Electron microscopy of DNA crosslinked with trimethylpsoralen: test of the secondary structure of eukaryotic inverted repeat sequences

Twenty years of t-loops: A case study for the importance of collaboration in molecular biology

Collaborative studies open doors to breakthroughs otherwise unattainable by any one laboratory alone. Here we describe the initial collaboration between the Griffith and de Lange laboratories that led to thinking about the telomere as a DNA template for homologous recombination, the proposal of telomere looping, and the first electron micrographs of t-loops. This was followed by collaborations that revealed t-loops across eukaryotic phyla. The Griffith and Tomáška/Nosek collaboration revealed circular telomeric DNA (t-circles) derived from the linear mitochondrial chromosomes of nonconventional yeast, which spurred discovery of t-circles in ALT-positive human cells. Collaborative work between the Griffith and McEachern labs demonstrated t-loops and t-circles in a series of yeast species. The de Lange and Zhuang laboratories then applied super-resolution light microscopy to demonstrate a genetic role for TRF2 in loop formation. Recent work from the Griffith laboratory linked telomere transcription with t-loop formation, providing a new model of the t-loop junction. A recent collaboration between the Cesare and Gaus laboratories utilized super-resolution light microscopy to provide details about t-loops as protective elements, followed by the Boulton and Cesare laboratories showing how cell cycle regulation of TRF2 and RTEL enables t-loop opening and reformation to promote telomere replication. Twenty years after the discovery of t-loops, we reflect on the collective history of their research as a case study in collaborative molecular biology.

The RNA Base-Pairing Problem and Base-Pairing Solutions

RNA molecules are folded into structures and complexes to perform a wide variety of functions. Determination of RNA structures and their interactions is a fundamental problem in RNA biology. Most RNA molecules in living cells are large and dynamic, posing unique challenges to structure analysis. Here we review progress in RNA structure analysis, focusing on methods that use the "cross-link, proximally ligate, and sequence" principle for high-throughput detection of base-pairing interactions in living cells. Beginning with a comparison of commonly used methods in structure determination and a brief historical account of psoralen cross-linking studies, we highlight the important features of cross-linking methods and new biological insights into RNA structures and interactions from recent studies. Further improvement of these cross-linking methods and application to previously intractable problems will shed new light on the mechanisms of the "modern RNA world."

Transcription of telomeric DNA leads to high levels of homologous recombination and t-loops

The formation of DNA loops at chromosome ends (t-loops) and the transcription of telomeres producing G-rich RNA (TERRA) represent two central features of telomeres. To explore a possible link between them we employed artificial human telomeres containing long arrays of TTAGGG repeats flanked by the T7 or T3 promoters. Transcription of these DNAs generates a high frequency of t-loops within individual molecules and homologous recombination events between different DNAs at their telomeric sequences. T-loop formation does not require a single strand overhang, arguing that both terminal strands insert into the preceding duplex. The loops are very stable and some RNase H resistant TERRA remains at the t-loop, likely adding to their stability. Transcription of DNAs containing TTAGTG or TGAGTG repeats showed greatly reduced loop formation. While in the cell multiple pathways may lead to t-loop formation, the pathway revealed here does not depend on the shelterins but rather on the unique character of telomeric DNA when it is opened for transcription. Hence, telomeric sequences may have evolved to facilitate their ability to loop back on themselves.

RNA Duplex Map in Living Cells Reveals Higher-Order Transcriptome Structure

RNA has the intrinsic property to base pair, forming complex structures fundamental to its diverse functions. Here, we develop PARIS, a method based on reversible psoralen crosslinking for global mapping of RNA duplexes with near base-pair resolution in living cells. PARIS analysis in three human and mouse cell types reveals frequent long-range structures, higher-order architectures, and RNA-RNA interactions in trans across the transcriptome. PARIS determines base-pairing interactions on an individual-molecule level, revealing pervasive alternative conformations. We used PARIS-determined helices to guide phylogenetic analysis of RNA structures and discovered conserved long-range and alternative structures. XIST, a long noncoding RNA (lncRNA) essential for X chromosome inactivation, folds into evolutionarily conserved RNA structural domains that span many kilobases. XIST A-repeat forms complex inter-repeat duplexes that nucleate higher-order assembly of the key epigenetic silencing protein SPEN. PARIS is a generally applicable and versatile method that provides novel insights into the RNA structurome and interactome. VIDEO ABSTRACT.

Chromatin structure analysis of single gene molecules by psoralen cross-linking and electron microscopy

Nucleosomes occupy a central role in regulating eukaryotic gene expression by blocking access of transcription factors to their target sites on chromosomal DNA. Analysis of chromatin structure and function has mostly been performed by probing DNA accessibility with endonucleases. Such experiments average over large numbers of molecules of the same gene, and more recently, over entire genomes. However, both digestion and averaging erase the structural variation between molecules indicative of dynamic behavior, which must be reconstructed for any theory of regulation. Solution of this problem requires the structural analysis of single gene molecules. In this chapter, we describe a method by which single gene molecules are purified from the yeast Saccharomyces cerevisiae and cross-linked with psoralen, allowing the determination of nucleosome configurations by transmission electron microscopy. We also provide custom analysis software that semi-automates the analysis of micrograph data. This single-gene technique enables detailed examination of chromatin structure at any genomic locus in yeast.

Following the Chromosome Path to the Garden of the Genome

I have been fascinated by chromosomes for longer than I care to mention; their beautiful structure, cell-type-specific changes in morphology, and elegant movements delight me. Shortly before I began graduate study, the development of nucleic acid hybridization made it possible to compare two nucleic acids whether or not their sequences were known. From this stemmed a progression of development in tools and techniques that continues to enhance our understanding of how chromosomes function. As my PhD project I contributed to this progression by developing in situ hybridization, a technique for hybridization to nucleic acids within their cellular context. Early studies with this technique initiated several lines of research, two of which I describe here, that I have pursued to this day. First, analysis of RNA populations by hybridization to polytene chromosomes (a proto-microarray-type experiment) led us to characterize levels of regulation during heat shock beyond those recognizable by puffing studies. We found also that one still-undeciphered major heat shock puff encodes a novel set of RNAs for which we propose a regulatory role. Second, localization of various multicopy DNA sequences has suggested roles for them in chromosome structure: Most recently we have found that Drosophila telomeres consist of and are maintained by special non-LTR (long terminal repeat) retrotransposons.

Visualization of the annealing of complementary single-stranded DNA catalyzed by the herpes simplex virus type 1 ICP8 SSB/recombinase

The rate of annealing of long linear complementary single-stranded (ss) DNAs can be increased greatly by certain DNA-binding proteins including the herpes simplex virus type 1 ICP8 SSB/recombinase. Using electron microscopy, we have investigated the DNA-protein structures involved in ICP8-mediated DNA annealing. We show that the formation of superhelical ICP8-ssDNA filaments is required for annealing. Two superhelices interact with each other to form a coiled-coil, which is the intermediate in annealing. In this process, the superhelices likely rotate and translocate relative to each other. Psoralen/UV photocrosslinking studies revealed that meta-stable contacts form at sites of limited sequence homology during the annealing. Partial proteolysis of ICP8 in the protein-ssDNA complexes showed that Mg2+ induces conformational changes in the N-terminal region (amino acid residues 1-305) of ICP8. In addition to Mg2+, Ca2+ and, to a significantly lesser extent, Cu2+ and Mn2+, were found to induce superhelix formation of the ICP8-ssDNA filament and to facilitate annealing. Mechanisms for how the coiled-coil structures facilitate annealing are discussed.

Role of the double-strand origin cruciform in pT181 replication

pT181 is a small rolling-circle plasmid from Staphylococcus aureus whose initiator protein, RepC, melts the plasmid's double-strand origin (DSO) and extrudes a cruciform involving IR II, a palindrome flanking the initiation nick site. We have hypothesized that the cruciform is required for initiation, providing a single-stranded region for the assembly of the replisome (R. Jin et al., 1997, EMBO J. 16, 4456-4566). In this study, we have tested the requirement for cruciform extrusion by disrupting the symmetry of the IR II palindrome or by increasing its length. The modified DSOs were tested for replication with RepC in trans. Rather surprisingly, disruption of the IR II symmetry had no detectable effect on replication or on competitivity of the modified DSO, though plasmids with IR II disrupted were less efficiently relaxed than the wild type by RepC. However, in conjunction with IR II disruption, modification of the tight RepC binding site IR III blocked replication. These results define two key elements of the pT181 initiation mechanism--the IR II conformation and the RepC binding site (IR III)--and they indicate that pT181 replication initiation is sufficiently robust to be able to compensate for significant modifications in the configuration of the DSO.

Mammalian telomeres end in a large duplex loop

Mammalian telomeres contain a duplex array of telomeric repeats bound to the telomeric repeat-binding factors TRF1 and TRF2. Inhibition of TRF2 results in immediate deprotection of chromosome ends, manifested by loss of the telomeric 3' overhang, activation of p53, and end-to-end chromosome fusions. Electron microscopy reported here demonstrated that TRF2 can remodel linear telomeric DNA into large duplex loops (t loops) in vitro. Electron microscopy analysis of psoralen cross-linked telomeric DNA purified from human and mouse cells revealed abundant large t loops with a size distribution consistent with their telomeric origin. Binding of TRF1 and single strand binding protein suggested that t loops are formed by invasion of the 3' telomeric overhang into the duplex telomeric repeat array. T loops may provide a general mechanism for the protection and replication of telomeres.

A 140-bp-long palindromic sequence induces double-strand breaks during meiosis in the yeast Saccharomyces cerevisiae

Palindromic sequences have the potential to form hairpin or cruciform structures, which are putative substrates for several nucleases and mismatch repair enzymes. A genetic method was developed to detect such structures in vivo in the yeast Saccharomyces cerevisiae. Using this method we previously showed that short hairpin structures are poorly repaired by the mismatch repair system in S. cerevisiae. We show here that mismatches, when present in the stem of the hairpin structure, are not processed by the repair machinery, suggesting that they are treated differently than those in the interstrand base-paired duplex DNA. A 140-bp-long palindromic sequence, on the contrary, acts as a meiotic recombination hotspot by generating a site for a double-strand break, an initiator of meiotic recombination. We suggest that long palindromic sequences undergo cruciform extrusion more readily than short ones. This cruciform structure then acts as a substrate for structure-specific nucleases resulting in the formation of a double-strand break during meiosis in yeast. In addition, we show that residual repair of the short hairpin structure occurs in an MSH2-independent pathway.

Inverted repeats, stem-loops, and cruciforms: significance for initiation of DNA replication

Inverted repeats occur nonrandomly in the DNA of most organisms. Stem-loops and cruciforms can form from inverted repeats. Such structures have been detected in pro- and eukaryotes. They may affect the supercoiling degree of the DNA, the positioning of nucleosomes, the formation of other secondary structures of DNA, or directly interact with proteins. Inverted repeats, stem-loops, and cruciforms are present at the replication origins of phage, plasmids, mitochondria, eukaryotic viruses, and mammalian cells. Experiments with anti-cruciform antibodies suggest that formation and stabilization of cruciforms at particular mammalian origins may be associated with initiation of DNA replication. Many proteins have been shown to interact with cruciforms, recognizing features like DNA crossovers, four-way junctions, and curved/bent DNA of specific angles. A human cruciform binding protein (CBP) displays a novel type of interaction with cruciforms and may be linked to initiation of DNA replication.

Kinetoplast maxicircle DNA replication in Crithidia fasciculata and Trypanosoma brucei

Kinetoplast DNA, the mitochondrial DNA of trypanosomatids, is composed of several thousand minicircles and a few dozen maxicircles, all of which are topologically interlocked in a giant network. We have studied the replication of maxicircle DNA, using electron microscopy to analyze replication intermediates from both Crithidia fasciculata and Trypanosoma brucei. Replication intermediates were stabilized against branch migration by introducing DNA interstrand cross-links in vivo with 4,5',8-trimethylpsoralen and UV radiation. Electron microscopy of individual maxicircles resulting from a topoisomerase II decatenation of kinetoplast DNA networks revealed intact maxicircle theta structures. Analysis of maxicircle DNA linearized by restriction enzyme cleavage revealed branched replication intermediates derived from theta structures. Measurements of the linearized branched molecules in both parasites indicate that replication initiates in the variable region (a noncoding segment characterized by repetitive sequences) and proceeds unidirectionally, clockwise on the standard map.

Repair of 8-methoxypsoralen photoinduced cross-links in yeast. Analysis by alkaline step-elution and electron microscopy

The repair of interstrand cross-links induced by 8-methoxypsoralen plus UVA (365 nm) radiation DNA was analyzed in diploid strains of the yeast Saccharomyces cerevisiae. The strains employed were the wild-type D7 and derivatives homozygous for the rad18-1 or the rad3-12 mutation. Alkaline step-elution and electron microscopy were performed to follow the process of induction and removal of photoinduced cross-links. In accordance with previous reports, the D7 rad3-12 strain failed to remove the induced lesions and could not incise cross-links. The strain D7 rad18-1 was nearly as efficient in the removal of 8-MOP photoadducts after 2 h of post-treatment incubation as the D7 RAD+ wild-type strain. However, as demonstrated by alkaline step-elution and electron microscopic analysis, the first incision step at DNA cross-links was three times more effective in D7 rad18-1 than in D7 RAD+. This is consistent with the hypothesis that the RAD18 gene product is involved in the filling of gaps resulting from persistent non-informational DNA lesions generated by the endonucleolytic processing of DNA cross-links. Absence of this gene product may lead to extensive strand breakage and decreased recognition of such lesions by structural repair systems.

Chemical and photochemical probing of DNA complexes

An overview of the chemical and photochemical probes which over the past ten years have been used in studies of DNA/ligand complexes and of non-B-form DNA conformations is presented with emphasis on the chemical reactions of the probes with DNA and on their present 'use-profile'. The chemical probes include: dimethyl sulfate, ethyl nitroso urea, diethyl pyrocarbonate, osmium tetroxide, permanganate, aldehydes, methidiumpropyl-EDTA-Fell (MPE), phenanthroline metal complexes and EDTA/FeII. The photochemical probes that have been used include: psoralens, UVB, acridines and uranyl salts. The biological systems analysed by use of these probes are reviewed by tabulation.

Localization of replication origins in pea chloroplast DNA

The locations of the two replication origins in pea chloroplast DNA (ctDNA) have been mapped by electron microscopic analysis of restriction digests of supercoiled ctDNA cross-linked with trioxalen. Both origins of replication, identified as displacement loops (D-loops), were present in the 44-kilobase-pair (kbp) SalI A fragment. The first D-loop was located at 9.0 kbp from the closest SalI restriction site. The average size of this D-loop was about 0.7 kbp. The second D-loop started 14.2 kbp in from the same restriction site and ended at about 15.5 kbp, giving it a size of about 1.3 kbp. The orientation of these two D-loops on the restriction map of pea ctDNA was determined by analyzing SmaI, PstI, and SalI-SmaI restriction digests of pea ctDNA. One D-loop has been mapped in the spacer region between the 16S and 23S rRNA genes. The second D-loop was located downstream of the 23S rRNA gene. Denaturation mapping of recombinants pCP 12-7 and pCB 1-12, which contain both D-loops, confirmed the location of the D-loops in the restriction map of pea ctDNA. Denaturation-mapping studies also showed that the two D-loops had different base compositions; the one closest to a SalI restriction site denatured readily compared with the other D-loop. The recombinants pCP 12-7 and pCB 1-12 were found to be highly active in DNA synthesis when used as templates in a partially purified replication system from pea chloroplasts. Analysis of in vitro-synthesized DNA with either of these recombinants showed that full-length template DNA was synthesized. Recombinants from other regions of the pea chloroplast genome showed no significant DNA synthesis activity in vitro.

Localization of replication origins in pea chloroplast DNA

The locations of the two replication origins in pea chloroplast DNA (ctDNA) have been mapped by electron microscopic analysis of restriction digests of supercoiled ctDNA cross-linked with trioxalen. Both origins of replication, identified as displacement loops (D-loops), were present in the 44-kilobase-pair (kbp) SalI A fragment. The first D-loop was located at 9.0 kbp from the closest SalI restriction site. The average size of this D-loop was about 0.7 kbp. The second D-loop started 14.2 kbp in from the same restriction site and ended at about 15.5 kbp, giving it a size of about 1.3 kbp. The orientation of these two D-loops on the restriction map of pea ctDNA was determined by analyzing SmaI, PstI, and SalI-SmaI restriction digests of pea ctDNA. One D-loop has been mapped in the spacer region between the 16S and 23S rRNA genes. The second D-loop was located downstream of the 23S rRNA gene. Denaturation mapping of recombinants pCP 12-7 and pCB 1-12, which contain both D-loops, confirmed the location of the D-loops in the restriction map of pea ctDNA. Denaturation-mapping studies also showed that the two D-loops had different base compositions; the one closest to a SalI restriction site denatured readily compared with the other D-loop. The recombinants pCP 12-7 and pCB 1-12 were found to be highly active in DNA synthesis when used as templates in a partially purified replication system from pea chloroplasts. Analysis of in vitro-synthesized DNA with either of these recombinants showed that full-length template DNA was synthesized. Recombinants from other regions of the pea chloroplast genome showed no significant DNA synthesis activity in vitro.

Electron microscopy of single-stranded structures in the DNA of competent Haemophilus influenzae cells

Chromosomal DNAs from exponential-phase and competent cells of Haemophilus influenzae were examined by electron microscopy to determine whether the chromosome undergoes structural changes during competence development. Single-stranded gaps and single-stranded tails formed in chromosomal DNA during competence development. The generation of gaps was dependent on the rec-2 function. Since the rec-2 mutant is defective in the translocation of donor DNA, it was inferred that the gaps were involved in the translocation step of transformation. The generation of single-stranded tails was independent of the rec-1 and rec-2 genes. Therefore, these structures were assumed to play no direct role in the interaction of donor and recipient DNAs during transformation. Gaps were preferentially associated with a readily denaturable, possibly A + T-rich fraction of the genome. This finding raised the possibility that hot spots for transformation might be associated with A + T-rich DNA.

In vivo homologous recombination intermediates of yeast mitochondrial DNA analyzed by electron microscopy

To study the structure of in vivo mitochondrial DNA recombination intermediates in Saccharomyces cerevisiae, we used a deletion mutant of the wild type mitochondrial genome. The mtDNA of this petite is composed of a direct tandem repetition of an approximately 4,600 bp monomer repeat unit with a unique HhaI restriction enzyme site per repeat. The structure of native mtDNA isolated from log phase cells, and mtDNA crosslinked in vivo with trioxsalen plus UVA irradiation, was studied by electron microscopy. Both populations contained crossed strand "Holliday" type recombination intermediates. Digestion of both non-crosslinked and crosslinked mtDNA with the enzyme HhaI released X and H shaped structures composed of two monomers. Electron microscopic analysis revealed that these structures had pairs of equal length arms as required for homologous recombination intermediates and that junctions could occur at points along the entire monomer length. The percentage of recombining monomers in both non-crosslinked and trioxsalen crosslinked mtDNA was calculated by quantitative analysis of all the structures present in an HhaI digest. The relationship between these values and the apparent dispersive replication of mtDNA in density-shift experiments and mtDNA fragility during isolation is discussed.

Perfect palindromic lac operator DNA sequence exists as a stable cruciform structure in supercoiled DNA in vitro but not in vivo

A perfect palindromic 66-base pair (bp) DNA sequence derived from the lac operator and cloned into plasmid pMB9 [Betz, J. L. & Sadler, J. R. (1981) Gene 13, 1-12] can exist in a 66-bp linear form or as two 33-bp cruciform arms. The fraction of the sequence in the cruciform depends on the superhelical density of the plasmid DNA. Relaxed DNA contains no cruciforms. The palindrome in the cruciform structure is cut by EcoRI endonuclease at the base of the cruciform arms, releasing 33-bp fragments; when in the linear form only 66-bp fragments are produced. The cruciform structure is fixed by trimethylpsoralen crosslinks in the cruciform arms. This together with the EcoRI cutting provides an assay for the cruciform structures in the DNA of living cells. Using this assay we show that the cruciform structure rarely if ever exists in vivo, but after DNA isolation greater than 90% of the sequence is in cruciforms. Results suggest that the plasmid DNA as organized in vivo either lacks sufficient torsional tension to form this cruciform or the palindrome is restrained in the linear form by other bound molecules.

Accessibility of ribosomal genes to trimethyl psoralen in nuclei of Physarum polycephalum

We have probed the accessibility of the genes for rRNA in Physarum polycephalum by using the photoreactive DNA cross-linking agent 4,5',8-trimethyl psoralen. Nuclei isolated from actively growing Physarum were treated with trimethyl psoralen and irradiated with 360-nm light in order to form cross-links. The palindromic, extrachromosomal rDNA then was isolated, and the positions of cross-links were determined by electron microscopy of the DNA under totally denaturing conditions. The results indicate that the frequency of cross-linking, after correction for base sequence bias of the reaction, is up to sixfold higher in the transcribed regions than in the central or the terminal spacer regions. There is no detectable heterogeneity among the different rDNA molecules or between the halves of a single molecule. Cross-linked molecules invariably occur in a linear as opposed to a cruciform structure. The preferential cross-linking of the transcribed region is nearly eliminated in spherules, a dormant transcriptionally inactive form in the Physarum life cycle.

Site-directed psoralen crosslinking of DNA

A technique has been developed for placing site-specific crosslinks into DNA by using a psoralen derivative containing a sulfhydryl group. Plasmid pBR322 was nicked at the unique BamHI restriction site, and mercurated pyrimidines were introduced by nick-translation. Then the psoralen was specifically directed to this site in the dark through a Hg-S linkage prior to irradiation to form a crosslink. The location of interstrand cross-links was examined by electron microscopy of Pst I-cut molecules and found to be at or near the BamHI site.

Accessibility of ribosomal genes to trimethyl psoralen in nuclei of Physarum polycephalum

We have probed the accessibility of the genes for rRNA in Physarum polycephalum by using the photoreactive DNA cross-linking agent 4,5',8-trimethyl psoralen. Nuclei isolated from actively growing Physarum were treated with trimethyl psoralen and irradiated with 360-nm light in order to form cross-links. The palindromic, extrachromosomal rDNA then was isolated, and the positions of cross-links were determined by electron microscopy of the DNA under totally denaturing conditions. The results indicate that the frequency of cross-linking, after correction for base sequence bias of the reaction, is up to sixfold higher in the transcribed regions than in the central or the terminal spacer regions. There is no detectable heterogeneity among the different rDNA molecules or between the halves of a single molecule. Cross-linked molecules invariably occur in a linear as opposed to a cruciform structure. The preferential cross-linking of the transcribed region is nearly eliminated in spherules, a dormant transcriptionally inactive form in the Physarum life cycle.

In situ photochemical crosslinking of HeLa cell mitochondrial DNA by a psoralen derivative reveals a protected region near the origin of replication

The in vivo association with proteins of HeLa cell mitochondrial DNA (mtDNA) has been investigated by analyzing the pattern of in situ crosslinking of the DNA by 4'-hydroxymethyl-4, 5',8-trimethylpsoralen (HMT). Either isolated mitochondria or whole cells were irradiated with long wavelength UV light in the presence of ths psoralen derivative, and the mtDNA was then isolated and analyzed in the electron microscope under totally denaturing conditions. No evidence of nucleosomal structure was found. The great majority of the molecules (approximately 90%) had a double-stranded DNA appearance over most of their contour length, with one to several bubbles occupying the rest of the contour, while the remaining 10% of the molecules appeared to be double-stranded over their entire length. Analysis of restriction fragments indicated the presence, in approximately 80% of the molecules, of a protected segment (300 to 1500 bp long) in a region which was centered asymmetrically around the origin of replication so as to overlap extensively the D-loop. Control experiments showed that at most 30% of the bubbles found near the origin could represent D-loops or expanded D-loops: furthermore, it could be excluded that some sequence peculiarity would account for the preferential location of bubbles near the origin of replication. The data have been interpreted to indicate that, in at least 55% of HeLa cell mtDNA molecules, the region around the origin is protected from in situ psoralen crosslinking by proteins or protein complexes which are associated in vivo with the DNA.

Structure and distribution of inverted repeats (palindromes). II. Analysis of DNA of the mouse

The size and distribution of renatured inverted repeats (palindromes) in Mus musculus DNA were examined by electron microscopy (EM). The majority (85%) of the palindromes were found to be clustered in about one half of the DNA strands. The rest of the DNA strands were seen with a solitary looped structure - The unlooped palindromes constituted 53% of all palindromes and were always clustered. There was a significant reduction in the number of unlooped palindromes in comparison to D. melanogaster DNA (Biezunski, 1981) and as a result the palindrome clusters were smaller and contained 2-8 palindromes [4-16 inverted repeats (ir)] per DNA strand. The looped palindromes had a wide and regular distribution with spacing lengths similar to those found in D. melanogaster DNA, and showed some periodicity. The average spacing between centers of all palindromes (inside a cluster) was 4.325 kb, and between centers of looped palindromes 8.544 kb. - The lengths of the ir of unlooped and looped palindromes were grouped (similar to D. melanogaster DNA) in one size-class with a range of 30-240 bp and an average length of 130 bp. Longer ir were also observed and the average length of ir in unlooped palindromes was 186 bp, in looped 588 bp, and the total average length was 375 bp. - It was calculated that there are about 224,000-320,000 palindromes (ir pairs) in the mouse genome, with the spacing between centers of all palindromes about 13-9 kb in length. - In high molecular weight mouse DNA, complex looped structures composed of rows of 5-8 looped palindromes one on "top" the other, formed by renaturation of multiple ir, were observed. It is suggested, that clustered repetitive sequences, in direct and inverted orientation, might be of one family and homologous to one another, and be able to reassociate, in vitro and in vivo, into structures of different forms, which could function as binding sites for various regulatory proteins during mouse development.

Alkaline gel electrophoresis of deoxyribonucleic acid photoreacted with trimethylpsoralen: rapid and sensitive detection of interstrand cross-links

Restriction fragments of phage lambda and phi X174 deoxyribonucleic acid (DNA) were photoreacted with 4,5',8-trimethylpsoralen to various extents, and the amount of covalent cross-linking was determined by electron microscopy of the DNA under totally denaturing conditions. The DNA was then analyzed by electrophoresis in alkaline agarose gels. A single cross-link in a DNA molecule produced a large decrease in its electrophoretic mobility. With DNA fragments 0.3--4 kilobase pairs in size, the apparent Mr (molecular weight) of the cross-linked DNA was 2.0 +/- 0.1 times and Mr of the unreacted, single-stranded DNA. A single cross-link in a larger DNA molecule resulted in an even greater increase in apparent Mr. Further cross-linking produced a decrease in the apparent Mr of the DNA, reaching a plateau at a value of 1.4 +/- 0.1 times the Mr of the unreacted, single-stranded DNA over a large range of fragment sizes (0.6--10 kilobase pairs). The apparent Mr of the cross-linked DNA was weakly dependent on the percentage of agarose in the gel. Although highly sensitive to interstrand cross-links the electrophoretic mobilities appeared to be unaffected by low levels of monoadducts (trimethylpsoralen covalently bound to one strand of the DNA). The DNA bandwidths increased by as much as 4-fold at low extents of cross-linking, presumably due to heterogeneity in the locations of the cross-links in the DNA molecules. The bands became sharp again at high levels of reaction. These observations from the basis of a new assay for interstrand DNA cross-links that is both more sensitive and more convenient than previous methods.

DNA electron microscopy

In recent years DNA electron microscopy has become a tool of increasing interest in the fields of molecular genetics and molecular and cell biology. Together with the development of in vitro recombination and DNA cloning, new electron microscope techniques have been developed with the aim of studying the structural and functional organization of genetic material. The most important methods are based on nucleic acid hybridizations: DNA-DNA hybridization (heteroduplex, D-loop), RNA-DNA hybridization (R-loop), or combinations of both (R-hybrid). They allow both qualitative and quantitative analysis of gene organization, position and extension of homology regions, and characterization of transcription. The reproducibility and resolution of these methods make it possible to map a specific DNA region within 50 to 100 nucleotides. Therefore they have become a prerequisite for determining regions of interest for subsequent nucleotide sequencing. Special methods have been developed also for the analysis of protein-DNA interaction: e.g., direct visualization of specific protein-DNA complexes (enzymes, regulatory proteins), and analysis of structures with higher complexity (chromatin, transcription complexes).

DNA-protein interactions in the Drosophila melanogaster mitochondrial genome as deduced from trimethylpsoralen crosslinking patterns

The location of proteins on the mitochondrial DNA (mtDNA) of Drosophila melanogaster was investigated by trimethylpsoralen photoreaction of embryos disrupted by gentle homogenization. After photoreaction, the mtDNA was isolated and the pattern of DNA crosslinking was determined by electron microscopy of the DNA under totally denaturing conditions. In contrast to nuclear DNA, which showed periodic crosslinks indicative of a nucleosome structure, most of each mtDNA molecule exhibited uniformly heavy crosslinking. A 10% region of the mtDNA was, however, protected from psoralen crosslinking in a distinctive manner: five uncrosslinked segments were closely clustered in the mtDNA. Four were 394 +/- 13 (SD) base pairs in size, while the fifth measured about 200 base pairs. These protected segments mapped within the A+T-rich region of the mtDNA, extending from the end of the A+T-rich region near the Bg1 II cleavage site to the center of the A+T-rich region. Protection of this part of the mtDNA from crosslinking was interpreted to be the result of association with proteins in the mitochondrion because mtDNA that was deproteinized before the photoreaction was uniformly crosslinked over its entire length. The origin of replication of the mtDNA is also located at the center of the A+T-rich region, which suggests that the protection from the psoralen photoreaction may be due to proteins involved in membrane attachment or replication.

Photoaddition of trimethylpsoralen as a probe for the intracellular organization of Escherchia coli DNA

It has been previously demonstrated that photoreaction with 4,5',8-tri-methylpsoralen (trioxsalen) can be used as a probe for the in vivo structure of eucaryotic chromatin. We have used this probe to analyze the organization of intracellular Escherichia coli DNA. In contrast to eucaryotic DNA, bacterial DNA within the intact cell is not protected from saturating doses of trioxsalen photoaddition. The final level of covalently bound trioxsalen upon saturation is identical to that found with purified DNA. In addition, the distribution of interstrand DNA cross-links formed by low doses of trioxsalen photoadducts does not exhibit the repeating pattern that has been observed with eucaryotic nucleosomes.

Telomere replication, kinetochore organizers, and satellite DNA evolution

Robertsonian rearrangements demonstrate one-break chromosome rearrangement and the reversible appearance and disappearance of telomeres and centromeres. Such events are quite discordant with classical cytogenetic theories, which assume all chromosome rearrangements to require at least two breaks and consider centromeres and telomeres as immutable structures rather than structures determined by mutable DNA sequences. Cytogenetic data from spontaneous and induced telomere-telomere fusions in mammals support a molecular model of terminal DNA synthesis in which all telomeres are similar and recombine before replication and subsequent separation. This, along with evidence for a hypothetical DNA sequence, the kinetochore organizer, readily explains latent telomeres, latent centromeres, and reversible (one-break) Robertsonian rearrangements. A second model, involving simply recombination between like satellite DNA sequences on different chromosomes, explains not only how one satellite can simultaneously evolve on different chromosomes, but also why satellite DNA is usually located near centromeres or telomeres and why it maintains a preferred orientation with respect to the centromere.

Improved methods for the formation and stabilization of R-loops

Improved methods for the formation and stabilization of R-loops for visualization in the electron microscope are presented. The two complementary strands of a duplex DNA are photochemically crosslinked once every 1 to 3 kb using 4, 5', 8 trimethylpsoralen. R-loops are then formed by incubation with RNA in 70% formamide at a temperature above the DNA melting temperature. Finally, the R-loops are stabilized by modifying the free single strand of DNA with glyoxal, thus minimizing the displacement of the hybridized RNA by branch migration. In this manner R-loops can be formed and visualized at a high frequency irrespective of the base composition of the nucleic acid of interest.

Yeast chromosomal DNA molecules have strands which are cross-linked at their termini

The microbial eukaryote Saccharomyces cerevisiae has 18 chromosomes, each consisting of a DNA molecule of 1 x 15 x 10(8) daltons (150 to 2,300 kilobase pairs). Interstand cross-links have now been found in molecules of all sizes by examining the ability of high molecular weight DNA to snap back, i.e., to rapidly renature after denaturation. Experiments in which snap back was assessed for molecules broken by shearing indicate that there are probably two cross-links in each chromosome. Evidence that the cross-links occur at specific sites in the genome was obtained by treating total chromosomal DNA with the endonuclease EcoRI which cleaves the yeast genome into approximately 2,000 discrete fragments. Cross-link containing fragments were separated from fragments without cross-links. This purification resulted in enrichment for about 18 specific fragments. To determine whether the cross-links are terminal or at internal sites in chromosomal DNA, large shear-produced fragments were examined by electron microsopy. With complete denaturation few fragments exhibit the X-shaped single strand configuration expected for internal cross-links. When partially denatured fragments were examined some ends had single strand loops as expected for (AT-rich) cross-linked termini. We propose that a duplex chromosomal DNA molecules have cross-linked termini. We propose that a duplex chromosomal DNA molecule in this eukaryote consists of a continuous, single, self-complementary strand of DNA. This structure has implications for the mechanism of chromosome replication and may be the basis of telomere behavior.

Heterogeneous nuclear RNA double-stranded regions probed in living HeLa cells by crosslinking with the psoralen derivative aminomethyltrioxsalen

The psoralen derivative aminomethyltrioxsalen (AMT, 4'-aminomethyl-4,5',8-trimethylpsoralen) has been employed as a probe for heterogeneous nuclear RNA (hnRNA) double-stranded regions in experiments with living HeLa cells. hnRNA ribonucleoprotein (hnRNP) particles were purified from untreated or AMT-treated cells after irradiation with 365-nm light, and double-stranded hnRNA regions (dsRNA) were isolated by RNase A + T1 digestion of hnRNP, followed by preparative Cs2SO4 isopycnic centrifugation. The purified, hnRNP-derived dsRNA was then assayed for interstrand crosslinks by measurement of its "snapback" to RNase-resistant form after thermal denaturation. By this procedure, the amount of crosslinked dsRNA was found to be increased 3- to 7-fold in cells exposed to AMT in vivo. The levels of crosslinking in vivo compared favorably with those observed in model experiments with pure dsRNA in vitro. These results establish that double-stranded hnRNA regions exist in the living cell, and they further demonstrate that these base-paired regions are organized as rather accessible sites within the nucleus.

Activation of endogenous type C virus in BALB/c mouse cells by herpesvirus DNA

Several virion and nonvirion DNAs were tested for the ability to activate endogenous type C virus in BALB/c-derived mouse cells using the calcium precipitation technique. The DNAs from all herpesviruses tested activated xenotropic type C virus synthesis. These included DNAs from herpes simplex virus types 1 and 2, Epstein-Barr virus, human cytomegalovirus, SA8 virus, infectious bovine rhinotracheitis virus, pseudorabies virus, and herpes saimiri virus (M-DNA). In contrast, DNAs from vaccinia virus, simian virus 40, primate cells, bacteria, mycoplasma, and salmon sperm showed no ability to activate type C virus when tested under the same conditions. Several herpesviruses and vaccinia virus, which were highly infectious for the BALB/c cells used, were tested for their ability to activate type C virus after UV irradiation. All herpesviruses tested were positive, while vaccinia virus was negative. Unirradiated simian virus 40 also showed no ability to activate type C virus. Activation of type C virus by DNA from herpes simplex virus was observed after shearing or sonication of the DNA to an average size of 3 x 10(6) daltons, but was not observed with DNA sonicated to an average size of 1 x 10(6) daltons. Alkali denaturation of DNA from herpes simplex virus or treatment with DNase, but not RNase, destroyed its ability to activate type C virus, as did crosslinking of the DNA with 4,5',8-trimethylpsoralen (psoralen) and light.

Photochemical addition of the cross-linking reagent 4,5', 8-trimethylpsoralen (trioxaslen) to intracellular and viral simian virus 40 DNA-histone complexes

We demonstrated here that 4,5', 8-trimethylpsoralen (trioxsalen) is a valuable probe for the structure of SV40 DNA-histone complexes. Trioxsalen readily penetrated intact cells and, in the presence of 340- to 380-nm light, covalently cross-linked DNA preferentially at the sites available for micrococcal nuclease digestion. Histograms of the lengths of the regions of SV40 DNA protected from cross-linking, as visualized by electron microscopy, indicated a repeating pattern of base pairs in DNA from both infected cells and virus particles. The ability of the trioxsalen probe to act in vivo and to map the location of protected regions may provide a powerful tool for analyzing the role of nucleosomes in the structure of the virus particle and in intracellular complexes such as transcription templates and replication intermediates.

Characterization of inverted repeated sequences in wheat nuclear DNA

The properties of inverted repeated sequences in wheat nuclear DNA have been studied by HAP(1) chromatography, nuclease S1 digestion and electron microscopy. Inverted repeated sequences comprise 1.7% of wheat genome. The HAP studies show that the amount of "foldback HAP bound DNA" depends on DNA length. Inverted repeats appear to be clustered with an average intercluster distance of 25 kb. It is estimated that there are approximately 3 x 10(6) inverted repeats per haploid wheat genome. The sequences around inverted repeats involve all families of repetition frequencies. Inverted repeats are observed as hairpins in electron microscopy. 20% of hairpins are terminated by a single-stranded spacer ranging from 0.3 to 1.5 kb in length. Duplex regions of the inverted repeats range from 0.1 to 0.45 kb with number average values of 0.24 kb and 0.18 kb for unlooped and looped hairpin respectively. Thermal denaturations and nuclease S1 digestions have revealed a length of about 100 bases for duplex regions. The methods used to study inverted repeated sequences are compared and discussed.

Production of DNA bifilarly substituted with bromodeoxyuridine in the first round of synthesis: branch migration during isolation of cellular DNA

Incubation of human lymphoid cells with bromodeoxyuridine (BrdUrd) for short periods produces three classes of DNA containing analog: DNAHL (hybrid DNA, density approximately equal to 1.75 g/cm3), DNAint (intermediate density DNA, density approximately equal to 1.71 g/cm3), and DNAHH (DNA with both strands containing analog, density approximately equal to 1.80 g/cm3). Preparations of DNAint yield DNAHH after extensive shearing and/or treatment with single strand specific endonuclease. Cross-linking of pulse-labeled (BrdUrd + 3HdT) DNA in cells by treatment with trioxsalen and near UV light before lysis prevents the appearance of DNAHH.Cross-linking after lysis has little effect. A large fraction of DNAHH is obtained after incubation of cells with caffeine. Extraction of DNA at high salt concentration or cross-linking with trioxsalen and near UV light drastically reduced the amount of DNAHH obtained from caffeine-treated cells. We conclude that most DNAHH arises from in vitro branch migration in isolated DNA growing points.

Analysis of single- and double-stranded nucleic acids on polyacrylamide and agarose gels by using glyoxal and acridine orange

We have developed a simple and rapid system for the denaturation of nucleic acids and their subsequent analysis by gel electrophoresis. RNA and DNA are denatured in 1 M glyoxal (ethanedial) and 50% (vol/vol) dimethyl sulfoxide, at 50 degrees. The glyoxalated nucleic acids are then subjected to electrophoresis through either acrylamide or agarose gels in a 10 mM sodium phosphate buffer at pH 7.0. When glyoxalated DNA molecules of known molecular weights are used as standards, accurate molecular weights for RNA are obtained. Furthermore, we have employed the metachromatic stain acridine orange for visualization of nucleic acids in gels. This dye interacts differently with double- and single-stranded polynucleotides, fluorescing green and red, respectively. By using these techniques, native and denatured DNA and RNA molecules can be analyzed on the same slab gel.

Psoralen-DNA photoreaction: controlled production of mono- and diadducts with nanosecond ultraviolet laser pulses

Samples of DNA which contain adducts of a new psoralen derivative, but no cross-links, have been prepared by irradiating mixtures of DNA and the derivative with single, 15-nanosecond pulses of laser light. Succeeding pulses introduce cross-links. The ability to rapidly and selectively create monoadducts and corss-links may allow the use of psoralens as probes of dynamic processes in living cells.