than in total DNA, while 82- and 125-bp repeats showed no differences. The relative distribution of individual adducts varied among the various sequences. After 9 days, all five sequences showed 1.3-1.7 times higher binding as compared to total DNA. In contrast, a random binding was observed when DNA reacted in vitro with an active metabolite, N-acetoxy-2-acetylaminofluorene. Taken together, these results suggest that the enrichment and differential excision of adducts in the repetitive DNA sequences may be a function of the nuclear organization of DNA. This application of the 32P assay constitutes a means to study the DNA damage and excision repair in vivo in chromatin structural components, including transcribed and nontranscribed multiple-copy genes, in a much more sensitive and precise way than has been hitherto possible.     

Magnesium ion-dependent triple-helix structure formed by homopurine-homopyrimidine sequences in supercoiled plasmid DNA.

DNA can be chemically cleaved at the site of chloroacetaldehyde-modified residues by the chemicals used for Maxam-Gilbert sequencing reactions. Use of this technique facilitates fine structural analysis of unpaired DNA bases in DNA with non-B-DNA structure. This method was used to study the non-B-DNA structure adopted by the poly-(dG).poly(dC) sequence under torsional stress at various ionic conditions. In the presence of 2 mM Mg2+, the 5' half of the deoxycytosine tract is very reactive to chloroacetaldehyde, while the 3' half is virtually unreactive. In the poly(dG) tract, chloroacetaldehyde reaction is restricted to the center guanine residues. In the absence of Mg2+, however, it is the 5' half of the deoxyguanine tract that is reactive to chloroacetaldehyde, while the 3' half is unreactive. And chloroacetaldehyde reaction is restricted to the center cytosine residues in the poly(dC) stretch. These results strongly suggest that the poly(dG).poly(dC) sequence is folded into halves from the center of the sequence to form a tetra-stranded-like structure. Such a structure contains either a triplex consisting of poly(dG).poly(dG).poly(dC) strands in the presence of Mg2+ or a triplex consisting of poly(dC).poly(dG).poly(dC) strands in the absence of Mg2+. The fourth strand, not involved in triplex formation, is closely associated with the triplex and is positioned in such a way that DNA bases are exposed and freely accessible to the chloroacetaldehyde reaction.     

mRNA in human cells contains sequences complementary to the Alu family of repeated DNA.

Approximately one-half of the polysomal poly(A)+RNA from CCRF-CEM human lymphoblastoid cells associates at low R0t (10 M.sec) [where R0 is the initial concentration of RNA (M) and t is time (sec)] to form branched complexes detectable by electron microscopy. The complexes typically involve 2-16 molecules associated over double-stranded regions 120 +/- 30 base pairs long. Formation of such complexes suggests that poly(A)+RNA contains repeated-sequence elements that are highly represented in the mRNA population. Hybridization of polysomal poly(A)+RNA with a recombinant human DNA plasmid, p lambda H15C, which is shown to contain at least three regions complementary to two different members of the Alu family of DNA repeat sequences, showed a total of five regions where R loops are formed. The hybridized regions comprising these groups are 260 +/- 180, 240 +/- 170, 150 +/- 70, 180 +/- 60, and 180 +/- 80 base pairs long. The relative frequencies of R loops formed at these different sites indicate that sequences in this recombinant DNA are represented in the mRNA population at different frequencies. The hybridizing sequence of the RNA molecules is located near one terminus in 13% of the R loops and internally in 53% of the R loops. Surprisingly, 35% of the R loops apparently involve RNA molecules hybridized over their entire length of only 200 +/- 110 base pairs.     

Dynamics of the B-to-Z transition in supercoiled DNA.

The sequence (dC-dG)16, inserted into the polylinker of plasmid pUC8, adopts a left-handed Z-DNA conformation at "natural" supercoil density. The radioactively labeled monoclonal antibody Z-D11, which has a very high affinity for this DNA conformation, provides a convenient sensitive tool to measure selectively the amount of Z-DNA. Chloroquine reversibly changes the supercoil density of plasmid DNA and thereby the equilibrium between right- and left-handed double-helical DNA. The time-dependent formation or disappearance of Z-DNA was measured by using the antibody either as a fast indicator of Z-DNA or as an additional effector of the B-to-Z equilibrium. In the middle of the transition, a relaxation time of about 1 hr is observed in 0.1 M NaCl at 22 degrees C. The kinetic data are compatible with an all-or-none transition between the two conformations. The overall rate constant for Z-DNA formation, kBZ, decreases with the square of the chloroquine concentration, while the reverse one, kZB, increases with about the fourth power.     

Mapping of inverted repeated DNA sequences within the genome of simian virus 40.

Single-stranded, linear DNA of simian virus 40 (SV40) created by denaturing the endonuclease EcoRI- or Hpa II-generated, linear, double-stranded products from form I DNA of SV40 was analyzed for regions of inverted repeated sequences by visualization with the electron microscope. Six hairpin loops were found at positions 0.11-0.30 (two loops forming a "rabbit ears" structure), 0.47-0.52, 0.63-0.68, 0.70-0.76, and 0.90-0.96. The nucleotide sequences within all of these inverted repeats may be related since the looped regions can crosshybridize with one another and, thus, the SV40 genome may contain regions of interspersed repeated and unique sequences. The map positions of the 3' and 5' ends of the early and late messenger RNAs, as determined by others, lie within regions of inverted repeated sequences. Previously recorded recombination events that occurred either within the SV40 genome or between SV40 DNA and other genomes have apparently occurred frequently at positions of inverted repeated sequences within the SV40 DNA.     

Plasmid P1 replication: negative control by repeated DNA sequences.

The incompatibility locus, incA, of the unit-copy plasmid P1 is contained within a fragment that is essentially a set of nine 19-base-pair repeats. One or more copies of the fragment destabilizes the plasmid when present in trans. Here we show that extra copies of incA interfere with plasmid DNA replication and that a deletion of most of incA increases plasmid copy number. Thus, incA is not essential for replication but is required for its control. When cloned in a high-copy-number vector, pieces of the incA fragment that each contain only three repeats destabilize P1 plasmids efficiently. This result makes it unlikely that incA specifies a regulatory product. Our in vivo results suggest that the repeating DNA sequence itself negatively controls replication by titrating a P1-determined protein, RepA, that is essential for replication. Consistent with this hypothesis is the observation that the RepA protein binds to the incA fragment in vitro.     

Tandemly repeated nonribosomal DNA sequences in the chloroplast genome of an Acetabularia mediterranea strain.

A purified chloroplast fraction was prepared from caps of the giant unicellular green alga Acetabularia mediterranea (strain 17). High molecular weight DNA obtained from these chloroplasts contains at least five copies of a 10-kilobase-pair (kbp) sequence tandemly arranged. This unique sequence is present in DNA from chloroplasts of all stages of the life cycle examined. A chloroplast rDNA clone from mustard hybridized with some restriction fragments from Acetabularia chloroplast DNA but not with the repeated sequence. An 8-kbp EcoRI-Pst I fragment of the repeated sequence was cloned into pBR322 and used as a hybridization probe. No homology was found between the cloned 8-kbp sequence and chloroplast DNA from related species Acetabularia crenulata or chloroplast DNA from spinach.     

Nuclear matrix DNA from chicken erythrocytes contains beta-globin gene sequences.

Nuclear matrices containing residual DNA were isolated from chicken erythrocytes after extraction of purified nuclei with buffered 2 M NaCl. After further purification of this residual DNA, it was found to contain high concentrations of beta-globin gene sequences as assayed by dot hybridization with 32P-labeled nick-translated pHB1001. Electron microscopy of a random sample of this residual DNA fraction shows the DNA to be intimately associated with protein at various intervals. A hypothesis for enrichment of active genes in residual DNA from purified chromatin or in nuclear matrix DNA is also discussed.     

Virus-specific sequences in nuclear DNA from non-producer hamster Rous sarcoma.

Proviral DNA from non-producer Rous sarcoma in Syrian hamster contains practically all the nucleotide sequences presented in 125I-labeled RNA from Rous sarcoma virus, Carr-Zilber strain. Virus-specific sequences consist of moderately reiterated and unique DNA regions. The amount of Rous sarcoma virus-specific provirus equivalents in hamster Rous sarcoma DNA is equal to 5.2 +/- 0.01. Experiments on transfection show that proviral DNA studied possesses biological activity in respect to cell transformation and virus production. The infectivity of DNA from hamster tumor does not depend on the expression of group-specific (gs) antigen in the recipient cells.     

Error-prone replication of repeated DNA sequences by T7 DNA polymerase in the absence of its processivity subunit.

We have examined the effect of thioredoxin, an accessory protein that confers high processivity to bacteriophage T7 DNA polymerase, on the fidelity of DNA synthesis. In the presence of thioredoxin, exonuclease-proficient T7 DNA polymerase is highly accurate. In fidelity assays that score errors that revert M13mp2 lacZ alpha-complementation mutants, error rates are < or = 2.2 x 10(-6) for base substitution and < or = 3.7 x 10(-7) and < or = 4.5 x 10(-7) for frameshifts that revert mutations in the +1 and -1 reading frames, respectively. Rates are more than 10-fold higher during synthesis by polymerase.thioredoxin complex lacking 3'-->5' exonuclease activity, demonstrating that frameshift as well as substitution errors are subject to proofreading. The contribution of thioredoxin to accuracy has been examined by comparing the fidelity of the exonuclease-deficient polymerase in the presence or absence of the accessory protein. Thioredoxin either enhances or reduces fidelity, depending on the type of error considered. In the absence of thioredoxin, T7 DNA polymerase is 3-fold more accurate for base substitutions and > or = 27-fold and 9-fold more accurate, respectively, for 1- and 2-nt deletion errors at nonreiterated nucleotide sequences. Higher fidelity for all three errors may reflect the inability of the polymerase to continue synthesis from the premutational intermediates in the absence of the accessory protein. In marked contrast, the rate for frameshift errors wherein one or more nucleotides has been added to a repeated DNA sequence increases 46-fold when thioredoxin is absent from the polymerization reaction. The error rate increases as the length of the repeated sequence increases, consistent with a model where strand slippage creates misaligned template-primers. Thus, replicative expansion of repetitive sequences occurs in the absence of a replication accessory protein.     

Lupus-inducing drugs alter the structure of supercoiled circular DNA domains.

We analyzed the effects of procainamide (PROC), hydralazine (HYD), N-acetylprocainamide (NAPA), and L-canavanine (CAN) on circular supercoiled plasmids as models for chromosomal loop domains. The supercoil-dependent B-Z equilibrium in recombinant plasmids was used as an indicator of structural changes induced in circular DNA. Two-dimensional gel electrophoresis showed that PROC and HYD strongly inhibited supercoil-induced Z-DNA formation, whereas NAPA caused less pronounced changes in the B-Z equilibrium, and CAN had no effect. Gel retardation assays showed that the binding of a Z-DNA-specific autoimmune antibody to a Z-DNA-containing plasmid was strongly perturbed by HYD, but not influenced by CAN. Both PROC and NAPA showed moderate inhibition of antibody binding. Our results demonstrate the different potentials of these 4 drugs to interact with DNA and to alter the tertiary topology of DNA domains. It is conceivable that the in vivo capacity of PROC and HYD to induce antinuclear antibodies may be related to their ability to influence structural features in chromosomal DNA domains or nucleosomes, thus liberating antigenic structural epitopes in DNA and/or DNA-associated proteins.     

Imaging of DNA sequences with chemiluminescence.

We have coupled a chemiluminescent detection method that uses an alkaline phosphatase label to the genomic DNA sequencing protocol of Church and Gilbert [Church, G. M. & Gilbert, W. (1984) Proc. Natl. Acad. Sci. USA 81, 1991-1995]. Images of sequence ladders are obtained on x-ray film with exposure times of less than 30 min, as compared to 40 h required for a similar exposure with a 32P-labeled oligomer. Chemically cleaved DNA from a sequencing gel is transferred to a nylon membrane, and specific sequence ladders are selected by hybridization to DNA oligonucleotides labeled with alkaline phosphatase or with biotin, leading directly or indirectly to deposition of enzyme. If a biotinylated probe is used, an incubation with avidin-alkaline phosphatase conjugate follows. The membrane is soaked in the chemiluminescent substrate (AMPPD) and is exposed to film. Dephosphorylation of AMPPD leads in a two-step pathway to a highly localized emission of visible light. The demonstrated shorter exposure times may improve the efficiency of a serial reprobing strategy such as the multiplex sequencing approach of Church and Kieffer-Higgins [Church, G. M. & Kieffer-Higgins, S. (1988) Science 240, 185-188].