ABSTRACT
We have identified five autonomously replicating sequences (ARSs) in a 100 kbp
region of the
Schizosaccharomyces pombe
chromosome II. Analyses of replicative intermediates of the chromosome DNA by
neutral/neutral two-dimensional gel electrophoresis demonstrated that at least three of these
ARS loci operate as chromosomal replication origins. One of the loci,
ori2004
, was utilized in almost every cell cycle, while the others were used less
frequently. The frequency of initiation from the respective chromosomal
replication origin was found to be roughly proportional to the efficiency of
autonomous replication of the corresponding ARS plasmid. Replication from
ori2004
was initiated within a distinct region almost the same as that for replication
of the ARS plasmid. These results showed that the
ori2004
region of
~
3 kbp contains all the
cis
elements essential for initiation of chromosome replication.
Replication origins are physically defined as the sites where DNA synthesis is
initiated, while the regions required for initiation of chromosome replication
are called replicators. Initiation of replication in eukaryotic cells is
tightly regulated during the cell division cycle. The regulation presumably
involves interactions of various proteins with unique DNA sequences
constituting the replicators.
The structures of eukaryotic replicators have not been clarified, except in the
budding yeast,
Saccharomyces cerevisiae
. Certain chromosomal fragments of the yeast have been shown to be capable of
autonomous replication (autonomously replicating sequences, ARS) (
1
,
2
). All the yeast ARSs contain a match to an 11 bp sequence, called the ARS
consensus sequence (ACS), that is essential for ARS function (
3
,
4
). In addition to the ACS, distinct elements within a 100 bp region are required
for ARS function (
5
). Some of the ARS segments function as chromosomal replication origins (
6
,
7
). An origin recognition complex (ORC) containing six protein subunits has been
purified and shown to bind the ACS (
8
).
In vivo
footprinting yields a pattern similar to that found by footprinting with ORC
in vitro
(
9
). Moreover, analyses of temperature-sensitive
orc2-1
and
orc5-1
mutants by 2D gel techniques have shown that binding of ORC to the ACS is
essential for initiation of replication of the yeast genome (
10
).
Physical mapping of the replication origins of higher eukaryotic chromosomes has suggested that replication is initiated from restricted
regions, ranging in size from 0.5 to 55 kb (
11
). The replication origins for the human [beta]-globin and
Drosophila
chorion genes are located within 2 and 3 kb regions respectively (
12
,
13
). For their function, however, regions apart from the actual initiation sites
are also required (
13
,
14
). The elements involved in the initiation of replication remain to be
identified. In contrast to the budding yeast, no short chromosome fragments
capable of autonomous replication have yet been isolated from mammalian cells,
although human chromosome fragments in the 10 kb range do exhibit significant
autonomous replication activity in human cells (
15
-
17
). These results suggest that the replicators in higher eukaryotes have more
complex structures than the budding yeast replicator.
In
Schizosaccharomyces pombe
, certain chromosome fragments can replicate autonomously (
18
-
22
), but the regions required for ARS function are much longer than those in the
budding yeast (
19
,
23
). Although a match to an 11 bp sequence similar to the budding yeast ACS exists
in the
S.pombe
ARS fragments, the sequence is not essential for ARS activity (
19
). Detailed analyses of two ARS elements, namely
ars1
and
ars3002
, suggest the importance of an asymmetric A+T-rich sequence clustered within several hundred base pair regions (
24
,
25
). Thus, the fission yeast replicators appear to more closely resemble those in
higher eukaryotes with regard to the level of complexity. Physical mapping of the
fission yeast replication origins using two-dimensional (2D) gel techniques has shown that replication of a 6 kb region upstream
of the
ura4
gene is initiated from multiple sites (
26
). The origin region contains three ARS elements that are required for
initiation from the corresponding replication origins (
23
). Although these replication origins do not individually operate in every cell
cycle, initiation from the region is achieved in every cell cycle by their
concurrent action. It remains unknown whether the
ura4
origin region is characteristic of fission yeast replicators.
In this report, we describe identification of five distinct ARS elements in a
continuous 100 kb region of
S.pombe
chromosome II and provide evidence that three of the ARS loci function as chromosomal replication origins. Our results suggest that a single ARS element
can promote chromosome replication in every cell cycle.
The
S.pombe
haploid strain used was HM123
h
-
leu1
(
27
). It was cultured in YPD complete medium (1% yeast extract, 2% polypeptone, 2%
glucose) and EMM minimal medium (
28
).
Escherichia coli
DH5[alpha] (
29
) was grown in LB (0.5% yeast extract, 1% polypeptone, 1% NaCl, pH 7.5). For EMM and LB plates, agar was added to 2 and 1.5% respectively. Plasmid DNA was prepared from
E.coli
transformants as described previously (
30
).
Restriction fragments (2-15 kb) of cosmid clones containing
S.pombe
chromosomal DNA, sp1029, 205, 1228 and 1558 (
31
), were subcloned into pYC11, a derivative of Bluescript KS(+) carrying the
S.cerevisiae
LEU2
gene (
32
). Derivatives of pARS2004 with an additional insert at either the left or right
end of the ARS segment were constructed as follows. A 564 bp
Hin
dIII fragment of bacteriophage [lambda] and the 3.2 kb
Not
I-
Xba
I fragment of pARS2004 were inserted into the
Hin
dIII and the
Not
I-
Xba
I sites respectively of pYC11, resulting in pARS2004H[lambda]. The phage [lambda] fragment was first cloned into the
Hin
dIII site of a pYC11 derivative whose
Xho
I site had been altered to a
Not
I site and then the
Not
I fragment of the resulting plasmid was inserted into the
Not
I site of pARS2004, resulting in pARS2004N[lambda].
The electroporation method (
33
) was employed to introduce plasmid DNA into
S.pombe
cells. HM123 cells (1 * 10
7
cells/ml) were washed three times and suspended in cold 1.2 M sorbitol at a concentration of 1 * 10
9
cells/ml. To the cell suspension (0.1 ml), 0.2 [mu]g plasmid DNA was added with 5 [mu]g sonicated salmon testis DNA. After electroporation at 2000 V, 200 [Omega] and 25 [mu]F, one twentieth of the suspension was spread on an EMM plate
and incubated for 4 days at 30oC.
For determination of the physical status and copy numbers of ARS plasmids, the
total cellular DNA of transformants was prepared as described previously (
34
). The DNA was separated by agarose gel electrophoresis before or after
digestion with a restriction enzyme and analyzed by Southern blot hybridization
with the vector probe.
The stability of ARS plasmids was determined by the method described by Heyer
et al
. (
35
). Transformants grown in EMM medium to 1 * 10
7
cells/ml were diluted and plated onto YPD plates. Colonies which formed after 2
days at 30oC were replica-plated onto both EMM and YPD plates to determine the percentage of
plasmid-containing cells under selective conditions (
A
). The cells in the EMM culture were then diluted to 1 * 10
3
/ml with YPD medium and grown at 30oC for ~10 generations without selection. After scoring the cell number (
n
), diluted cells were plated onto YPD plates. The colonies formed were then
replicated on EMM and YPD plates to determine the percentage of plasmid-containing cells under non-selective conditions (
B
). Plasmid loss rate per generation was calculated with the equation 1 - (
B
/
A
)
1/N
, where N = 3.3 log
10
n
- 10.
The neutral/neutral 2D gel electrophoresis method described by Brewer and
Fangman (
36
) was employed to determine the initiation sites of replication on the
S.pombe
chromosome and ARS plasmids. Total cellular DNA was prepared as described by
Shinomiya
et al
. (
37
) with some modifications (
38
). A culture of log phase cells (5 * 10
9
cells) in YPD medium was mixed with an equal volume of cold stop solution
containing 3% toluene, 95% ethanol and 20 mM EDTA. The cells collected by
centrifugation were washed twice with SE buffer (75 mM NaCl, 100 mM EDTA) and
suspended in the same buffer at ~1 * 10
8
cells/ml. The cell suspension (5 ml) was then mixed with an equal volume of 1%
low melting point agarose (SeaPlaque GTG Agarose; FMC BioProducts) and 4 vol.
paraffin prewarmed to 42oC. After vigorous shaking for 1 min, emulsions were poured into 5 vol. cold
SE buffer and stirred for 5 min. The agarose beads were collected by
centrifugation, incubated at 37oC for 30 min in an equal volume of SE buffer containing Zymolyase 20T (Seikagaku Kogyo) at 5 mg/ml and then in 2 vol. 1% SDS and 25 mM EDTA for 10 min at room temperature. After further incubation at 37oC for 1 h in 2 vol. 1% Salkosyl, 25 mM EDTA containing proteinase K at
0.5 mg/ml and washing with TE buffer (10 mM Tris-HCl, pH 7.4, and 1 mM EDTA) containing 0.1 mM phenylmethylsulfonyl
fluoride, the DNA in agarose beads was digested with appropriate restriction
enzymes and electroeluted in a dialysis bag. After removal of agarose by centrifugation at 12 000 r.p.m. for 15 min at 4oC, DNA was concentrated with isobutyl alcohol and precipitated with 2-propanol. Electrophoresis for the first dimension was carried out in 0.35% agarose gels at 1 V/cm in TBE buffer (100 mM Tris, 100 mM boric acid and 2 mM EDTA) at 4oC and for the second dimension in 0.875% agarose gels at 8 V/cm
in the presence of 0.3 [mu]g/ml ethidium bromide. The DNA was blotted onto Hybond-N
+
membrane and hybridized with
32
P-labeled probes. The membranes were placed in contact with imaging plates
for 2 days and the stored images were analyzed with an Image Analyzer Bas1000Mac (Fuji Film, Tokyo).
To isolate chromosome fragments capable of replicating autonomously from
S.pombe
, we cloned overlapping fragments of 2-15 kb from a 100 kb region of chromosome II into a vector carrying the
S.cerevisiae
LEU2
gene. Each clone of the library was introduced into
S.pombe
leu1
cells to measure the ARS activity. Plasmids carrying five distinct fragments
gave leu
+
transformants at a high efficiency of ~1 * 10
6
/pmol DNA (Fig.
1
). Plasmids with other fragments were unable to transform at a frequency higher
than 1 * 10
4
/pmol DNA. Thus, only distinct regions of the
S.pombe
genome had the ability for autonomous replication. The ARS regions identified
were designated
ars2001
-
2005
from proximal to
cen2
(Fig.
1
).
Earlier studies have shown that
S.pombe
ARS plasmids are readily rearranged into large multimeric forms (
21
,
39
). To examine the physical status of newly isolated ARS plasmids, total cellular DNA
extracted from the leu
+
transformants was analyzed by Southern hybridization with the vector fragment. Most of the ARS plasmids examined, except for pARS2001L and pARS2002S, existed in a monomeric
form (Table
1
). pARS2001L, a larger derivative containing
ars2001
, was found in a large multimeric form, while pARS2002S, a shorter derivative with
ars2002
, was in dimeric and trimeric forms. The transformants with pARS2001L and pARS2002S grew at a much slower rate than those with the
other ARS plasmids (data not shown). Thus, plasmids that replicated efficiently were maintained as monomers, while those replicating inefficiently were rearranged into multimers.
To estimate the efficiency of replication of ARS plasmids that were present as
monomers, we examined their copy numbers in the transformants. By hybridization
analysis after linearization with a restriction enzyme, the copy number of
pARS2004 was estimated to be ~14/cell (Table
1
). pARS2002 and pARS2003 were present at about half the copy number of pARS2004,
while pARS2001 and pARS2005 at about one fourth (Table
1
). These results suggeste that pARS2004 replicated at a higher efficiency than
the other plasmids isolated.
Since the vector used in this study did not contain a centromere, the stability
of the plasmid during cell cycles would depend mainly on the efficiency of
autonomous replication. The rate of mitotic loss of pARS2004 under the non-selective conditions (see Materials and Methods) was 2.3% per generation.
Plasmids pARS2002 and pARS2003 were slightly less stable than pARS2004, while pARS2001 and pARS2005 were relatively unstable during cell cycles (Table
1
). From comparison of the results on the copy numbers of plasmids and mitotic
stability, the relative order of autonomous replication activity was thus
judged to be ARS2004 > 2002 or 2003 > 2001 or 2005.
To examine whether chromosomal replication was initiated from the ARS loci or
from regions outside, we employed the neutral/neutral 2D gel electrophoresis
technique (
36
). The total cellular DNA from exponentially growing HM123 cells was separated by 2D gel
electrophoresis after digestion with appropriate restriction enzymes. By hybridization with the ARS2004 probe, a complete bubble arc that started from the 1N position and extended beyond the 2N position was detected (Fig.
2
B), demonstrating that chromosomal replication had been initiated within the restriction fragment. Such a bubble arc was also observed with the
ARS2003 and ARS2002 probes (Fig.
2
D and E). The origins of replication in the
ars2002
,
2003
and
2004
regions of chromosome II were designated
ori2002
,
2003
and
2004
respectively.
To determine the initiation site in the
ori2004
region, cellular DNA was digested with various restriction enzymes and analyzed
by 2D gel electrophoresis. If the initiation site was located at the center of
a restriction segment, a complete bubble arc would be expected. On the other
hand, if replication was initiated from another position away from the center,
a bubble arc would be detected near the 1N position and a Y arc would be formed
with progression of replication. As shown in Figure
3
A, the
Bgl
II-
Pst
I fragment formed a complete bubble arc, suggesting that replication was initiated near the center of the fragment. On the other hand, the
Bgl
II-
Xba
I fragment, lacking a 1 kb region at the right end of the
Bgl
II-
Pst
I fragment, formed a shortened bubble arc and a Y arc whose intensity near the
2N position was stronger than that near the 1N position (Fig.
3
B). Furthermore, the
Kpn
I-
Pvu
II fragment lacking ~2 and 0.5 kb regions from the left and right ends of the
Bgl
II-
Pst
I fragment respectively formed a short bubble arc ending half way from the 1N to
2N positions and a strong Y arc starting at the middle and ending at the 2N
position (Fig.
3
C). These results show that replication in the
ori2004
region was initiated at a distinct site away from the center of the
Bgl
II-
Xba
I and
Kpn
I-
Pvu
II fragments. The
Bam
HI fragment, which overlapped with the right half of the
Kpn
I-
Pvu
II fragment, formed only a Y arc (Fig.
3
D). These results suggest that replication was initiated from a site in the left
half of the
Kpn
I-
Pvu
II segment.
To examine whether replication of plasmid pARS2004 was initiated from a distinct
site corresponding to the initiation site on the chromosome, cellular DNA from
the plasmid-carrying HM123 cells was analyzed by 2D gel electrophoresis. Upon
digestion of the DNA with
Pvu
II, which excised a 3.4 kb fragment containing
ars2004
, a strong bubble arc extending to nearly the 2N position was detected using the ARS fragment as probe (Fig.
4
A), indicating that replication was predominantly initiated within the ARS segment.
In addition to the bubble arc, a strong Y arc starting half way between the 1N
and 2N positions and ending at the 2N position was observed (Fig.
4
A). The transition from the bubble to Y arc indicated that replication was
initiated from a distinct site located away from the center of the fragment.
Based on the estimated molecular mass of the smallest Y-shaped molecule (6.0 +- 0.2 kb), the initiation site was mapped to 1.4 +- 0.2 kb from either end of the
Pvu
II fragment (half arrowheads in Fig.
4
). For determination of the initiation site, a 0.5 kb fragment was inserted near
the left or right end of the
Pvu
II fragment to shift the relative location of the initiation site. With
insertion at the left end, the bubble arc was greatly extended and a transition
to a very short Y arc was found (Fig.
4
B), showing that the initiation site in the fragment was at a position near the
center. In contrast, the fragment with the insertion at the right end formed a
shortened bubble arc and a slightly extended Y arc, compared with those without insertion (Fig.
4
C). The initiation sites estimated from the molecular masses of the smallest Y-shaped molecules were located at 2.0 +- 0.2 or 1.3 +- 0.2 kb from the
Pvu
II sites on plasmids with insertions at the left or right end of the ARS fragment. From the results we conclude that replication on plasmid pARS2004 was initiated
from a distinct location near the
Bam
HI site at 1.3 kb from the left end of the ARS fragment (see upper part of Fig.
4
). These results, together with those presented in Figure
3
, indicate that replication on the pARS2004 plasmid was initiated from the same site as that in the
ori2004
region on the chromosome.
We have identified five ARS elements in a 100 kb region of the fission yeast
chromosome II, at least three of which function as chromosomal replication origins. The fact that ARS-lacking regions did not show any bubble arcs on 2D gel analyses indicates that
chromosomal replication is likely to be initiated from distinct regions that are capable of autonomous replication. Initiation from
ori2004
occurs in every or almost every cell cycle, while that from
ori2002
and
ori2003
may take place once every few cell cycles. The efficiency of autonomous
replication of the corresponding ARS plasmid is roughly proportional to the frequency of utilization of the
corresponding chromosomal replication origin. These results suggest that
elements involved in initiation of chromosomal replication are localized within
limited regions and that the elements cloned into plasmids function as in the chromosomal locations.
The replication of pARS2004 is initiated from a unique site in the ARS fragment, which is identical, within the resolution of the 2D gel technique, to that in
the
ori2004
region on the chromosome. We can thus conclude that the ARS2004 fragment
contains all the
cis
elements required for initiation from the specific site. Analyses of deletion
derivatives from either end of the 3.2 kb long ARS2004 fragment revealed that
the initiation site is present within a 940 bp segment essential for ARS
function (H.Satoh and H.Masukata, unpublished results). These results imply
that the essential region would be directly involved in reactions that promote
replication.
It has been shown that the
ura4
origin region contains three clustered ARS elements that are functional as chromosomal replication origins (
23
,
26
). Although none of the origins operates in every cell cycle, initiation of
replication in the
ura4
region does occur in every cell cycle. From the results, the authors proposed
that the
ura4
origin region is a model for a replication initiation zone in higher eukaryotic
chromosomes (
40
). In contrast, the
ori2004
region appears to contain a unique replication origin that operates in every
cell cycle. Therefore, our results suggest that a single ARS element can
function as a predominant chromosomal replicator and clustering of multiple
origins as in the
ura4
region is not a general feature of fission yeast replicators.
It has been shown that the ARS elements of the fission yeast are several fold
larger than those of the budding yeast. The ARS activity is gradually reduced
by deletions from the ends of the fragments and no consensus sequence essential
for ARS function has been identified (
19
,
23
). These characteristics are similar to those of human ARS fragments (
15
,
17
,
41
). However, the latter did not show any significant ARS activity in fission
yeast and fission yeast ARS fragments did not replicate in human cells at a
higher efficiency than non-ARS fragments (H.Satoh and H.Masukata, unpublished results). Presumably,
the sequences required for initiation differ between these organisms. Extensive
analyses of two ARS elements,
ars1
and
ars3002
, have suggested that asymmetric A+T-rich sequences clustered in one or two regions within a several hundred
base pair segment are important for ARS activity (
24
,
25
). Detailed analyses of the ARS elements obtained in this study seem to be
important to deduce the essential sequences and common structures of the
S.pombe
replicators.
In the budding yeast, recognition of the ARS consensus sequence by the ORC
protein complex plays a crucial role in the initiation of replication (
8
,
42
). Recently, genes homologous to
ORC1
or
ORC2
have been identified in many organisms (
43
-
45
) and the homologs in fission yeast,
orp1
+
and
orp2
+
, have been confirmed to be essential for cell growth (
43
,
46
,
47
). The Orp1 and Orp2 proteins of fission yeast, like those of the budding yeast
(
8
), may interact with some element(s) in the replicator. The presently described
ars2004
that promotes efficient replication from a specific site could be used as a
model replicator for studies of initiation of replication of fission yeast
chromosomes.
We thank Y.Sakakibara, J.Tomizawa, H.Ogawa and T.Yonesaki for critical reading
of the manuscript and helpful discussions, M.Yanagida for providing cosmids and yeast strains and K.Shirahige for advice on 2D gel analysis. This work was supported by a Grant-in-Aid for Scientific Research on Priority Areas from the Ministry of
Education, Science and Culture of Japan (to H.M.).
*To whom correspondence should be addressed at: Department of Biology, Graduate
School of Science, Osaka University, 1-1 Machikaneyama, Toyonaka, Osaka 560, Japan. Tel: +81 6 850 5432; Fax: +81
6 850 5440; Email: masukata@bio.sci.osaka-u.ac.jp
+
Present address: Department of Biology, Graduate School of Science, Osaka
University, 1-1 Machikaneyama, Toyonaka, Osaka 560, Japan.
REFERENCES
Return