A yeast replication origin consists of multiple copies of a small conserved sequence

Autonomously replicating sequences (ARSs) of the yeast S. cerevisiae function as replication origins on plasmids and probably also on chromosomes. ARS function requires a copy of the ARS core consensus (5'-[A/T]TTTAT[A/G]TTT[A/T]-3') and additional sequences 3' to the T-rich strand of the consensus. Our analysis of an ARS from chromosome III, the C2G1 ARS, suggests that ARS function depends on the presence of an exact match to the core consensus and the presence of additional near matches in the 3' flanking region. We have demonstrated that ARS function can be mediated by multiple matches to the core consensus by constructing synthetic ARS elements from oligonucleotides containing copies of the consensus sequence. We find that two copies of the core consensus are sufficient for ARS activity and that an artificial ARS as efficient as a natural chromosomal ARS can be constructed from multiple core consensus elements in a specific orientation.     

DNA sequence analysis of ARS elements from chromosome III of Saccharomyces cerevisiae: identification of a new conserved sequence.

Four fragments of Saccharomyces cerevisiae chromosome III DNA which carry ARS elements have been sequenced. Each fragment contains multiple copies of sequences that have at least 10 out of 11 bases of homology to a previously reported 11 bp core consensus sequence. A survey of these new ARS sequences and previously reported sequences revealed the presence of an additional 11 bp conserved element located on the 3' side of the T-rich strand of the core consensus. Subcloning analysis as well as deletion and transposon insertion mutagenesis of ARS fragments support a role for 3' conserved sequence in promoting ARS activity.     

Identification and purification of a protein that binds the yeast ARS consensus sequence

We have identified a yeast protein that binds specifically to the ARS consensus sequence. By two-step chromatography we have purified the factor to apparent homogeneity as a single polypeptide of 67 kd. The purified ARS consensus-binding protein (ACBP) recognizes the ARS consensus of the four genomic ARS elements tested, binding preferentially to the T-rich single strand. Point mutations in the consensus significantly reduce the affinity of the single-strand binding. At the histone H4 ARS, ACBP recognizes both the perfect ARS consensus and a 9/11 match 3' of it. These two binding sites correlate with the boundaries of the minimal functional H4 ARS element. A similar configuration of binding sites is found at ARS1. We propose a model implicating this factor in an early step of the initiation of DNA replication.     

Interaction of the H4 autonomously replicating sequence core consensus sequence and its 3''-flanking domain.

Yeast autonomously replicating sequence (ARS) elements are composed of a conserved 11-base-pair (bp) core consensus sequence and a less well defined 3'-flanking region. We have investigated the relationship between the H4 ARS core consensus sequence and its 3'-flanking domain. The minimal sequences necessary and sufficient for function were determined by combining external 3' and 5' deletions to produce a nested set of ARS fragments. Sequences 5' of the core consensus were dispensable for function, but at least 66 bp of 3'-flanking domain DNA was required for full ARS function. The importance of the relative orientation of the core consensus element with respect to the 3'-flanking domain was tested by precisely inverting 14 bp of DNA including the core consensus sequence by oligonucleotide mutagenesis. This core inversion mutant was defective for all ARS function, showing that a fixed relative orientation of the core consensus and 3'-flanking domain is required for function. The 3'-flanking domain of the minimal functional H4 ARS fragment contains three sequences with a 9-of-11-bp match to the core consensus. The role of these near-match sequences was tested by directed mutagenesis. When all near-match sequences with an 8-of-11-bp match or better were simultaneously disrupted by point mutations, the resulting ARS construct retained full replication function. Therefore, multiple copies of a sequence closely related to the core consensus element are not required for H4 ARS function.     

Ease of DNA unwinding is a conserved property of yeast replication origins.

Autonomously replicating sequence (ARS) elements function as plasmid replication origins. Our studies of the H4 ARS and ARS307 have established the requirement for a DNA unwinding element (DUE), a broad easily-unwound sequence 3' to the essential consensus that likely facilitates opening of the origin. In this report, we examine the intrinsic ease of unwinding a variety of ARS elements using (1) a single-strand-specific nuclease to probe for DNA unwinding in a negatively-supercoiled plasmid, and (2) a computer program that calculates DNA helical stability from the nucleotide sequence. ARS elements that are associated with replication origins on chromosome III are nuclease hypersensitive, and the helical stability minima correctly predict the location and hierarchy of the hypersensitive sites. All well-studied ARS elements in which the essential consensus sequence has been identified by mutational analysis contain a 100-bp region of low helical stability immediately 3' to the consensus, as do ARS elements created by mutation within the prokaryotic M13 vector. The level of helical stability is, in all cases, below that of ARS307 derivatives inactivated by mutations in the DUE. Our findings indicate that the ease of DNA unwinding at the broad region directly 3' to the ARS consensus is a conserved property of yeast replication origins.     

A single-stranded DNA binding protein from S. cerevisiae specifically recognizes the T-rich strand of the core sequence of ARS elements and discriminates against mutant sequences.

A protein named ssARS-T binding protein has been purified from yeast that specifically binds to the T-rich strand of the consensus core sequence of yeast autonomously replicating sequence (ARS) elements. As assayed from gel mobility shift experiments the ssARS-T protein shows characteristics of a sequence specific single-stranded DNA binding protein. The complementary A-rich strand of the ARS core sequence is bound much more weakly and no binding can be detected for the double-stranded form of the core sequence. Three single base substitutions in the core sequence that are known to abolish ARS function in vivo also lead to weaker binding of the core sequence to the ssARS-T protein in vitro. The strong correlation between the binding of mutated sequences in vitro and the ARS properties of these sequences in vivo points to an essential function of the ssARS-T protein during replication initiation in yeast ARS elements.     

Effect of ARS1 mutations on chromosome stability in Saccharomyces cerevisiae.

We have used a set of deletion mutations in the ARS1 element of Saccharomyces cerevisiae to measure their effect on chromosome stability. This work establishes the previously proposed existence of three domains in ARS1. Domain C, which we have previously inferred, but not proved, to be a part of ARS1, is now established. In addition, we show that increasingly large deletions of the domain have increasingly large effects, which was not realized before. Furthermore, we have provided the first positive evidence for the central importance of a 14-base-pair core sequence containing the ARS consensus element by showing that it has the ability to act as a replicator on a plasmid containing no other ARS1 flanking sequence. The method of analyzing plasmid stability used in our study employs a novel and sensitive flow cytometry assay for beta-galactosidase. We discuss ways in which flow cytometry, based on this assay, could be generalized beyond its particular application in this work to studying other aspects of the cell biology of yeast and higher cells. The actual flow cytometry method will be described in detail elsewhere.     

cis-acting components in the replication origin from ribosomal DNA of Saccharomyces cerevisiae.

The ribosomal DNA (rDNA) repeats of Saccharomyces cerevisiae contain an autonomously replicating sequence (ARS) that colocalizes with a chromosomal origin of replication. We show that a minimal sequence necessary for full ARS function corresponds to a 107-bp rDNA fragment which contains three 10-of-11-bp matches to the ARS consensus sequence. Point mutations in only one of the 10-of-11-bp matches, GTTTAT GTTTT, inactivate the rDNA ARS, indicating that this consensus sequence is essential. A perfect match to a revised ARS consensus is present but not essential. Sequences up to 9 bp 5' from the essential consensus are dispensable. A broad DNA region directly 3' to the essential consensus is required and is easily unwound as indicated by: (i) hypersensitivity to nicking of an approximately 100-bp region by mung bean nuclease in a negatively supercoiled plasmid and (ii) helical instability determined by thermodynamic analysis of the nucleotide sequence. A correlation between DNA helical instability and replication efficiency of wild-type and mutated ribosomal ARS derivatives suggests that a broad region 3' to the essential ARS consensus functions as a DNA unwinding element. Certain point mutations that do not stabilize the DNA helix in the 3' region but reduce ARS efficiency reveal an element distinct from, but overlapping, the DNA unwinding element. The nucleotide sequence of the functionally important constituents in the ARS appears to be conserved among the rDNA repeats in the chromosome.     

Identification of an essential core element and stimulatory sequences in a Kluyveromyces lactis ARS element, KARS101

A Kluyveromyces lactis chromosomal sequence of 913 bp is sufficient for replication in Saccharomyces cerevisiae and K. lactis . This fragment contains a 12 bp sequence 5'-ATTTATTGTTTT-3' that is related to the S. cerevisiae ACS (ARS consensus sequence). This dodecamer was removed by site-directed mutagenesis and the effect on K. lactis and S. cerevisiae ARS (autonomous replicating sequence) activity was determined. The dodecamer is essential for S. cerevisiae ARS function but only contributes to K. lactis ARS activity; therefore, its role in K. lactis is unlikely to be the same as that of the essential S. cerevisiae ACS.
A 103 bp subclone was found to retain ARS activity in both yeasts, but the plasmid was very unstable in S. cerevisiae . Deletion and linker substitution mutagenesis of this fragment was undertaken to define the DNA sequence required for K. lactis ARS function and to test whether the sequence required for ARS activity in K. lactis and S. cerevisiae coincide. We found a 39 bp core region essential for K. lactis ARS function flanked by sequences that contribute to ARS efficiency. The instability of the plasmid in S. cerevisiae made a fine-structure analysis of the S. cerevisiae ARS element impossible. However, the sequences that promote high-frequency transformation in S. cerevisiae overlap the essential core of the K. lactis ARS element but have different end-points.     

Three domains in the simian virus 40 core origin orchestrate the binding, melting, and DNA helicase activities of T antigen.

The simian virus 40 (SV40) core origin of replication consists of three functional domains. The sequence 5'-CACTACTTCTGGAATAG-3' with an imperfect inverted repeat (underlined), a palindrome with four 5'-GAGGC-3' pentanucleotide repeats, and a 17-base-pair A + T-rich segment. We have been able to assign primary functions to each domain. Remarkably, SV40 large T antigen melted the inverted repeat domain in the complete absence of other origin sequences. Presumably, this protein-DNA interaction initiates a replication bubble that leads to daughter strand DNA synthesis. The pentanucleotide domain alone docked and arranged T antigen at the origin. The A + T-rich domain had no independent function, but, in the presence of the other two domains, allowed bound T antigen to extend the replication bubble. Thus, three domains of the origin coordinate the binding, melting, and DNA helicase activities of T antigen in an ordered sequence of events to initiate DNA replication.     

The yeast protein encoded by PUB1 binds T-rich single stranded DNA.

We have characterized binding activities in yeast which recognise the T-rich strand of the yeast ARS consensus element and have purified two of these to homogeneity. One (ACBP-60) is detectable in both nuclear and whole cell extracts, while the other (ACBP-67) is apparent only after fractionation of extracts by heparin-sepharose chromatography. The major binding activity detected in nuclear extracts was purified on a sequence-specific DNA affinity column as a single polypeptide with apparent mobility of 60kDa (ACBP-60). This protein co-fractionates with nuclei, is present at several thousand copies per cell and has a Kd for the T-rich single strand of the ARS consensus between 10(-9) and 10(-10) M. Competition studies with simple nucleic acid polymers show that ACBP-60 has marginally higher affinity for poly dT30 than for a 30 nt oligomer containing the T-rich strand of ARS 307, and approximately 10 fold higher affinity for poly rU. Internal sequence information of purified p60 reveals identity with the open reading frames of genes PUB1 and RNP1 which encode polyuridylate binding protein(s). The second binding activity, ACBP-67, also binds specifically to the T-rich single strand of the ARS consensus, but with considerably lower affinity than ACBP-60. Peptide sequence reveals that the 67kDa protein is identical to the major polyA binding protein in yeast, PAB1.     

Evidence for binding of at least two factors, including T-rich strand-binding factor(s) to the single-stranded ARS1 sequence in Saccharomyces cerevisiae.

Summary To study the mechanism of initiation of eukaryotic chromosomal replication, we examined protein factors interacting with the ARS1 region located near the centromere of chromosome IV in Saccharomyces cerevisiae. Using the gel shift assay, we found protein factor(s) which specifically bound to the T-rich strand of the region containing the core consensus and its flanking sequences in ARS1, but not to the opposite strand. We designated this factor ATS (ARS1, T-rich strand-binding factor(s)). Similar specific complexes were also detected with oligonucleotide probes specific for the H4 or C2G1 ARS. As we have previously identified another binding factor, we conclude that at least two factors bind to the single-stranded ARS1 sequence.     

Bending of DNA segments with Saccharomyces cerevisiae autonomously replicating sequence activity, isolated from basidiomycete mitochondrial linear plasmids

Summary Previous studies have indicated that DNA bending is a general structural feature of sequences (ARSs) from cellular DNAs of yeasts and nuclear and mitochondrial genomic DNAs of other eukaryotes that are capable of autonomous replication in Saccharomyces cerevisiae. Here we showed that bending activity is also tightly associated with S. cerevisiae ARS function of segments cloned from mitochondrial linear DNA plasmids of the basidiomycetes Pleurotus ostreatus and Lentinus edodes. Two plasmids, designated pLPO2-like (9.4 kb), and pLPO3 (6.6 kb) were isolated from a strain of P. ostreatus. A 1029 by fragment with high-level ARS activity was cloned from pLPO3 and it contained one ARS consensus sequence (A/T)TTTAT(A/G)TTT(A/T) indispensable for activity and seven dispersed ARS consensus-like (10/11 match) sequences. A discrete bent DNA region was found to lie around 500 by upstream from the ARS consensus sequence (T-rich strand). Removal of the bent DNA region impaired ARS function. DNA bending was also implicated in the ARS function associated with a 1430 by fragment containing three consecutive ARS consensus sequences which had been cloned from the L. edodes plasmid pLLE1 (11.0 kb): the three consecutive ARSs responsible for high-level ARS function occurred in, and immediately adjacent to, a bent DNA region. A clear difference exists between the two plasmid-derived ARS fragments with respect to the distance between the bent DNA region and the ARS consensus sequence(s).     

An autonomously replicating sequence of pSRI plasmid is effective in two yeast species, Zygosaccharomyces rouxii and Saccharomyces cerevisiae

The autonomously replicating sequences (ARSs) of pSR1, a cryptic circular DNA plasmid detected in a strain of Zygosaccharomyces rouxii, were delimited by subcloning and deletion analysis and by the isolation of nucleotide substitution mutations. A 30 base-pair (bp) sequence from inverted repeat 1 (IR1) and presumably the same region from IR2 of pSR1 functions as an ARS in the native host, Z. rouxii, and in a heterologous host, Saccharomyces cerevisiae. Thus, pSR1 has two ARSs per molecule, either of which is sufficient for replication of the plasmid molecule in both hosts. These hosts, however, respond differently to nucleotide substitutions in the 30 bp sequence, suggesting that the sequences required for ARS function in the two organisms are not exactly the same. In addition, a 137 bp sequence that overlaps the 30 bp sequence by 11 bp also functions as an ARS in Z. rouxii but not in S. cerevisiae. However, this 137 bp sequence enhances the stability of plasmids carrying the pSR1 ARS in S. cerevisiae. The 30 bp and 137 bp sequences each contain a single copy of the 11 bp ARS consensus sequence, which is essential for ARS function in S. cerevisiae. Small insertions between the 11 bp overlapping region and the 11 bp ARS consensus sequence showed that a proper distance between these two 11 bp sequences is essential for the ARS function of the 30 bp sequence. Point mutations that inactivate ARS function show that the ARS consensus sequence, as well as a short A:T segment in the overlapping sequence, is required for the ARS function of the 30 bp sequence.     

The ARS consensus sequence is required for chromosomal origin function in Saccharomyces cerevisiae.

Replication origins have been mapped to positions that coincide, within experimental error (several hundred base pairs), with ARS elements. To determine whether the DNA sequences required for ARS function on plasmids are required for chromosomal origin function, the chromosomal copy of ARS306 was deleted and the chromosomal copy of ARS307 was replaced with mutant derivatives of ARS307 containing single point mutations in domain A within the ARS core consensus sequence. The chromosomal origin function of these derivatives was assayed by two-dimensional agarose gel electrophoresis. Deletion of ARS306 deleted the associated replication origin. The effects on chromosomal origin function of mutations in domain A paralleled their effects on ARS function, as measured by plasmid stability. These results demonstrate that chromosomal origin function is a property of the ARS element itself.     

An ARS element from Drosophila melanogaster telomeres contains the yeast ARS core and bent replication enhancer.

We have sequenced the 0.7-kb-long fragment of Drosophila DNA which ensures the autonomous replication of plasmids in yeast. Deletion mapping has shown the ARS element to consist of at least two domains: the core having the consensus 11-bp sequence TAAATATAAAT and the enhancer which is no more than 90 bp long and is located at the 3'-end of the A-rich core strand. Neither domain per se ensures plasmid replication in yeast. A comparison of the enhancer sequence with the sequences of 14 different ARS elements failed to reveal significant homology areas. Most probably the ARS flanks that are adjacent to the core and act as enhancer do not carry any consensus. They may determine a peculiar structural feature of DNA (for example bends) which are necessary for the protein-ARS interaction.     

Multiple DNA elements in ARS305 determine replication origin activity in a yeast chromosome.

A yeast autonomously replicating sequence, ARS305, shares essential components with a chromosome III replicator, ORI305. Known components include an ARS consensus sequence (ACS) element, presumed to bind the origin recognition complex (ORC), and a broad 3'-flanking sequence which contains a DNA unwinding element. Here linker substitution mutagenesis of ARS305 and analysis of plasmid mitotic stability identified three short sequence elements within the broad 3'-flanking sequence. The major functional element resides directly 3' of the ACS and the two remaining elements reside further downstream, all within non-conserved ARS sequences. To determine the contribution of the elements to replication origin function in the chromosome, selected linker mutations were transplaced into the ORI305 locus and two-dimensional gel electrophoresis was used to analyze replication bubble formation and fork directions. Mutation of the major functional element identified in the plasmid mitotic stability assay inactivated replication origin function in the chromosome. Mutation of each of the two remaining elements diminished both plasmid ARS and chromosomal origin activities to similar levels. Thus multiple DNA elements identified in the plasmid ARS are determinants of replication origin function in the natural context of the chromosome. Comparison with two other genetically defined chromosomal replicators reveals a conservation of functional elements known to bind ORC, but no two replicators are identical in the arrangement of elements downstream of ORC binding elements or in the extent of functional sequences adjacent to the ACS.