Discrete Start Sites for DNA Synthesis in the Yeast ARS1 Origin

See allHide authors and affiliations

Science  02 Jan 1998:
Vol. 279, Issue 5347, pp. 95-98
DOI: 10.1126/science.279.5347.95


Sites of DNA synthesis initiation have been detected at the nucleotide level in a yeast origin of bidirectional replication with the use of replication initiation point mapping. The ARS1origin of Saccharomyces cerevisiae showed a transition from discontinuous to continuous DNA synthesis in an 18–base pair region (nucleotides 828 to 845) from within element B1 toward B2, adjacent to the binding site for the origin recognition complex, the putative initiator protein.

An origin of bidirectional DNA replication is characterized by the transition between continuous DNA synthesis (proceeding in one direction) and discontinuous synthesis (proceeding in the opposite direction). We have developed replication initiation point (RIP) mapping to determine this transition in the autonomously replicating sequence (ARS) 1 of the yeastSaccharomyces cerevisiae.

ARS1 functions as an origin of DNA replication (ORI) both on a plasmid and in its normal context on chromosome IV (1).ARS1-containing plasmids respond normally to the cell cycle, duplicating once per cycle (2), and replication is initiated by the same cellular protein machinery acting on chromosomes.

ARS1 is composed of subdomains A, B1, B2, and B3 (3). Subdomains A and B1 are recognized by the origin recognition complex (ORC) (4), the putative initiator protein (5) indispensable for origin function (6,7). Element B2 is easily unwound DNA (8) and element B3 is a binding site for the ARS binding factor I (ABFI) (9).

RIP mapping, described here, has sufficient sensitivity for study of eukaryotic origins, unlike an earlier method (10). It allows precise mapping of initiation sites for DNA synthesis and was applied to a 193–base pair (bp) fragment containing yeastARS1. We chose the wild-type ARS1 construct (pARS1/WTA) transformed into S. cerevisiae strain SP1 (approximately two copies per cell), also used for other analyses of origin function (3, 11). Nuclear DNA was isolated from asynchronously growing cultures, and replicative intermediate (RI) DNA was enriched by benzoylated naphthoylated DEAE (BND) cellulose column chromatography (12). RI DNA was treated with λ-exonuclease to eliminate nicked DNA (12) (Fig.1A). Nascent DNA strands with an attached RNA primer are resistant to λ-exonuclease digestion (13) and were subsequently used as template DNA to extend a primer to the junction with the RNA primer.

Figure 1

Application of RIP mapping to the SV40 origin (B), with the use of λ-exonuclease to degrade phosphorylated DNA (A). (A) Double-stranded DNA cleaved with Bgl I, phosphorylated or dephosphorylated, was incubated for 30 min with (+) or without (−) λ-exonuclease. Heat-denatured, single-stranded DNA orEscherichia coli tRNA was incubated for 12 hours. Lane M: Molecular size marker (in kilobases) is λ DNA cut with Hind III. (B) RIP mapping of the SV40 origin. RI DNA untreated (RI-λ) or digested with λ-exonuclease (RI) was used as template DNA. Sequencing reactions were prepared with dideoxyadenosine 5′-triphosphate (ddATP) (A), dideoxycytidine 5′-triphosphate (ddCTP) (C), dideoxyguanosine 5′-triphosphate (ddGTP) (G), or dideoxythymidine 5′-triphosphate (ddTTP) (T). Selected nucleotide positions (lane T) are numbered for orientation.

RIP mapping of simian virus (SV) 40 ORI (Fig. 1B) indicates essentially the same start sites as those previously identified by a different method (10). Multiple individual 5′ DNA ends represent multiple start sites for SV40 DNA synthesis. Blank areas on the gel reflect regions of continuous DNA synthesis. The nucleotide position at the 5′ end of the smallest detectable fragment marks the transition point between discontinuous and continuous DNA synthesis (Fig. 1B) and thus the site of leading strand initiation.

RIP mapping was next used to analyze ARS1. To demonstrate that the primers annealed specifically, we used double-stranded nonreplicating DNA (dsDNA) cleaved at a unique restriction site as the template for primer extension to the restriction site (Fig. 2). No endogenous pause sites were detected for primer extension on the top strand. However, when the bottom strand was analyzed, a second fragment sometimes appeared that we interpreted as an endogenous pause site because the band also occurred when the yeast ARS1 plasmid was replicated in bacteria, indicating that it was not linked to the replication process in yeast (14). When nascent DNA was analyzed with primers rev IV and −90 (Fig. 2, lanes RI), which annealed 100 nucleotides (nt) beyond the left and right borders ofARS1, respectively, multiple individual start sites for DNA synthesis were detected with a distinct transition from discontinuous to continuous initiation. We confirmed the start sites mapped on each strand using different primers (14). Thus, within the 400-bp region containing ARS1 and adjacent vector sequences, the transition point for the top strand mapped to nt 845 and for the bottom strand to nt 828 (Fig. 3B). Hence, there is an 18-bp transition region (including the bands at each end), which partly overlaps the ORC binding site. The polyomavirus origin also shows a transition zone of 18 bp (15), but the SV40 origin transition occurs within 2 to 3 nt (10) (Fig. 3A).

Figure 2

RIP mapping of the yeast ARS1 origin. (A) Analysis of the top and (B) bottom strand. RI DNA from SP1/pARS1/WTA digested with λ-exonuclease was used as template DNA. Double-stranded, nonreplicating yeast DNA cleaved with Eco RI (ds/E) or Hind III (ds/H) served as a control. Sequencing reactions with ddATP (A), ddCTP (C), ddGTP (G), or ddTTP (T) are shown. The position of a polymerase pause site at nt 845 seen with primer −90 is indicated by an asterisk. tp, transition point.

Figure 3

Map of start sites detected for DNA synthesis. DNA 5′ ends detected as start sites for DNA synthesis in this study are indicated by arrows. Stronger bands on the gels are shown by longer arrows and weaker bands by shorter arrows. (A) SV40 origin (10). The striped bar and asterisks indicate the transition point and DNA start sites, respectively, previously mapped by Hay and DePamphilis (10). DUE, DNA unwinding element. (B) YeastARS1. DNase I–hypersensitive sites mapped by genomic footprinting and in vitro (4) are shown below the map, with the most pronounced site indicated by an asterisk.

Additional evidence that the initiation sites we observed were linked to the process of replication was provided by analyzing nascent strands derived from an inactive copy of ARS1, carrying a mutation in element A, that is replicated through function of an active wild-type ARS1 on the same plasmid (16). If replication forks generated by the wild-type ARS1 origin proceed at the same rates in both directions, they should meet at approximately nt 2900 of the construct pARS/A&WT (Fig.4A). DNA replication through the inactiveARS1/A would be discontinuous on the bottom strand and continuous on the top strand. The replication initiation point patterns determined by RIP mapping were consistent with this model (Fig. 4B, lanes A). No primer extension products smaller than several kilobases were detected on the top strand, whereas multiple start sites were observed throughoutARS1/A on the bottom strand (Fig.4B).

Figure 4

(A) Scheme of the pARS/A&WT construct and replication events. Replication initiates within ARS/WT. Leading strands are shown as bold, semicircular arrows. Lagging strands synthesized by multiple Okazaki fragments are represented by short arrows. (B) RIP mapping of the inactive mutant ARS/A copy. Details are as in Fig. 2. RI DNA was isolated from SP1/pARS/A&WT cells (A) as well as from SP1/pARS1/WTA cells (WT). Open circles indicate the positions of the fragments in lane A.

The distance between ARS1 sites most frequently used (longest arrows in Fig. 3B) was ∼20 to 50 nt. However, Okazaki fragments in higher eukaryotes average over 100 nt (17,18). We analyzed the length of nascent strands with intact bi- or triphosphorylated RNA primers in a labeling reaction with the RNA-specific “capping enzyme” guanylyltransferase (19). Labeled fragments of 20 to 35, 40 to 50, 60, 70 to 80, and 125 nt accumulated preferentially (Fig.5), similar to the size distribution of nascent strands in SV40 DNA (20) and Drosophila(21). Therefore, this seems to be a conserved feature among eukaryotes. Whether this size distribution is due to DNA polymerase δ pauses or reflects a discontinuous mechanism underlying the formation of Okazaki fragments (“discontinuity model”) (20) remains to be determined.

Figure 5

Size distribution of nascent strands in yeast. (p)ppRNA-DNA chains from RI DNA and dsDNA and from total nuclear yeast DNA were radiolabeled with [α-32P]GTP with the use of vaccinia guanylyltransferase (19). The label was removed by alkali treatment, indicating that the label was on RNA primers (14). The double-stranded fraction contains “full-length” Okazaki fragments of 125 nt that are not efficiently bound by BND cellulose. “Full-length” Okazaki fragments of 125 nt are absent in RI DNA, as they are already processed or ligated to leading strands. Lane M: Molecular size marker (in bases) is 0X174 RF DNA cut with Hae III.

What accounts for the multiple initiation sites mapped in this study and for SV40 (10) and polyomavirus (15)? It is likely that most of these sites reflect positions where discontinuous (lagging strand) synthesis initiates, and individual replicating molecules might choose different sites to initiate DNA synthesis, resulting in population polymorphism. However, we cannot determine whether some are alternate start sites for continuous (leading strand) synthesis. Indeed, the primer extension product that maps to the transition point was not the strongest band for the SV40 origin or for yeast ARS1, as would have been expected if it represented the only leading strand initiation site used by every replicating molecule.

An origin of bidirectional replication can be defined as a cis-acting sequence upon which the trans-acting replication machinery assembles. An indispensable part of this machinery is the initiator protein (5). In SV40 the transition region is close to the binding site for the viral initiator protein, large tumor (T) antigen, and adjacent to a DNA unwinding element (DUE) (22) (Fig.3A). Similarly, the transition region in ARS1 is flanked by the binding site of ORC, the putative initiator protein, and by element B2 (8) (Fig. 3B). Mutations in B2 of ARS1 reduce replication efficiency (3), whereas mutations in the transition region, between B1 and B2, do not affect replication efficiency in vivo (3). This suggests that the transition region itself has no cis-regulatory function.

The initiation points we detected in the transition region coincide with deoxyribonuclease I–hypersensitive sites that are exposed on each strand of ARS1 upon ORC binding in vivo and in vitro (4) (Fig. 3). The most pronounced hypersensitive site in element B1 (4) is at the transition point on the top strand. The coincidence of ORC-induced hypersensitive sites with DNA initiation sites suggests that ORC defines the transition region.

  • * To whom correspondence should be addressed. E-mail: susan gerbi{at}


Stay Connected to Science

Navigate This Article