The origin of adenovirus DNA replication: minimal DNA sequence requirement in vivo.

Adenovirus mini-chromosomes which contain two cloned, inverted adenovirus termini replicate in vivo when supplied with non-defective adenovirus as a helper. This system has been used to define the minimum cis acting DNA sequences required for adenovirus DNA replication in vivo. Deletions into each end of the adenovirus inverted terminal repeat (ITR) were generated with Bal31 exonuclease and the resulting molecules constructed into plasmids which contained two inverted copies of the deleted ITR separated by the bacterial neomycin phosphotransferase gene. To determine the effect of the deletion in vivo plasmids cleaved to expose the adenovirus termini were co-transfected with adenovirus type 2 DNA into tissue culture cells. The replicative ability of the molecules bearing adenovirus termini was assayed by Southern blotting of extracted DNA which had been treated with DpnI, a restriction enzyme which cleaves only methylated and therefore unreplicated, input DNA. Molecules containing the terminal 45 bp of the viral genome were fully active whereas molecules containing only 36 bp were in-active in this assay. Therefore sequences required for DNA replication are contained entirely within the terminal 45 bp of the viral genome. Thus, both the previously described highly conserved region (nucleotides 9-18) and the binding site for the cellular nuclear factor I (nucleotides 19-48) are essential for adenovirus DNA replication in vivo.     

Origin of adenovirus DNA replication. Role of the nuclear factor I binding site in vivo

An assay is described that detects in vivo a single round of initiation and DNA synthesis directed by a linear molecule containing an exposed single copy of an adenovirus (Ad) origin of replication. This and a previously described assay, which measures multiple rounds of DNA replication, were used to identify DNA sequences within the Ad2 and Ad4 origins of replication that are important for ori function. Linear DNA molecules containing sequences from the Ad2 or Ad4 genome termini were cotransfected with homologous and heterologous helper virus, and net amounts of DNA synthesis were compared. Linear molecules containing the Ad4 inverted terminal repeats were replicated 20-fold better in the presence of the homologous helper, whereas both Ad2 and Ad4 inverted terminal repeats were utilized efficiently by Ad4. DNA sequence analysis of the Ad2 ori and the corresponding region in Ad4 indicated that, although there are only ten variant base-pairs, eight are located within the Ad2 DNA sequence recognized by the cellular protein nuclear factor I. This protein is required to achieve the maximal rate of Ad2 DNA replication in vitro, and these differences therefore identify DNA sequences that are crucial to Ad2 ori function. The Ad4 ITR does not contain a functional nuclear factor I binding site, and deletion analysis has demonstrated that this region of the Ad4 genome is not required for ori function. In contrast to Ad2, the DNA sequences required for the initiation of Ad4 DNA replication were shown to reside entirely within the terminal 18 base-pairs of the Ad4 inverted terminal repeat.     

Most chloroplast DNA of maize seedlings in linear molecules with defined ends and branched forms

We used pulsed-field gel electrophoresis, restriction fragment mapping, and fluorescence microscopy of individual DNA molecules to analyze the structure of chloroplast DNA (cpDNA) from shoots of ten to 14 day old maize seedlings. We find that most of the cpDNA is in linear and complex branched forms, with only 3-4% as circles. We find the ends of linear genomic monomers and head-to-tail (h-t) concatemers within inverted repeat sequences (IRs) near probable origins of replication, not at random sites as expected from broken circles. Our results predict two major and three minor populations of linear molecules, each with different ends and putative origins of replication. Our mapping data predict equimolar populations of h-t linear concatemeric molecules differing only in the relative orientation (inversion) of the single copy regions. We show how recombination during replication can produce h-t linear concatemers containing an inversion of single copy sequences that has for 20 years been attributed to recombinational flipping between IRs in a circular chromosome. We propose that replication is initiated predominantly on linear, not circular, DNA, producing multi-genomic branched chromosomes and that most replication involves strand invasion of internal regions by the ends of linear molecules, rather than the generally accepted D-loop-to-theta mechanism. We speculate that if the minor amount of cpDNA in circular form is useful to the plant, its contribution to chloroplast function does not depend on the circularity of these cpDNA molecules.     

Covalently closed circles of human adenovirus DNA are infectious

Replication of the linear adenovirus DNA molecule is thought to result from semiconservative synthesis off linear templates, starting from origins at either end of the genome. Recently, however, it has been shown that in cells infected with adenovirus type 5 (Ad5) a significant fraction of the ends of viral DNA molecules become joined head-to-tail due at least in part to the formation of covalently closed circles. Circular DNA is not present in virions but joining of the ends of viral DNA is detectable shortly after infection, well before the onset of viral DNA replication. To learn more about the structure and possible function of these circular forms of viral DNA, I have cloned Ad5 circles as plasmids replicating in Escherichia coli. Two plasmids have been analyzed in detail and shown to generate infectious virus with an efficiency comparable with that of virion DNA following transfection into human cells. These results suggest that circles are not totally inert or functionless but that, once formed, they are capable of re-entering the pool of replicating molecules to generate linear progeny.     

Protein-primed DNA replication: role of inverted terminal repeats in the Escherichia coli bacteriophage PRD1 life cycle.

Escherichia coli bacteriophage PRD1 and its relatives contain linear double-stranded DNA genomes, the replication of which proceeds via a protein-primed mechanism. Characteristically, these molecules contain 5'-covalently bound terminal proteins and inverted terminal nucleotide sequences (inverted terminal repeats [ITRs]). The ITRs of each PRD1 phage species have evolved in parallel, suggesting communication between the molecule ends during the life cycle of these viruses. This process was studied by constructing chimeric PRD1 phage DNA molecules with dissimilar end sequences. These molecules were created by combining two closely related phage genomes (i) in vivo by homologous recombination and (ii) in vitro by ligation of appropriate DNA restriction fragments. The fate of the ITRs after propagation of single genomes was monitored by DNA sequence analysis. Recombinants created in vivo showed that phages with nonidentical genome termini are viable and relatively stable, and hybrid phages made in vitro verified this observation. However, genomes in which the dissimilar DNA termini had regained identical sequences were also detected. These observations are explained by a DNA replication model involving two not mutually exclusive pathways. The generality of this model in protein-primed DNA replication is discussed.     

Adenoviruses with nonidentical terminal sequences are viable.

Adenovirus genomes consist of linear DNA molecules containing inverted terminal repeat sequences (ITRs) of 100 to 200 base pairs. The importance of identical termini for viability of adenoviruses was investigated. The viral strains used in this study were wild-type adenovirus type 5 (Ad5) and a variant Ad2 strain with termini which were distinct from those of all other human adenoviruses sequenced to date. A hybrid virus (sub54), obtained by recombination between Ad2 and Ad5, derived the left 42 to 52% of its genome from Ad2 and the right 58 to 48% from Ad5. Southern blotting analysis with labeled oligodeoxynucleotides indicated that both Ad2 and Ad5 ITRs were present in sub54 viral DNA preparations, and successive plaque purifications of sub54 demonstrated that viruses with nonidentical terminal sequences were viable but were rapidly converted to viruses with identical ends. Cloning of the sub54 genome as a bacterial plasmid supported the observations made by analysis of sub54 virion DNA. A plasmid, pFG154, was isolated which contained the entire adenovirus genome with an Ad2 ITR at the left terminus covalently linked to an Ad5 ITR at the right terminus. Upon transfection of mammalian cells with pFG154, viral progeny were obtained which had all possible combinations of termini, thus confirming that molecules with nonidentical termini are viable. Pure populations of viruses with nonidentical termini could not be isolated, suggesting efficient repair of one end with the opposite terminus used as a template. A model for this process is proposed involving strand displacement replication and emphasizing the importance of panhandle formation (annealing of terminal sequences) as a replicative intermediate.     

A size analysis of the adenovirus replicon

The linear double-stranded genome of adenovirus DNA replicates semiconservatively from two origins of replication at either of the two molecular ends. Using an in vitro replication system which is able to initiate de novo DNA synthesis we have mapped the origin of DNA replication within the terminal 19 bp of the viral genome. Our conclusions are based on the use of different natural DNA templates, i.e., adenovirus type 2 and mouse adenovirus Fl DNA. In addition, we have employed linearized plasmid DNA templates which contain cloned terminal restriction enzyme fragments as well as chemically synthesized adenovirus termini of different length.     

Structure of the inverted terminal repetition of adenovirus type 2 DNA.

Several secondary structure features involving the ends of single strands of adenovirus type 2 DNA have been studied by electron microscopy by both the gene 32-ethidium bromide technique and a modification of the standard formamide-cytochrome c technique. A duplex stem of length 115 +/- 10 nucleotide pairs due to pairing between the two members of the inverted terminal repetition is observed in the single-stranded circles that form upon annealing single-stranded linear molecules. This duplex stem is shown to lie at the ends of the DNA by using several reference markers: (i) a newly discovered secondary structure feature (a loop of length ca. 500 nucleotides with a 20-nucleotide pair duplex stem) that maps 73% of the full length from the left end of the molecule and (ii) a duplex region due to a hybridized restriction fragment. There is also some secondary structure within each end of linear single strands. There is some variation in the morphology of the end strucures, and we propose that these involve base pairing, as in a tRNA clover leaf, rather than an exact single hairpin-type inverted repeat. These observations are consistent with the hypothesis that there is a foldback structure at the 3' ends of the DNA that functions as a primer for the initiation of replication.     

The nucleotide sequence of the right-hand terminus of adenovirus type 5 DNA: implications for the mechanism of DNA replication.

The nucleotide sequence of the right-hand terminal 3% of adenovirus type 5 (Ad5) DNA has been determined, using the chemical degradation technique developed by Maxam and Gilbert (1977). This region of the genome comprises the 1003 basepair long HindIII-I fragment and the first 75 nucleotides of the adjacent HindIII-F fragment, extending from the right-hand terminus to the sequences from which the main body of the mRNA of early region 4 is transcribed. One of the origins of adenovirus DNA replication is located within this part of the genome. The sequencing results are discussed in relation to several models proposed for the mechanism of replication of linear DNA molecules, which invariably depend on the presence of specific arrangements of nucleotides at the termini of those linear DNAs.     

Replication of a linear mini-chromosome with terminal inverted repeats from the Kluyveromyces lactis linear DNA plasmid k2 in the cytoplasm of Saccharomyces cerevisiae

The k1 and k2 linear DNA plasmids of Kluveromyces lactis replicate in the cytoplasm under the control of plasmid-encoded genes. These plasmids can also replicate autonomously in the cytoplasm of mitochondrial DNA-deficient strains of Saccharomyces cerevisiae. Essential for replication are plasmid-specific terminal inverted repeats (TIRs) to which a terminal protein (TP) is attached at the 5' ends. A plasmid was constructed with k2 TIRs in opposite orientations and with a selectable marker (URA3) under the control of k1UCS2 (upstream conserved sequence 2, the promoter of k1 open reading frame 2) in between the TIRs. Transformation of k1- and k2-containing S. cerevisiae with a fragment generated by releasing the TIR-flanked fragment from the plasmid by restriction digestion was very efficient, despite the absence of a TP. Transformation was also achieved with a fragment generated by PCR. Southern blotting demonstrated that transformants contained multiple copies of DNA fragments with the same size as the transforming DNA, supporting the hypothesis that these were replicating linear mini-chromosomes. The high frequency of transformation strongly suggests that these mini-chromosomes readily replicate supported by k2. Derivatives with a heterologous gene, firefly luciferase (LUC), expressed luciferase at high levels provided the gene was adjacent to a cytoplasmic plasmid promoter (k2UCS5).     

Characterization of the origins of replication of bacteriophage phi 29 DNA.

The origins of replication of phi 29 DNA have been studied by analyzing the activity as templates in the phi 29 in vitro replication system of E. coli recombinant plasmids and M13 derivatives containing phi 29 DNA terminal sequences. Plasmid pITR, containing the 6 bp long inverted terminal repeat of phi 29 DNA, was shown to be essentially inactive. The analysis of a series of deletion derivatives of plasmid pID13, that contains the 73 and 269 bp from the left and right phi 29 DNA ends, respectively, indicated that the minimal origins of replication are comprised within the mutagenesis at these sequences was carried out. Changes of the second or third A into a C completely abolished the template activity. In the case of changes at position from 4 to 12, only 3 out of 14 mutations reduced the template activity; these 3 mutations were double changes and 2 of them affected the inverted terminal repeat. The results suggest that the sequence requirement at the end-proximal region of the origin of replication is more strict than that at the distal region.     

Linear molecules of tobacco ptDNA end at known replication origins and additional loci

Higher plant plastid DNA (ptDNA) is generally described as a double-stranded circular molecule of the size of the monomer of the plastid genome. Also, the substrates and products of ptDNA replication are generally assumed to be circular molecules. Linear or partly linear ptDNA molecules were detected in our present study using pulsed-field gel electrophoresis and Southern blotting of ptDNA restricted with ‘single cutter’ restriction enzymes. These linear DNA molecules show discrete end points which were mapped using appropriate probes. One possible explanation of discrete ends would be that they represent origins of replication. Indeed, some of the mapped ends correlate well with the known origins of replication of tobacco plastids, i.e. both of the oriA sequences and—less pronouncedly—with the oriB elements. Other ends correspond to replication origins that were described for Oenothera hookeri, Zea mays, Glycine max and Chlamydomonas reinhardtii, respectively, while some of the mapped ends were not described previously and␣might therefore represent additional origins of replication.     

In vitro resolution of covalently joined AAV chromosome ends

We have developed an assay for a key step in the replication of adeno-associated virus (AAV) DNA. We demonstrate the covalently joined ends of linear AAV DNA can be resolved in vitro to the open duplex configuration. Only extracts prepared from human cells that have been infected with both adenovirus and AAV are capable of carrying out the reaction. The reaction is initiated by a site-specific and strand-specific endonucleolytic cut at a terminal resolution site near the end of the AAV terminal palindrome. During resolution the orientation of the terminal palindrome is inverted, and the 3' viral strand is extended by DNA synthesis. The size of the newly synthesized 3' strand is nearly identical to that found in viral particles. These observations provide direct biochemical evidence for an essential step in the model for AAV DNA replication.     

Sequence of inverted terminal repetitions from different adenoviruses: demonstration of conserved sequences and homology between SA7 termini and SV40 DNA.

We have established the nucleotide sequence for the inverted terminal repetition of human adenovirus type 3, a subgroup B adenovirus. The repetition, which is 136 bp long, shows a high degree of homology with the known sequence for the inverted repetition of adenovirus type 5 (Steenbergh et al., 1977) a subgroup C adenovirus. Partial sequence information convering 120 bp of the inverted terminal repetitions of human serotype 12, a subgroup A member, and of simian adenovirus type 7 has also been obtained. A comparison of the established sequences shows that the terminal repetitions, in particular the first 50 bp from the ends, contain sequences that have been well conserved in adenovirus evolution. For instance, only six mismatched base pairs were detected among the first 50 bp in the repetitions of simian adenovirus type 7 and human adenovirus type 5, although the homology between simian adenovirus 7 and human subgroup C adenoviruses was estimated to be only 30%. A 14 bp sequence located 9-22 nucleotides from the ends is present in DNAs from all the human serotypes examined as well as in simian adenovirus 7 DNA. Furthermore, the simian adenovirus 7 repetition contains a 21 bp sequence which is present in SV40 DNA, close to the origin of DNA replication.     

Adenovirus DNA replication in vitro: a protein linked to the 5' end of nascent DNA strands.

Soluble nuclear extracts prepared from adenovirus-infected HeLa cells supported adenovirus DNA replication with exogenous DNA-protein complex as template, but protease-treated, phenol-extracted DNA was less active. Replication was enhanced when creatine phosphate and creatine phosphokinase were included in the reaction mixture, rendering the reaction independent of exogenous ATP. Genomic-length, newly synthesized DNA strands were first observed 30 min after initiation of replication and continued to increase in amount for at least 4 h. Thus, the rate of replication is consistent with previous estimates of the rate of replication in vivo. Nascent DNA strands bound to benzoylated, naphthoylated DEAE-cellulose due to their association with protein. The 5' termini of nascent DNA strands were resistant to the 5'- to 3'-specific T7 exonuclease, and the 3' termini of nascent strands were sensitive to the 3'- to 5'-specific exonuclease III. These results suggest that a protein becomes covalently linked to the 5' termini of nascent DNA strands replicated in vitro. Nuclear extracts prepared from adenovirus type 2-infected cells also supported replication of DNA-protein complex prepared from the unrelated type 7 adenovirus. The limited sequence homology between these two viruses at the origin of replication further defines recognition sequences at the origin. These results are discussed in terms of a model for adenovirus DNA replication in which the terminal protein and sequences within the inverted terminal repetition are involved in the formation of an initiation complex that is able to prime DNA replication.     

Restriction endonuclease mapping of adenovirus 35, a type isolated from immunocompromised hosts.

The DNA of adenovirus 35 (Ad35), a type recently associated with infections in immunocompromised hosts, was mapped by the use of BamHI, SmaI, PstI, EcoRI, and HpaI restriction endonucleases. In addition to standard mapping procedures, we used the in vitro adenovirus DNA replication system with origins at both physical ends of the linear molecule to determine the terminal fragments. Deletions of single restriction endonuclease sites in a group of closely related adenovirus isolates from patients with acquired immunodeficiency syndrome helped determine the location of some DNA fragments on the genome.     

Origin-Dependent Inverted-Repeat Amplification: Tests of a Model for Inverted DNA Amplification

DNA replication errors are a major driver of evolution—from single nucleotide polymorphisms to large-scale copy number variations (CNVs). Here we test a specific replication-based model to explain the generation of interstitial, inverted triplications. While no genetic information is lost, the novel inversion junctions and increased copy number of the included sequences create the potential for adaptive phenotypes. The model—Origin-Dependent Inverted-Repeat Amplification (ODIRA)—proposes that a replication error at pre-existing short, interrupted, inverted repeats in genomic sequences generates an extrachromosomal, inverted dimeric, autonomously replicating intermediate; subsequent genomic integration of the dimer yields this class of CNV without loss of distal chromosomal sequences. We used a combination of in vitro and in vivo approaches to test the feasibility of the proposed replication error and its downstream consequences on chromosome structure in the yeast Saccharomyces cerevisiae. We show that the proposed replication error—the ligation of leading and lagging nascent strands to create “closed” forks—can occur in vitro at short, interrupted inverted repeats. The removal of molecules with two closed forks results in a hairpin-capped linear duplex that we show replicates in vivo to create an inverted, dimeric plasmid that subsequently integrates into the genome by homologous recombination, creating an inverted triplication. While other models have been proposed to explain inverted triplications and their derivatives, our model can also explain the generation of human, de novo, inverted amplicons that have a 2:1 mixture of sequences from both homologues of a single parent—a feature readily explained by a plasmid intermediate that arises from one homologue and integrates into the other homologue prior to meiosis. Our tests of key features of ODIRA lend support to this mechanism and suggest further avenues of enquiry to unravel the origins of interstitial, inverted CNVs pivotal in human health and evolution.     

Functional homology between the sequence-specific DNA-binding proteins nuclear factor I from HeLa cells and the TGGCA protein from chicken liver.

Nuclear factor I from HeLa cells, a protein with enhancing function in adenovirus DNA replication, and the chicken TGGCA protein are specific DNA-binding proteins that were first detected by independent methods and that appeared to have similar DNA sequence specificity. To test whether they are homologous proteins from different species we have compared (i) their DNA binding properties and (ii) their function in reconstituted adenovirus DNA replication systems. Using deletion and substitution mutants derived from the DNA binding site on the adenovirus 2 inverted terminal repeat, it was found that the two proteins protect the same 24-nucleotide region of both strands against DNase I digestion and that they have identical minimal recognition sequences of 15 bp containing dyad symmetry. Like nuclear factor I, the TGGCA protein enhances the initiation reaction of adenovirus 2 DNA replication in vitro in a DNA recognition site-dependent manner.     

Infectious circular DNA of human adenovirus type 5: regeneration of viral DNA termini from molecules lacking terminal sequences.

A series of plasmids containing the entire human adenovirus genome with viral DNA termini joined 'head to tail' has been isolated. Several plasmids were able to generate infectious virus following transfection of human cells in spite of having small deletions and rearrangements at the junctions of termini. One plasmid has lost 2 bp of DNA from one end of the viral genome and 11 bp from the other end yet produced viruses with complete wild-type sequences at both ends of the genome. We propose a model for replication of viral DNA off circular templates in which regeneration of terminal information involves translocation of primer and polymerase during initiation of DNA replication. The model suggests a novel mechanism for extension of the 5' ends of linear DNA molecules which could be applicable to chromosomal telomeres.     

Two complementary strand-specific termination sites for adenovirus DNA replication

Adenovirus type 2 DNA, specifically labeled at the termini for DNA replication, was prepared by isolation of viral DNA molecules which were completed during short pulses with ³H-thymidine. The distribution of radioactivity in the two complementary strands at the termini for DNA replication was determined by liquid phase hybridization and gel electrophoresis. At the right-hand terminus, nearly all radioactivity was found in the viral h strand, whereas at the left-hand terminus, most radioactivity was confined to the viral I strand. The results suggest that both molecular ends serve as origins and termini for replication of adenovirus type 2 DNA.