Conservation of ARS Elements and Chromosomal DNA Replication Origins on Chromosomes III of Saccharomyces cerevisiae and S. carlsbergensis
Chen Yang, James F. Theis, Carol S. Newlon


DNA replication origins, specified by ARS elements in Saccharomyces cerevisiae, play an essential role in the stable transmission of chromosomes. Little is known about the evolution of ARS elements. We have isolated and characterized ARS elements from a chromosome III recovered from an alloploid Carlsberg brewing yeast that has diverged from its S. cerevisiae homeologue. The positions of seven ARS elements identified in this S. carlsbergensis chromosome are conserved: they are located in intergenic regions flanked by open reading frames homologous to those that flank seven ARS elements of the S. cerevisiae chromosome. The S. carlsbergensis ARS elements were active both in S. cerevisiae and S. monacensis, which has been proposed to be the source of the diverged genome present in brewing yeast. Moreover, their function as chromosomal replication origins correlated strongly with the activity of S. cerevisiae ARS elements, demonstrating the conservation of ARS activity and replication origin function in these two species.

LITTLE is known about the evolution of noncoding elements of chromosomes. One class of these noncoding elements, the origins of DNA replication, is well characterized in the budding yeast, Saccharomyces cerevisiae. They are dependent upon ARS elements, sequences of 100–200 bp that were identified by their ability to promote the extrachromosomal maintenance of plasmids (reviewed by Newlon 1996). As is the case for centromeres, fission yeast ARS elements are larger than those of S. cerevisiae, and replication appears to initiate in broad zones in metazoans (reviewed by DePamphilis 1996).

ARS elements have been identified systematically on chromosome III (Newlonet al. 1993; A. Poloumienko, A. Dershowitz, J. Shah and C. S. Newlon, unpublished results), chromosome V (Tanakaet al. 1996), chromosome VI (Shirahigeet al. 1993), and in a 140kb region of chromosome XIV (Friedmanet al. 1995). Many, but not all, of the ARS elements function detectably as chromosomal replicators (Deshpande and Newlon 1992; Greenfeder and Newlon 1992b; Rivier and Rine 1992; Zhuet al. 1992; Newlonet al. 1993; Friedman et al. 1995, 1997; Theis and Newlon 1997; Yamashitaet al. 1997).

Several observations make the question of whether the positions of replication origins are conserved an interesting one. Brewer (1988) has noted that bacterial and viral genomes are organized so that replication forks traverse heavily transcribed genes in the same direction as RNA polymerase, suggesting that there may be constraints on the location of replication origins in chromosomes. However, the analysis of origin deletions on chromosome III has shown that the direction of fork movement has little or no effect on chromosome stability (Dershowitz and Newlon 1993; Newlonet al. 1993; A. Dershowitz and C. S. Newlon, unpublished results). Moreover, the observation that deletion of several origins had no effect on chromosome stability suggests that more origins are present than are needed to replicate the chromosome and that some could be lost during evolution. One explanation of inactive replication origins is that they represent origins that are active in other species, but no longer function in S. cerevisiae. Alternatively, the inactive or inefficient ARS elements might represent fortuitous binding sites for the replication initiator protein, origin recognition complex (ORC; reviewed by Newlon 1996) that allows plasmid, but not chromosomal, replication origin activity. A third explanation is that inactive ARS elements are used for some function other than replication. Two inactive origins, ARS301 and ARS302, are associated with the silent mating type locus, HML, where they function as transcriptional silencers. ORC clearly plays a role in silencing because mutations in two different subunits of the complex result in derepression of HML (Fosset al. 1993; Looet al. 1995). In addition, mutations in ORC have been found that separate its role in silencing from its role in DNA replication, indicating that the two functions are independent (Foxet al. 1995).

As a first approach toward addressing these questions, we chose to isolate and characterize ARS elements from a chromosome III of a brewing yeast that has diverged from S. cerevisiae. The brewing yeast is an alloploid that carries genomes from two different Saccharomyces species: one is like S. cerevisiae; the other, referred to as S. carlsbergensis, has diverged significantly at the DNA sequence level. Six chromosomes of the brewing strain have been transferred into S. cerevisiae (reviewed by Kielland-Brandtet al. 1995). Two or more copies of each chromosome were recovered. For some chromosomes, one copy was indistinguishable from S. cerevisiae, and the other was S. carlsbergensis-like, hybridizing with S. cerevisiae probes at low stringency and having a different restriction map. Other chromosomes were mosaic, having segments from both S. cerevisiae and S. carlsbergensis, suggesting that they were products of rare recombination events between the homeologous chromosomes in the alloploid brewing strain.

Figure 1.

—Physical maps of chromosomes III of S. cerevisiae and S. carlsbergensis: All numbers represent ARS elements on the chromosome. A “+” following the number denotes that the ARS element functions as a chromosomal replication origin, while a “–” indicates that it does not. The region from MAT to the right telomere is homologous between S. cerevisiae and S. carlsbergensis, i.e., the sequences are similar, and recombination occurs between the two chromosomes, while the region from the left telomere to MAT is homeologous, i.e., the sequences have diverged and there is little recombination between the two chromosomes. Despite the sequence divergence, all the essential genes in this region are functional, as indicated by the fact that the S. carlsbergensis chromosome III can substitute for its S. cerevisiae counterpart. The homeologous region is shaded gray.

The S. carlsbergensis chromosome III can substitute for the S. cerevisiae chromosome III (Nilsson-Tillgrenet al. 1981). The S. carlsbergensis chromosome III is actually a mosaic chromosome (Figure 1). The region between the left telomere and the MAT locus, in the middle of the right arm, presumably was derived from the S. carlsbergensis genome and has diverged from the corresponding region in S. cerevisiae sufficiently that meiotic recombination between the homeologues is reduced at least 100-fold (Nilsson-Tillgrenet al. 1981; Priebeet al. 1994). Despite these DNA sequence differences, equivalent genes appear to be present in the same order on both chromosomes. In contrast, the region distal to MAT on the right arm is homologous to S. cerevisiae with normal levels of meiotic recombination (Nilsson-Tillgrenet al. 1981).

In this article, we present data demonstrating that ARS elements are located in the brewing yeast chromosome within the same intergenic intervals that they occupy in S. cerevisiae. Moreover, their ability to function as replication origins is conserved both in S. cerevisiae and in S. monacensis, which has been proposed to be the source of the S. carlsbergensis genome in the Carlsberg brewing strain (Hansen and Kielland-Brandt 1994).


Strains and plasmids: The S. cerevisiae strain YP45 (Sikorski and Hieter 1989) was utilized for transformations to examine ARS activity. The S. carlsbergensis chromosome III, initially recovered from Carlsberg production strain 244, was maintained in the S. cerevisiae chromosome III substitution strain M1253 (C80-1253; Nilsson-Tillgrenet al. 1981). The S. monacensis type strain was obtained from the American Type Culture Collection (#76670, an isolate of CBS1503). The Escherichia coli strains JA226 (Devenish and Newlon 1982) and DH5α (Life Technologies, Grand Rapids, NY) were used for routine plasmid manipulations. The shuttle vector pRS306 (Sikorski and Hieter 1989) was used for constructing of S. carlsbergensis chromosome III libraries and for subcloning the ARS elements.

A plasmid pRS333 was constructed to test for ARS activity in S. monacensis. CEN3carl and a G418 resistance cassette were inserted into the S. cerevisiae-E. coli shuttle vector, pRS303 (Sikorski and Hieter 1989). The 0.5-kb XbaI-PstI fragment of 19EG1, containing the centromere (T. H. Andersen and T. Nilsson-Tillgren, personal communication), was treated with Mung Bean Nuclease (New England Biolabs, Beverly, MA) and cloned into pFA-KanMX4 (Wachet al. 1994) that had been digested with BglII and treated with Mung Bean Nuclease to generate blunt ends. This construct was digested with AscI plus SacI, then treated with T4 polymerase (New England Biolabs) plus dNTPs; the 1.9-kb fragment containing the centromere and the kanamycin-resistant cassette was isolated. This fragment was ligated to pRS303, which had been digested with AatII and treated with T4 DNA polymerase and dNTPs to generate blunt ends. The plasmid pRS333 was recovered in which the kanamycin-resistant cassette is transcribed in the same direction as the ampicillin-resistant gene.

The ARS-containing fragments were moved into pRS333 as follows. The plasmids p305K/D0.55, p307k/s-2′/2, and p310Bg/B0.56 were digested with FspI, and the ARS-containing fragments ligated to the 4.7-kb FspI fragment of pRS326 (Theis and Newlon 1994). The ARS305carl- and ARS310carl-containing plasmids were digested with BsaHI plus NgoMI, and the ARS-containing fragments ligated to the 4.1-kb BsaHI-NgoMI fragment of pRS333. The ARS307carl-containing plasmid was digested with BsaI plus NgoMI, and the ARS-containing fragment was ligated to the 4.5-kb BsaI-NgoMI fragment of pRS333.

Construction of S. carlsbergensis chromosome III libraries: S. carlsbergensis chromosome III was separated from other chromosomes by pulsed-field gel electrophoresis. Plugs from the substitution strain M1253 were prepared (Roseet al. 1990), and electrophoresis performed in a transverse alternating field electrophoresis apparatus (Beckman, Fullerton, CA) using the following conditions: 1% low-melting-point agarose (FMC, Rockland, ME) in 10 mm Tris, 0.4 mm EDTA, 0.025% (v/v) acetic acid; 30 min at 200 mA switching at 4-sec intervals followed by 20 hr at 220 mA switching at 15-sec intervals. A slice containing chromosome III was excised from the gel. The DNA was digested in situ with restriction enzymes prior to isolation by β-agarase (New England Biolabs) digestion. The resulting BamHI or EcoRI fragments were cloned in polylinker of pRS306. Ampicillin-resistant colonies were picked to form the BamHI and EcoRI libraries.

A BglII library was constructed for the isolation of the S. carlsbergensis ARS element corresponding to ARS309. Genomic DNA from M1253 was digested with BglII and separated on a 0.7% agarose gel; a slice containing fragments of 4.0–4.5 kb was excised, and the DNA eluted and then cloned in BamHI-digested pRS306. The ampicillin-resistant transformants represent the BglII library.

Screening of S. carlsbergensis chromosome III libraries for ARS-containing clones: A total of 40–48 BamHI library clones were grown in patches on an LB-Amp plate overnight; the cells were pooled, and plasmid DNA prepared. The plasmid DNA pools were then used to transform S. cerevisiae strain YP45. The stabilities of the plasmids were examined by streaking transformants onto a nonselective plate, growing overnight at 30°, then replica plating onto a Ura plate. ARS-bearing plasmids from yeast transformants that gave >80% Ura colonies after nonselective growth were characterized further.

Plasmid DNA was recovered from S. cerevisiae by an alkaline lysis procedure adapted from Birnboim and Doly (1979). Cells from a large yeast colony (about 3 mm in diameter) were washed in 1 ml SCE (1 m sorbitol, 0.1 m sodium citrate, 60 mm sodium EDTA), then resuspended in 200 μl of SCE solution with 0.1% (v/v) β-mercaptoethanol and 0.5 mg/ml zymolyase (U.S. Biologicals, Swampscott, MA). After digestion at 37° for about an hour, 400 μl of 0.2 n NaOH/1% SDS was added, and then samples were put on ice for 5 min. A total of 300 μl of precooled 2.7 m KOAc, pH 4.8, was added, and the samples left on ice for another 5 min. The supernatants were then collected after a 1-min spin in a microfuge. The DNAs were precipitated by adding 0.6 volumes of isopropanol. The DNA pellets were washed with 70% ethanol, dried, and dissolved in TE. These DNAs were used to transform E. coli strain JA226. Plasmids from ampicillin-resistant transformants were analyzed by restriction digestion. ARS activity was confirmed by transformation of yeast strain YP45.

The S. carlsbergensis libraries were also screened by hybridization to S. cerevisiae chromosome III probes. Different ARS-containing fragments (see legend of Figure 3) were used to probe the S. carlsbergensis BamHI, EcoRI, and BglII libraries at low stringency. Positive clones were recovered and tested for ARS activity by transformation of yeast strain YP45.

The Ars+ plasmids were labeled with [α-32P]dCTP (Multiprime kit; Amersham, Arlington Heights, IL) and hybridized to several blots: (1) pulsed-field gel-separated chromosomes from S. cerevisiae strain YP45 and the substitution strain M1253, (2) phage plaques of the ordered S. cerevisiae chromosome III λ library (Rileset al. 1993; low stringency), and (3) BamHI-, EcoRI-, and/or HindIII-digested DNA of S. cerevisiae strain YP45 (high stringency). The results allowed us to identify the regions of S. cerevisiae chromosome III to which the S. carlsbergensis ARS elements were homologous.

Hybridization conditions: High-stringency hybridizations were performed at 65° in 5× SSC, 0.1% SDS, 1× Denhardt's solution, 100 μg/ml sonicated calf thymus DNA. Blots were washed at 65° in 1× SSC, 0.1% SDS. Low-stringency hybridizations of the libraries were performed at 55° in 5× SSC, 1× Denhardt's solution. Filters were washed in 5× SSC at 55°.

Yeast transformation: S. cerevisiae cells were transformed with plasmid DNA using the lithium acetate procedure described by Elble (1992). S. monacensis cells were transformed by electroporation. Cells were grown in YM medium (Difco) to a density of ∼3.0 × 107/ml. Cells were prepared for electroporation as described by Becker and Guarente (1991). Approximately 100 ng of plasmid DNA and 2 μg of sonicated calf thymus DNA (Sigma, St. Louis) were added to 50 μl of cell suspension. Electroporation was carried out at 1.43 kV, 200 ohms, and 25 μF in a 0.2-cm cuvette. After electroporation, cells were resuspended in 1 ml of 1 m sorbitol and incubated in a roller drum for 3–5 hr prior to spreading on YM plates containing 50 μg/ml G418 (Geneticin, Life Technologies, Inc.). Plates were incubated at 23°. For Ars+ plasmids, 500–2000 transformants were obtained; pRS333 gave no transformants.

Subcloning and sequencing analysis: The ARS-containing clones isolated from the S. carlsbergensis chromosome III libraries were subcloned in pRS306, using either convenient restriction sites or limited Exonuclease III digestion (Henikoff 1987) to generate smaller clones. For each ARS element, the smallest subclone was sequenced (cycle sequencing kit, Life Technologies, Inc.). If the smallest ARS-containing subclone did not contain an open reading frame (ORF) to anchor it to the S. cerevisiae chromosome III sequence, the ends of larger clones were sequenced to detect open reading frames. Sequencing primers were provided by the New Jersey Medical School Molecular Resources Facility and Life Technologies, Inc. Sequence analysis was performed using the Wisconsin package (Genetics Computer Group, Inc.) provided by UMDNJ Academic Computing Services.

Analysis of replication intermediates by 2-D gel electrophoresis: All the ARS elements isolated from the S. carlsbergensis chromosome III were examined for their chromosomal replication origin activity by the 2-D gel electrophoresis method of Brewer and Fangman (1987). S. monacensis cells were harvested at a density of 2–4 × 107 cells/ml. DNA was prepared as described by Theis and Newlon (1994). The fragments labeled to detect replication intermediates were the 1.5-kb EcoRI from 4EH2 (ARS304carl), the 1.3-kb DraI fragment of D/D1.3 (ARS305carl), the 3.9-kb XbaI-BamHI from Xb/B3.9 (ARS306carl), the 2.2-kb HindIII from H/H2.2 (ARS307carl), the 3.0-kb BamHI-EcoRI from 19EG1 (ARS308carl), the 3.4-kb EcoRI-XhoI from –14 (ARS309carl), and the 1.0-kb BamHI from pB5B1(ARS310carl).


Identification of S. carlsbergensis ARS elements: To identify ARS elements from the S. carlsbergensis chromosome III, a BamHI library enriched for fragments of the gel-purified chromosome was constructed in the integrating shuttle vector pRS306 (Sikorski and Hieter 1989) as described in materials and methods. DNA from pools of 40–48 clones was screened for plasmids that transformed S. cerevisiae at high frequency, yielding unstable transformants. Using this strategy, 35 independent ARS-containing plasmids were identified. The chromosomal locations of these ARS elements were determined by Southern blot analysis (as described in materials and methods). Four different S. carlsbergensis chromosome III ARS elements were identified. Two were from the diverged region of the chromosome, one hybridizing at or near ARS305 and the second at or near ARS310 (Figure 2). The other two corresponded to ARS315 and ARS316 (A. Dershowitz, A. Polou-mienko, J. Shah and C. S. Newlon, unpublished results), which are located in the S. cerevisiae-like part of the mosaic chromosome, and were not analyzed further.

Figure 2.

—Genomic Southern hybridization. Southern blots of genomic DNA from strains M1253 and YP45 digested with BamHI, EcoRI, or HindIII were hybridized with the S. carlsbergensis ARS-containing fragments (A) 1B3a and (B) pB5B1 at high stringency. The divergence of the S. carlsbergensis sequences is indicated by the fact that the pB5B1 probe fails to hybridize with S. cerevisiae sequences and that the 1B3a probe hybridizes only weakly (arrows, Figure 2A).

This preliminary screening suggested that the chromosomal locations of ARS elements are conserved between the two homeologous chromosomes III, enabling further screening for S. carlsbergensis ARS elements using specific S. cerevisiae chromosome III ARS-containing fragments as probes to hybridize at low stringency to the S. carlsbergensis chromosome III libraries. Plasmids were identified that hybridized to ARS304, ARS306, ARS307, ARS308, and ARS309. Each of these clones transformed S. cerevisiae at high frequency, yielding transformants that carried unstable plasmids. These results indicated that each of these plasmids carries one or more ARS elements. To distinguish the S. cerevisiae ARS elements from their S. carlsbergensis homeologues, we employ the superscript “carl”; e.g., the ARS304 homeologue is ARS304carl.

To further define the locations of the new ARS elements, the S. carlsbergensis inserts in these plasmids were subcloned and tested for ARS activity (Figure 3); a single ARS element was present in each of the initial plasmids. Analysis of the DNA sequences of the subclones revealed that the S. carlsbergensis ARS elements are located within intergenic regions corresponding to their S. cerevisiae counterparts. For example, the ARS305carl is downstream of an ORF that is 60% identical to the C-terminal 112 aa of YCL049C, just as ARS305 is downstream of YCL049C (Figure 3B). ARS307carl is between ORFs homologous to PEL1 (76% identity over 78 amino acids) and YCL005W (82% identity over 34 aa), just as ARS307 is located between PEL1 and YCL005W. ARS310carl is upstream of an ORF 68% identical to the predicted N-terminal 106 aa of the YCR026C ORF (Figure 3G), placing it in the intergenic region that corresponds to ARS310.

The DNA sequences of the remaining four S. carlsbergensis ARS elements failed to reveal the presence of ORFs. Therefore, the ends of larger clones were sequenced to find homologies to ORFs that would allow alignment with S. cerevisiae DNA sequences. The ends of the ARS306carl clone, Xb/B3.9, are homologous to FUS1 (83% identity over 216 bp) and YCL025C (77% identity over 232 bp), placing ARS306carl in a location similar to ARS306 (Figure 3C). Similarly, ARS309carl is in a position analogous to ARS309 (Figure 3F), and ARS304carl is in the same position as ARS304. CEN3carl (located within clone H/H1.1) is 63% identical to CEN3 over 158 bp (T. H. Anderson and T. Nilsson-Tillgren, personal communication). While ARS308 is closely associated with CEN3 (Greenfeder and Newlon 1992a), ARS308carl is found in the adjacent fragment (Figure 3E).

Analysis of chromosomal replication origin activity: Only about two-thirds of the ARS elements on S. cerevisiae chromosome III function detectably as replication origins in their natural locations (Newlonet al. 1993). Therefore, we examined the replication intermediates of chromosomal DNA fragments carrying the S. carlsbergensis ARS elements using the 2-D gel electrophoresis technique developed by Brewer and Fangman (1987). If the fragment being examined lacks a replication origin, it is replicated by a fork originating outside the fragment, and the replication intermediates are Y-shaped and migrate in the pattern labeled “Y arc” in Figure 4. If the fragment contains a replication origin, replication intermediates are bubble shaped and migrate more slowly than Y-shaped intermediates in the second dimension, falling along the arc labeled “bubble” in Figure 4. If the origin is asymmetrically located within the fragment, a composite pattern is produced in which early replication intermediates are bubble shaped and late intermediates are Y-shaped. The detection of bubble-shaped intermediates indicates that the fragment contains an active replication origin. If a particular origin is active in every cell in the culture, then no small Y-shaped intermediates should be detected. If the origin is active in only a fraction of the cells, then some of the time the fragment is replicated by a fork entering the fragment from outside, resulting in the presence of Y-shaped intermediates of all sizes.

Figure 3.

—Subcloning of S. carlsbergensis ARS elements. The top line of each panel represents the physical map of S. cerevisiae chromosome III. On the map line, arrows represent open reading frames, while boxes denote the smallest ARS-containing restriction fragments, the recombination enhancer (RE; Wu and Haber 1995), and the centromere (CEN3); names are written underneath the line. The position of the essential ARS consensus sequence (ACS) of each ARS element is indicated by a bar or bars within the box; similarly, the bar within the CEN3 box denotes CDEI, CDEII, and CDEIII. Below the map lines, each box represents an S. carlsbergensis restriction fragment. The shaded boxes represent Ars+ fragments, while the open boxes represent Ars fragments. The dark shading denotes the regions which have been sequenced. The dark bars between the boxes represent the boundaries of the adjacent subclones. The capital letters used in the name of the subclones show the restriction sites used for cloning: B, BamHI; Bg, BglII; D, DraI; H, HindIII; K, KpnI; N, NcoI; R, EcoRI; S, SpeI; Sa, SacI; Sn,SnaBI; Xb, XbaI; and Xh, XhoI. All the maps are drawn to the same scale. For simplicity, only the elevant restriction sites are shown; in most cases, there are additional sites for the enzymes that are not shown. Numbering is based on the chromosome III sequence (Munich Information Center for Protein Sequences, 02.09.96 release) (A) ARS304carl. Clones 8BH2 and 4EH2 were identified in the BamHIand EcoRI libraries, respectively, using the indicated 5.3 kb BamHI fragment as a probe. (B) ARS305carl. Clone 1B3a was identified in the BamHI library using the indicated 6.2-kb BamHI fragment. (C) ARS306carl. 5BA1 was obtained from the BamHI library using the 3.2-kb HindIII/XbaI fragment as a probe. (D) ARS307carl. 16BD1 was obtained by probing the BamHI library with the 1.3-kb EcoRI/PstI fragment. (E) ARS308carl. 22BB2 and 19EG1 were pulled out of the BamHI and EcoRI libraries, respectively, by probing with the 3.4 kb BamHI fragment. (F) ARS309carl. Clone 14 was obtained from the BglII library probed with the 2.7-kb BglII/EcoRI fragment. (G) ARS310carl. pB5B1 was an Ars+ clone obtained from the BamHI library.

We first examined the origin activity of the S. carlsbergensis ARS elements in the chromosome III substitution strain M1253. Bubble-shaped replication intermediates were detected in the 2-D gels for ARS305carl, ARS306carl, ARS307carl, ARS309carl, and ARS310carl (Figure 4, panels 305, 306, 307, 309, and 310), indicating that these ARS elements function as chromosomal replication origins during normal cell growth. In addition to the bubble arc, a complete Y arc of uniform intensity was present on the 2-D gels for ARS307carl, ARS309carl, and ARS310carl, suggesting that these origins are active in only a fraction of the population. In contrast, the 2-D gels for ARS305carl and ARS306carl showed an intense signal on the late part of the Y arc with little signal from small Y-shaped intermediates, suggesting that these two origins are active in most or all of the cells in the population (Figure 4, compare panels 305 and 306 to panels 307, 309, and 310). ARS304carl and ARS308carl did not function detectably as chromosomal replication origins because only Y-shaped intermediates were detected (Figure 4, panels 304 and 308). A conspicuous feature of the 2-D analysis of ARS308carl is the presence of an intense spot of hybridization along the Y arc, mapping to the position of the centromere (Figure 4, panel 308), and reflecting a replication-fork pause site similar to the pause site coincident with CEN3 (Greenfeder and Newlon 1992a).

Analyses of S. monacensis: The preceding analyses were performed in an S. cerevisiae background. We were concerned that the replication of the S. carlsbergensis chromosome by S. cerevisiae proteins might make use of different replicator sequences than are used in the species from which the chromosome was derived. Indeed, studies of Schizosaccharomyces pombe and Kluyveromyces lactis have shown that the DNA sequences required for ARS activity differ from those required in S. cerevisiae (Maundrell et al. 1985, 1988; Fabianiet al. 1996). The obvious choice for these experiments is the species that contributed the S. carlsbergensis set of chromosomes to the brewing strain, which seems to be S. monacensis (Hansen and Kielland-Brandt 1994).

Figure 4.

—2-D gel analysis in the chromosome III substitution strain. Each of the S. carlsbergensis ARS elements was analyzed for origin function using DNA prepared from M1253. For each ARS element, the restriction fragment examined is indicated. The position of the ARS within the fragment is denoted by the dot, and the fragment used as a probe is indicated by the box below the line. The lower right panel is a diagram of the migration patterns of bubble- and Y-shaped replication intermediates. The presence of bubble-shaped intermediates indicates that a fragment contains a replication origin.

Using primers designed from our sequences of ARS-305carl and ARS306carl, we amplified by PCR the homologues from S. monacensis. The sequences of the PCR products were identical to the sequences obtained from our clones (Theiset al. 1999, and data not shown), supporting the idea that the S. carlsbergensis chromosomes are derived from S. monacensis. On the basis of this observation, we decided to analyze the function of the S. carlsbergensis ARS elements in S. monacensis. Because S. monacensis is a prototroph, we constructed a plasmid that carries CEN3carl and KanMX, encoding G418 resistance (see materials and methods). Three ARS elements were inserted into this plasmid and examined for ARS activity: ARS305carl-fragment D/Sn0.3, ARS-307carl-fragment k/s-2′/2, and ARS310carl-fragment Y. All three of these constructs yielded G418-resistant colonies, while the vector pRS333 did not. The plasmids were unstable in the S. monacensis transformants, and, in the case of the ARS310carl transformants, Southern blot analysis showed that the plasmid had not integrated into one of the S. monacensis chromosomes (data not shown). We conclude that these three ARS elements are able to function in both S. monacensis and S. cerevisiae.

We examined replication origin activity in S. monacensis (Figure 5). ARS304carl was inactive while ARS307carl and ARS309carl were active, though not in every cell cycle as shown by the Y arc of uniform intensity (compare panels in Figure 4 with those in Figure 5). ARS305carl and ARS306carl appeared to function less efficiently as chromosomal replication origins in S. monacensis, as indicated by the uniform intensity of the Y arc detected in Figure 5. ARS308carl was weakly active as an origin in S. monacensis, as indicated by the faint bubble arc, while it was inactive in the substitution strain (Figure 5, panel 308).

Figure 5.

—2-D gel analysis in S. monacensis. Origin function was examined using DNA isolated from S. monacensis cells. Restriction digests and probes are the same as shown in Figure 4.


We have isolated nine ARS elements from the S. carlsbergensis chromosome III. The first approach was to screen a plasmid library enriched for clones from the S. carlsbergensis chromosome. The advantage of this approach is that it requires no assumptions about the positions of ARS elements on the chromosome. Of the 35 independent Ars+ clones obtained in the screen of the BamHI library, 18 were derived from chromosome III and represented four different ARS elements. Two of these were derived from the S. cerevisiae region of the chromosome and were not studied further. ARS305carl was isolated twice, and ARS310carl was isolated four times. While this screen was certainly not saturated, it was relatively inefficient. Moreover, it showed that ARS elements are located in analogous positions on the two chromosomes. To supplement the data from the initial screen, five more ARS elements were identified by hybridization to the S. carlsbergensis libraries at reduced stringency using probes containing S. cerevisiae ARS elements.

The analysis of subclones indicated that each of the clones obtained from the S. carlsbergensis libraries contained a single ARS element. These seven ARS elements were located between homologous ORFs on the two chromosomes, despite the fact that the large clones often contained many intergenic regions. Assuming the organization of ORFs is identical on the two chromosomes, only 7 of the 27 intergenic regions present on the seven S. carlsbergensis clones contain ARS elements. The precise correspondence of the S. carlsbergensis ARS elements to the only ARS elements in a 150-kb segment of chromosome III (nucleotides 30,000–180,000; Newlonet al. 1991) makes it unlikely that the use of the directed strategy missed one or more ARS elements in the corresponding region of the S. carlsbergensis chromosome.

We addressed the ability of these ARS to function as chromosomal replication origins by examining replication intermediates in both the chromosome III substitution strain and S. monacensis. In S. cerevisiae, ARS305, ARS306, ARS307, ARS309, and ARS310 are all strong origins, i.e., active in most cells (Deshpande and Newlon 1992; Greenfeder and Newlon 1992b; Zhuet al. 1992; Newlonet al. 1993; Theis and Newlon 1997). The S. carlsbergensis counterparts of these ARS elements are also active as chromosomal replication origins, although their levels of activity appear to differ. For example, ARS305carl and ARS306carl appear to be active in most S. cerevisiae cells, as indicated by the absence of small Y-shaped intermediates (Figure 4), yet these origins appear to be less efficiently used in S. monacensis, as suggested by the uniform intensity of the Y arc present below the bubble arc (Figure 5). ARS309carl and ARS310carl appear to be partially active in both species, as indicated by the similar intensities of both bubble- and Y-shaped intermediates (Figures 4 and 5). ARS304carl is not active in either case, nor is ARS304 an active origin in S. cerevisiae (Dubeyet al. 1991; Newlonet al. 1993). ARS308carl does not appear to be used as an origin in the chromosome III substitution strain, but is weakly active in S. monacensis, just as ARS308 is a weak origin in S. cerevisiae (Greenfeder and Newlon 1992b). Therefore, the ability of these ARS elements to function as chromosomal replication origins is also conserved.

While the analysis of S. monacensis suggests that these ARS elements are likely to function in the yeast that contributed the S. carlsbergensis genome to the alloploid brewing strain, our further analysis of the S. monacensis type strain suggests that these results must be interpreted with some caution. We have found that S. monacensis is also an alloploid, and it appears to carry S. cerevisiae-like chromosomes I and III (J. F. Theis and C. S. Newlon, unpublished results).

It was somewhat surprising that all five of the efficient origins of replication in the region of S. cerevisiae chromosome III to the left of MAT, ARS305, ARS306, ARS307, ARS309, and ARS310 were conserved on the S. carlsbergensis chromosome because many of the origins can be deleted from the S. cerevisiae chromosome without affecting its stability (Dershowitz and Newlon 1993; Newlonet al. 1993). Even more surprising was the conservation of ARS304, which is not active as a chromosomal replication origin. While it is possible that ARS304 is active under conditions that we have not tested, we consider it unlikely because our analysis was done on cultures grown in rich medium at optimal temperatures, a situation in which the rapid growth rate might be expected to require the use of all available origins. Moreover, we have examined origin function under a very different set of conditions, premeiotic S phase, and have found that ARS304 is not active during meiosis either (Collins and Newlon 1994). Alternatively, ARS304 might be used for some purpose other than replication. ORC binding sites are an important component of the silencers at HML and HMR (reviewed by Laurenson and Rine 1992), and the function of the ORC in silencing is distinct from its role in replication (Fox et al. 1995, 1997; Dillin and Rine 1997). Perhaps ORC binding at ARS304 is being used for some other purpose.

Our finding that the ARS elements on chromosome III of S. cerevisiae and S. carlsbergensis are in conserved locations provides strong evidence that these pairs of ARS elements are homologues. Comparisons of different ARSs with one another have failed to identify any common elements beyond the ARS consensus sequence (Broachet al. 1983). This may indicate that, while ORC binding may be a common feature of all ARS elements, the accessory elements may be different. By comparing homologous ARS elements from different species, we hope to identify ARS-specific functional elements.


This project was initiated during a sabbatical year at the Carlsberg Laboratory in Copenhagen. C.S.N. thanks Dr. Morten Kielland-Brandt for his enthusiastic support, hospitality, and many helpful discussions during this period. We thank Dr. Torsten Nilsson-Tillgren for sharing unpublished information about S. carlsbergensis CEN3 and for many helpful discussions, Dr. Cletus Kurtzman for yeast strains and advice, Dr. S. G. Oliver for helpful comments on the manuscript, and Dr. Lynn Ripley for helpful discussions. Support provided by the National Institutes of Health (NIH) Senior Fellowship HG-00027 and the Carlsberg Research Center is gratefully acknowledged. This work was supported by NIH grant GM-35679 to C.S.N. Partial support for C.Y. was provided by a fellowship from UMDNJ–Graduate School of Biomedical Sciences.


  • Communicating editor: M. Johnston

  • Received September 3, 1998.
  • Accepted April 1, 1999.


View Abstract