ABSTRACT
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.
Activation of DNA replication origins within a eukaryotic chromosome is a
tightly regulated event, limited to a single occurrence during the S phase of
the cell cycle (reviewed in
1
). The collection of
cis
-acting DNA elements that determine the activity of an origin are called
the replicator, in general usage, and the
ori
or
ORI
, for a particular genetic locus, e.g.
ori
[lambda] (
2
-
4
). Among eukaryotic replicators, those in the yeast
Saccharomyces cerevisiae
are currently best defined. Yeast replicators were first identified as
chromosomal DNA fragments that, when inserted into plasmids, permitted high
frequency transformation of cells without integration (
5
,
6
). Such DNA fragments are called autonomously replicating sequence (ARS)
elements.
ARS activity requires a specific sequence that matches the ARS consensus
sequence or ACS:
A
/
T
TTTA
T
/
C
G
/
A
TTT
A
/
T
(
7
-
9
). The ACS is the core of a larger functional sequence called element A (
10
,
11
). The ACS is sensitive to point mutations and serves as a binding site for the origin recognition complex (ORC), a multiprotein complex
that contributes to the initiation of DNA replication (
12
-
15
). Sequences in the A element but outside the ACS appear non-conserved (
7
) and little is known about the precise requirements for such sequences in
different ARS elements.
In addition to the A element, a broad region called domain B (
10
), located downstream (3') of the T-rich strand of the ACS, is also required for ARS function (reviewed
in
8
). Domain B spans ~100 bp and exhibits no extended sequence conservation among ARSs (
7
). In several ARSs, domain B is easily unwound (
16
). A functional role for a DNA unwinding element (DUE) within domain B is
supported by mutations that reduce both the ease of DNA unwinding and
replication origin activity and by the functional substitution of domain B
sequences with dissimilar sequences that are easily unwound (
3
,
17
-
19
). Linker substitution analysis has identified short sequence elements that
contribute to the function of domain B in
ARS1
and
ARS307
(
11
,
20
,
21
).
ARS1
function requires three short sequence elements in domain B. Element B3 binds
ABF1, a transcription factor and a replication enhancer protein (
22
). Element B1 contributes to ORC binding
in vitro
and may have additional roles (
23
). Element B2 interacts with a single-stranded binding protein (RP-A) in supercoiled DNA and has been suggested to function as a DUE (
24
).
ARS307
contains functionally conserved elements at positions overlapping
ARS1
B1 and B2 elements (
20
,
21
). The
ARS307
B2 element is embedded in a broader region that contains a DUE (
17
). Analogous to models for protein-induced activation of
Escherichia coli
oriC
and the SV40
ori
(
25
,
26
), including a general role for a DUE (
27
,
28
), a model for activation of yeast replication origins is that DNA-bound ORC, when appropriately modified or associated with other factors (
13
,
29
), facilitates opening of the DNA helix in the DUE and subsequent entry of the
replication machinery.
Experiments aimed towards genetically defining replicators within yeast
chromosomes have shown that ARS elements share essential components with chromosomal replicators. At four different chromosomal loci a specific sequence that matches the ACS is
essential for origin function (
3
,
30
-
32
). Also essential is domain B, including the DUE region in
ARS305
(
3
), as well as the three B elements of
ARS1
(
31
). The contribution of element A sequences outside the ACS to the chromosomal replicator are not known. Also
unknown is why the contribution of certain B elements to plasmid ARS function
differs from the contribution of those same elements to chromosomal replicator
function (
21
,
31
). Definition of the replicator by direct measurement of replication origin
activity in the natural context of the chromosome is important, since ARS
activity only indirectly reflects replication origin activity (
33
,
34
) and can be strongly influenced by plasmid context (
35
-
37
).
We have performed a linker substitution analysis of
ARS305
to identify short DNA sequence elements that contribute to plasmid ARS function
and to define the contribution of such elements to replication origin function
in the natural chromosomal locus. In addition to the conserved ACS, we find
that a non-conserved sequence in element A is a major determinant of
ARS305
and chromosomal replicator function. Also, two B elements contribute
quantitatively to ARS and chromosomal replication efficiency. Comparison of
functional elements in three genetically defined chromosomal replicators shows
that an element A-B1 array, the bipartite ORC binding site (
23
), is conserved, but no two replicators are identical in the extent of the A
element outside the ACS or in the arrangement of B elements downstream of B1.
The
E.coli
strain used for plasmid propagation was DH5 (BRL).
Saccharomyces cerevisiae
strain YPH98 (haploid, MAT
a
,
ade2-101
,
lys2-801
,
ura3-52
,
trp1-1
,
leu2-1
) was obtained from Philip Hieter (Johns Hopkins University). Bacteria and yeast
cells were propagated and plasmid DNA was isolated as described in Huang and
Kowalski (
3
).
AmpliTaq polymerase was obtained from Perkin Elmer Cetus. Restriction enzymes
and T4 DNA ligase were obtained from New England BioLabs Inc. [alpha]-
32
P-Labeled deoxyribonucleotides were purchased from Amersham Corp. Media
components were from Difco Laboratories. DNA sequencing gel solutions (Sequagel) were from National
Diagnostics. Other chemicals were purchased from Sigma Chemical Co.
Mutations in
ARS305
were generated by the method of Perrin and Gilliland (
38
). The p305BP plasmid (
3
) containing the 2 kb
Bam
HI-
Pst
I fragment derived from the A6C region of chromosome III (
39
) was used as a template in the PCR. Oligonucleotides were synthesized by Al
Cairo (Biopolymer Facility, Roswell Park Cancer Institute). The mutagenesis
procedure involved two rounds of PCR and was performed as described previously
(
3
) except for the following modifications. In the first amplification a double-stranded fragment with the mutation of interest was prepared using 2.5 ng
p305BP DNA along with the
ARS305
mutant primer (10 pmol) at the 3'-end and a wild-type primer 305L3 (10 pmol) at the 5'-end. The DNA fragments were separated by gel electrophoresis (1.5%
Nusieve/1.5% SeaPlaque LM agarose; FMC) and isolated using the PCR Magic Prep
kit (Promega). The second amplification to generate a full-length mutant fragment was performed using a wild-type primer 305R3 (10 pmol) as 3'-flanking primer and the first PCR product (estimated 10
pmol) as the source of 5'-primer. The mutant primers for linker substitutions contain an 8 bp
Xho
I linker sequence flanked by 16 or 17 bp wild-type sequences on each side. For some of the point mutations degenerate
primers containing the three alternative nucleotides at the specific position
of interest were used as mutant primers. The sequences of the primers are:
305L3
5'-CGTGTAAGCTGGGGTGACTTTTGAGCTATTCGC AC-3';
305R3
5'-TGATCTTAGTTGGTAGCACTTTGATGAGGTCTCTAG-3'.
Each of the PCR products was cleaved with
Sac
I and
Cla
I and the electrophoretically purified 336 bp
Sac
I-
Cla
I fragments were then ligated with purified vector DNA (p305BP without the corresponding wild-type
Sac
I-
Cla
I region of
ARS305
) and transformed into DH5[alpha] cells. Plasmids with a
Xho
I linker substitution were identified by their susceptibility to cleavage by
Xho
I restriction enzyme. The DNA sequence of each mutant derivative was verified by
the dideoxy chain termination method using Sequenase (US Biochemicals) and the
305R1 primer (
3
).
The
Sph
I-
Eco
RI fragments from p305BP derivatives were subcloned into the
Sph
I and
Eco
RI sites of a centromere-containing vector, pVHA (
40
), a gift from Carol Newlon (UMD-New Jersey Medical School). Plasmids were transfected into the yeast cells using the
lithium acetate procedure (
41
). After transfection, cells were plated on synthetic minimal medium minus
uracil (SMM/U
-
) at 30oC for 2-4 days and scored for HFT (>200 transformants/g DNA). Transformants
were scored as Ars
+
if they could be passaged in SMM/U
-
liquid culture.
The mitotic stability assays were performed as described (
42
) with the following modifications. Single transformants from the SMM/U
-
plates were each inoculated in 3 ml selective medium and grown for 22-24 h at 30oC. Cultures were plated on non-selective (YPD) and selective plates (SMM/U
-
) at dilutions that gave rise to ~200 colonies on non-selective plates to determine the initial percentage of plasmid-containing cells under selection (
I
). Cultures were also diluted (1 * 10
4
cells/ml) and grown in 3 ml non-selective medium for 12 generations at 30oC. Cultures were plated on non-selective and selective media as above to determine the final
percentage of plasmid-containing cells after growth in the absence of selection (
F
). For transformants with extremely low growth rates, twice and 200 times the
amount of cells were plated on the selective plates to determine the
I
and
F
values respectively. Mitotic stabilities (percentage of plasmid-containing cells) before release of selection and after growth in non-selective medium for a specific number of generations (
n
) correspond to
I
and
F
respectively. The value of
n
is 12 generations in this study and the mitotic stability values reported
correspond to F
12
. Plasmid retention per generation (
PR
, %) is expressed as [
F
/
I
]
1/
n
(intuitively
Fn
=
I
[
PR
]
n
). The plasmid replication efficiency was calculated by normalizing
PR
for the centromere (CEN)-containing plasmid retention assumed to occur in the absence of any plasmid replication (50% per generation)
and is expressed as 2[
PR
- 50].
Linker substitution mutations were introduced into the
ORI305
locus by transplacement using procedures previously described (
3
). The desired homologous recombinants, each containing a
Xho
I linker substitution in the
ORI305
locus, were identified after restriction digestion of genomic DNA with
Xho
I and
Eco
RI and detection of the appropriate sized fragment following gel electrophoresis
and Southern blotting. DNA replication intermediates isolated from high
molecular weight yeast genomic DNA were isolated, enriched and separated using
two-dimensional gel electrophoresis and the signals analyzed using a
phosphorimager (Molecular Dynamics), as previously described (
3
).
A modified two-dimensional agarose gel electrophoresis method was used to determine the
direction of replication fork movement adjacent to the mutant
ORI305
loci. The method involves an `in gel' digestion with a second restriction
enzyme between the first and second dimensions of electrophoresis and is
essentially as described by Brewer
et al
. (
43
). After electrophoresis, DNA was transferred to a nylon membrane and hybridized
with a
32
P-labeled DNA probe. The radioactive signals were detected by scanning in a
phosphorimager and were quantified using ImageQuant software (Molecular
Dynamics) by analyzing parallel, non-overlapping regions of the two Y arcs.
To identify short sequence elements that are essential or otherwise important for ARS function we constructed a series of linker substitution mutations in
ARS305
. Oligonucleotide-directed mutagenesis via PCR was used to substitute an 8 bp G+C-rich sequence (
Xho
I linker) in place of the wild-type sequence at a variety of positions in
ARS305
. The ARS was present in a vector containing the
URA3
selectable marker and a CEN for proper plasmid segregation.
In contrast to Lin12, which caused a severe reduction in ARS mitotic stability,
Lin14, the same linker sequence introduced only 2 nt further downstream, had
little or no effect (Table
1
). Introduction of a different linker sequence at positions 14-21 (Lin14a) also had little effect compared with that seen for the Lin12
mutation (Table
1
). One of the possible interpretations of these results is that one or two
residues adjacent to the 11 bp ACS are crucial for replication origin function.
The residue 3' of the ACS (position 12) is known to be important but not essential in
both
ARS307
(
40
) and the H0
ARS
(
45
). As seen in Table
1
, point mutations at position 12 in
ARS305
reduce mitotic stability while all point mutations at position 13 had little or
no influence. These results indicate that a G at position 12 is important for
ARS305
function and that point substitutions lead to decreased function in the
following order G >= T > C > A; however, none of the single point mutations at position 12 or 13
led to the extremely low mitotic stability seen for Lin12, which substitutes
positions 12-19.
A double point mutation that matches the first two residues of Lin12 was
introduced at positions 12 and 13 to test whether the extremely low mitotic
stability seen for Lin12 was due solely to an effect of simultaneous mutations
at positions 12 and 13. As shown in Table
1
, this was not the case, since the mitotic stability of the double point mutant
C12C13 was much greater than that of Lin12.
Since separate mutations in positions 12-13 (C12C13) or 14-19 (Lin14 and Lin14a) did not severely reduce mitotic stability,
the extremely low mitotic stability of derivative Lin12 appeared to result from
a synergistic effect of mutations in two regions: positions 12-13 and positions 14-19. Three different linker sequences at positions 14-19 in combination with the C12C13 mutation led to an
extremely low mitotic stability (Table
1
, Lin12, Lin12a and Lin12b), confirming the synergistic effect and supporting
the functional importance of the region containing positions 14-19.
Reversion of the C mutation at position 13 in Lin12b to a wild-type T in an otherwise constant mutant background in derivative Lin12c led
to a high level of mitotic stability (Table
1
, Lin12c). Thus the T at position 13 is functionally important when
substitutions are present simultaneously at position 12 and in the region
containing positions 14-19. This observation is surprising, since position 13 appeared relatively
unimportant when point mutated in the wild-type context; however, this observation is consistent with the synergistic
effects seen for other box 3' mutations. Overall, the results show that
ARS305
function is influenced by several nucleotide residues at positions 12-19 which comprise box 3' and that simultaneous mutations in these residues can act
synergistically to essentially inactivate ARS function.
To determine the contribution that elements revealed by linker substitution of
ARS305
make to replication origin activity in the chromosome we first replaced the
natural sequence at the
ORI305
locus in a haploid yeast strain with mutated sequences containing linker
substitutions by using gene transplacement techniques. Then we analyzed replication origin activity around the mutated
ORI305
locus by two-dimensional gel electrophoresis (
46
). Chromosomal DNA was cleaved at specific restriction enzyme sites and the
fragments were separated by mass in the first dimension gel electrophoresis and
by shape and mass in the second dimension. DNA replication intermediates with
the sequence of interest were detected after Southern transfer by hybridization
using a specific
32
P-labeled DNA probe. If the specific DNA segment in the chromosome contains
an active replication origin a high rising arc, termed a bubble arc, is
detected, resulting from the progressively retarded mobility of intermediates
as the two replication forks move away from the centrally located origin (Fig.
3
F). If the replication origin in the specific DNA segment is inactivated as a
result of a mutation then the segment is passively replicated from an external
origin. In this case the bubble arc is diminished or not detected, depending on
the severity of the mutation, and intermediates with a single replication fork
give rise to a distinct arc, called a Y arc (Fig.
3
F). The relative intensities of the bubble and Y arcs reflect the relative
frequencies of replication origin activity and passive replication.
Figure
3
shows two-dimensional gel analysis of chromosomal origin function at the wild-type
ORI305
locus and at mutant
ORI305
loci in the yeast strains we constructed. Analysis of the wild-type
ORI305
locus shows a prominent bubble arc with little or no early Y arc, indicative of
efficient replication origin function (Fig.
3
A). A strong late Y arc signal is seen, indicating that one of the replication
forks from the active origin reaches an end of the specific restriction
fragment before the other fork, converting the bubble-shaped intermediates to Y-shaped.
In strain YRHLin12 the
ORI305
locus contains the Lin12 mutation. Lin12 identified box 3' as extremely important for plasmid ARS efficiency (Fig.
2
and Table
1
, Lin12). Two-dimensional gel analysis of replication intermediates from YRHLin12 shows
only a strong complete Y arc, indicating that the replication origin is not
detectably active in the chromosome (Fig.
3
B). Thus the Lin12 mutation identifies box 3' as extremely important for replication origin function mediated by
ORI305
in chromosome III.
Table 1
Strains YRHLin22 (Fig.
3
C), YRHLin-8 (Fig.
3
D) and YRHLin102x-8 (Fig.
3
E) contain Lin22, Lin-8 and Lin102 plus Lin-8 mutations respectively. The Lin-8 mutation substitutes the 8 bp immediately 5' of the ACS and has little effect on plasmid mitotic
stability (Fig.
2
). The Lin22 and Lin102 mutations define the B1 and B4 elements respectively and
significantly decrease plasmid mitotic stability (Fig.
2
). In all three strains two-dimensional gel analysis shows prominent bubble arcs, indicating that the
mutant
ORI305
loci mediate a high level of replication origin activity (Fig.
3
C-E); however, all three strains also show evidence of early Y arc-like signals, raising the possibility that origin activity may be diminished (see quantitative analysis
below).
Standard two-dimensional gel analysis of replication intermediates provides mostly
qualitative information. To quantify the effects of mutations on origin
activity we used a modified two-dimensional gel method that assesses the direction of replication fork
movement in the chromosome. Unlike the standard two-dimensional gel method, the modified method is free of possible
interference from Y arc-like signals that result from breakage of replication bubbles (
47
,
48
). After the first dimension electrophoresis of
Xho
I-cleaved genomic DNA, the DNA was digested in the gel with
Hin
dIII, subjected to second dimension gel electrophoresis and, after blotting, probed
for the large
Xho
I-
Hin
dIII fragment (Fig.
4
A). Rightward moving forks first enter the large
Xho
I-
Hin
dIII fragment, forming an arc of Y-shaped intermediates which arises from the intense spot containing non-replicating DNA (Fig.
4
B, R). Leftward moving forks first enter the small
Xho
I-
Hin
dIII fragment that is removed (Fig.
4
A, L
1
). Consequently, leftward moving forks form an arc of simple-Y intermediates that is displaced from the intense spot (Fig.
4
B, L). If
ORI305
is active, replication forks are expected to move rightward through the probed
fragment. If
ORI305
is inactivated by mutation, replication forks are expected to move leftward,
since the 306 origin (Fig.
4
A) and other centromere-proximal origins are active and since there are no active origins on the
telomere side of
ORI305
(
49
,
50
). The proportion of rightward moving forks reflects the replication origin
efficiency mediated by
ORI305
.
Linker substitution analysis of DNA is useful in identifying short sequences
that serve as functional elements, since the substitution alters the sequence
of an individual element but not the potentially important spacing between
elements (
51
). Linker substitutions in
ARS305
and analysis of plasmid mitotic stability revealed three elements downstream of
the ACS. The major functional element is box 3', which resides directly 3' of the essential ACS and is part of element A. Two additional
elements, B1 and B4, were detected further downstream in sequences in the B
domain. All three elements reside in sequences that are not conserved among
ARSs (
7
; data not shown). Our results show that
ARS305
function requires multiple short sequence elements in non-conserved ARS sequences downstream of the ACS.
Since plasmid ARS activity only indirectly reflects replication origin activity
and can be influenced by plasmid context, we examined whether and to what
extent each element that affects plasmid
ARS305
function also contributes to authentic replication origin function at the
ORI305
locus within chromosome III. Box 3' in element A is a major determinant, since linker substitution
essentially inactivates the replication origin in the chromosome. This is the
first identification of a short sequence element outside the ACS that is
crucial for replication origin function in a yeast chromosome.
Previous studies suggested that B elements contribute differently to plasmid and
chromosomal replicator function (
21
,
31
). For
ARS1
the B1 and B2 elements appeared to make different relative contributions to
plasmid stability and origin activity (
31
), although origin activity was not analyzed quantitatively. For
ARS307
B1 mutations affected quantitative origin activity less dramatically than they
affected plasmid stability over multiple generations (
21
). Here the contribution that the B1 and B4 elements make to chromosomal origin
efficiency was demonstrated by modified two-dimensional gel analysis, which permits quantitation of replication fork
directions. Furthermore, we compared the plasmid replication efficiency in a
single S phase, as opposed to the plasmid stability over multiple generations.
Our findings show that the contributions the B elements make to plasmid
ARS305
function reflect the quantitative contributions that they make to replication
origin function in the chromosome. Thus B elements detected in the ARS plasmid
likely serve the same function at the natural replicator locus within the
chromosome.
The overall organization of genetic elements in
ARS305
is compared with that in
ARS1
(
11
) and
ARS307
(
20
,
21
) in Figure
5
A. In addition to an ACS in element A, all three ARSs contain a B1 element and
one or two additional B elements. No element was detected by linker
substitution in
ARS305
at positions corresponding to B2 elements in
ARS1
and
ARS307
. The B4 element in
ARS305
overlaps the position of the B3 element in
ARS1
; however, B3 corresponds to an ABF1 binding site, but the sequence of B4 does
not match an ABF1 consensus and cannot be functionally substituted by B3, B2 or
B1 (S.Lin and D.Kowalski, unpublished data). At the B3/B4 position no element
is present in
ARS307
. All three ARSs differ in the extent of their sequence requirements directly 3' and 5' of the ACS. Only in
ARS305
and
ARS307
does element A extend significantly 3' of the ACS element, corresponding to the box 3' region of
ARS305
. The comparison reveals that an element A-B1 array, the bipartate ORC binding site (
23
), is conserved among the three replicators, however, no two replicators are
identical in the extent of element A outside the ACS or in the arrangement of B
elements downstream of B1.
We thank Martha Eddy for expert technical assistance and Joel Huberman and Sluan
Lin for helpful comments on the manuscript. This work was supported in part by
grants from the American Cancer Society (NP-872) and the National Institutes of Health (GM30614).
Derivative
Sequence
a
Mitotic stability (%)
b
12 19
[middot] [middot]
WT
GTTATGTATT
74 +- 9
Lin12
ccTcgagg
TT
0.03 +- 0.01
Lin14
GT
ccTcgAgg
63 +- 12
Lin14a
GT
ggaGctcc
55 +- 4
A12
a
TTATGTATT
20 +- 7
C12
c
TTATGTATT
40 +- 7
T12
t
TTATGTATT
56 +- 10
A13
G
a
TATGTATT
61 +- 8
C13
G
c
TATGTATT
69 +- 10
G13
G
g
TATGTATT
62 +- 13
C12C13
cc
TATGTATT
35 +- 6
Lin12a
ccatacat
TT
0.04 +- 0.04
Lin12b
ccaAaGTt
TT
0.02 +- 0.01
Lin12c
cTaAaGTt
TT
47 +- 11
REFERENCES
Return