Fission yeast genes which disrupt mitotic chromosome segregation when
overexpressed
Fission yeast genes which disrupt mitotic chromosome segregation when overexpressed
Jean-Paul
Javerzat*
,
Gwen
Cranston
and
Robin C.
Allshire
M.R.C. Human Genetics Unit, Western General Hospital, Crewe Road,
Edinburgh
EH4 2XU,
UK
Received August 21, 1996;
Revised and Accepted October 15, 1996
DDBJ/EMBL/GenBank accession nos U50769, U73337, L42550, L42551
ABSTRACT
An interference assay has been devised in
Schizosaccharomyces pombe
to rapidly identify and clone genes involved in chromosome segregation. Random
S.pombe
cDNAs were overexpressed from an inducible promoter in a strain carrying an
additional, non-essential minichromosome. Overexpression of cDNAs derived from four genes,
two known (
nda3
+
and
ubc4
+
, encoding
[beta]
-tubulin and a ubiquitin conjugating enzyme, respectively) and two unknown,
named
mlo2
+
and
mlo3
+
(
m
issegregation &
l
ethal when
o
ver expressed) caused phenotypes consistent with a failure to segregate
chromosomes. Full overexpression of all four cDNAs was lethal. Cells
overexpressing
nda3
+
and
ubc4
+
cDNAs arrested with condensed unsegregated chromosomes and cells overexpressing
mlo2
+
displayed an asymmetric distribution of nuclear chromatin. Sublethal levels of
overexpression of
nda3
+
,
ubc4
+
and
mlo2
+
cDNAs caused elevated rates of minichromosome loss. A third cDNA
mlo3
+
, displayed no increase in the frequency of minichromosome loss at sublethal
levels of overexpression but full overexpression caused a complete failure to
segregate chromosomes. Our results confirm the assumption that
[beta]
-tubulin overexpression is lethal in
S.pombe
, implicate
ubc4
+
in the control of metaphase-anaphase transition in fission yeast and finally
identify two new genes,
mlo2
+
and
mlo3
+
, likely to play an important role for chromosome transmission fidelity in
mitosis.
INTRODUCTION
Mitotic chromosome segregation is a highly accurate process ensuring the
conservation of chromosomal euploidy of dividing eucaryotic cells. Successful
chromosome segregation relies on the correct execution of a number of processes
that include DNA replication and repair, chromosome condensation, spindle
formation, kinetochore attachment to the spindle, sister chromatid separation
and subsequent chromosome movement to the poles, followed by spindle elongation
and chromosome decondensation. Defects in any of these processes or alteration
in their temporal order can lead to aneuploidy and generally cell death.
Many of the genetic approaches first used in
Saccharomyces cerevisiae
have been proven successful in
Schizosaccharomyces pombe
. Conditional lethal mutations identified several fission yeast genes that
function in chromosome transmission. For example, the nuclear-division-arrest (nda) mutant group identified the
nda2
+
and
nda3+
loci encoding [alpha]- and [beta]-tubulin, respectively (
1
-
2
) while
cut7
+
, a member of the
cut
(
c
ell
u
ntimely
t
orn) mutant family was found to encode a kinesin related motor protein (
3
-
4
). A search for mutations causing elevated rates of a tester minichromosome
defined another group of mutants (designated
mis
) among which several of them are affected for mitotic transmission of
chromosomes (
5
). It is very likely that many of the genes important for mitosis will be
uncovered from the above mutant screens. However, identifying all these genes
by mutation has some limitations. As pointed out previously by Meeks-Wagner
et al.
(
6
), not all genes are equally susceptible to mutations conferring
thermosensitivity and many genes exist in two or more copies in the genome
therefore hampering their identification by recessive mutation. To circumvent
these limitations, they set up a new approach based on gene overexpression. The
rationale being that mitotic structures, such as the mitotic chromosome, the
kinetochore or the mitotic spindle are likely to consist of multisubunit
complexes of interacting polypeptides which must be present in precise amounts
to achieve assembly. A change in the stoichiometry of some of these components
may lead to the disruption of these structures, resulting in elevated levels of
chromosome loss, non-disjunction events and increase in ploidy. Indeed the true success of the
original screen has only recently been realised with the demonstration that the
MIF2
gene product interacts with
S.cerevisiae
centromeres and shares protein sequence similarity with the mammalian
kinetochore protein CENP-C (
7
-
8
).
With the aim of identifying genes encoding structural components of mitotic
structures such as the spindle, the mitotic chromosome and the
centromere/kinetochore protein complex we adapted a similar gene dosage
strategy to fission yeast.
MATERIALS AND METHODS
DNA sequences
The new sequences reported in this paper have been deposited in the GenBank
database with the following accession numbers; U50769 (
rpc19
+
), U73337 (
mlo1
+
), L42550 (
mlo2
+
) and L42551 (
mlo3
+
).
Strains and plasmids
The strain used in this study is FY205:
h
-
, ura4 D-18, leu1-32, ade6-210, Ch16 (ade6-216).
The
S.pombe
cDNA library was provided by C. J. Norbury and B. Edgar (ICRF, Cell Cycle
Group, Oxford). cDNAs derived from exponentially growing
S.pombe
cells are directionally cloned into the
S.pombe
expression vector pREP3X (
9
). In this plasmid, expression is controlled by the thiamine regulated
nmt
promoter (
10
). [beta]-tubulin cDNAs were isolated from the library by colony hybridisation
using a fragment of the
nda3
+
gene as a probe. Partial 5' end sequencing allowed the identification of a full length clone called
pREP
nda3
+
.
Media and transformation
Media were essentially as described (
12
). When required, media were supplemented with uracil, leucine or adenine at a
final concentration of 75 mg/l each. Minimal medium with limiting amounts of
adenine contained 0.15* the standard concentration (EMG 0.15* ade). For transformation, the lithium acetate procedure (
12
) was used. Transformants were plated onto minimal plates with appropriate supplements and grown at 32oC.
Screening for overexpression lethal cDNAs
The cDNA library was transformed into the recipient strain and cells plated onto
EMG with uracil (EMGura) containing 20 [mu]M thiamine to repress expression from
nmt
but lacking leucine and adenine to select for transformants and keep selection
on the
Ch16
minichromosome. To identify transformants unable to grow upon
nmt
driven overexpression, colonies were replica plated twice at 24 h interval onto
medium lacking thiamine (EMGura) and clones that were unable to grow after the
second replica were collected.
Chromosome loss assay
Transformed cells from EMGura 20 [mu]M thiamine plates were dispersed into 0.5 ml of sterile water at a density
of ~10
6
cells/ml. An aliquot of each cell suspension was placed in the first row of a
microtitre plate and five 1:5 serial dilution were performed. Aliquots (10 [mu]l) of each dilution were then spotted onto EMGura 0.15* ade, EMGura 0.15* ade 0.01 [mu]M thiamine, EMGura 0.15* ade 0.05 [mu]M thiamine and EMGura 0.15* ade 20 [mu]M thiamine. After 4-5 days at 32oC, colonies were
examined by light microscopy to detect the presence of red sectors indicative
of
Ch16
loss events.
DAPI staining and immunofluorescence microscopy
Cell fixation was essentially as described previously (
13
). Cells from 1 ml of exponentially growing culture were fixed with ice-cold 2.5% glutaraldehyde for 10 min, washed twice with water and
resuspended in 20 [mu]l of 0.1% sodium azide. Five millilitres of cell suspension were mixed with
an equal volume of 1 [mu]g/ml DAPI solution, spread onto a poly-l-Lysine coated slide and mounted in Vectashield mounting medium
(Vector labs). Cells were observed with a Zeiss Axioplan fluorescence
microscope and pictures were taken using a KODAK TMAX 400 black and white film.
Flow cytometry
Ethanol fixed cells were treated for DNA content analysis (
14
). The analysis was carried out using a Becton-Dickinson FACScan and the software CELLFIT.
DNA methods
Plasmids from
S.pombe
transformants were recovered in
Escherichia coli
DH5
[alpha] using the extraction procedure previously described (
15
). Sequence data from the 5'end of cDNA inserts was obtained by solid-phase DNA sequencing essentially as described by KERR
et al
. (
16
). Sequence data from the 3'end of cDNA inserts was obtained by the double-stranded dideoxy method (
18
) described for Sequenase Version 2.0 (USB). Complete sequence of cDNA inserts
was achieved by a combination of primer walks and subcloning of restriction
fragments into Bluescript SK
-
(Stratagene).
RESULTS
Experimental design
The strategy adopted was designed to identify genes which interfere with mitotic
chromosome segregation when overexpressed. A decrease in chromosome stability can be monitored by the use of a marker
chromosome. A visual screen of haploid
S.pombe
cells undergoing chromosome loss events is possible by the use of a tester
strain carrying the
Ch16
minichromosome, a 500 kb chromosome III derivative produced by irradiation of a
S.pombe
strain disomic for chromosome III. The
Ch16
minichromosome retained all centromere 3 sequences and therefore is very stable
in mitotically dividing cells (the loss rate is 10
-4
) and segregates normally through meiosis (
19
). The minichromosome carries the
ade6-216
`pink' allele so that in a strain bearing the
ade6-210
`red' allele on chromosome III, colonies are ade
+
and white due to intragenic complementation. Loss of the minichromosome results
in adenine prototrophy and to the development of red coloured colonies on a
medium containing a limiting amount of adenine. Therefore, if loss events occur
during colony growth these are visualized as red sectors on a white background.
Gene overexpression was achieved by using a library of
S.pombe
cDNAs directionally cloned in pREP3X (
9
) under the control of the thiamine regulated
nmt
promoter first described by Maundrell (
10
). Expression from the
nmt
promoter is induced on medium lacking thiamine and repressed on medium
containing 20 [mu]M thiamine. Uptake of thiamine in
S.pombe
is known to be very efficient (up to 9000 pmol/10
7
cells) and transcription from the
nmt
promoter is rapidly repressed. Upon transfer to medium lacking thiamine, the
intracellular pool decays by dilution with increasing mass and reactivation of
nmt
occurs when the intracellular thiamine concentration falls below 50 pmol/10
7
cells (
11
). Upon induction, the
nmt
promoter is very strong, giving rise to expression levels similar to the
S.pombe
adh
promoter and 1000-fold greater than repressed levels (
9
-
20
).
The screening procedure was tested by the overexpression of two
S.pombe
cDNAs derived from genes encoding components of the mitotic spindle :
nda3
+
, encoding [beta]-tubulin (
2
) and
cut7
+
, encoding a kinesin related motor protein (
3
). A full length [beta]-tubulin cDNA (pREP
nda3
+
) was isolated from the cDNA library by colony hybridisation and the
cut7
+
ORF cloned in pREP1 (pREP
cut7
+
) was also utilised. The two plasmids were introduced into the tester strain and
individual transformants tested for minichromosome stability by spotting serial
dilution of cells onto media allowing the development of red sectors upon loss
of the minichromosome and containing various thiamine concentrations. As shown
in Figure
1
, all strains grew perfectly well under repressed conditions (20 [mu]M thiamine). In contrast, only pREP3X transformed cells were able to form
colonies on medium lacking thiamine, showing that both
cut7
+
and
nda3
+
gene products are lethal when overexpressed. However, colony growth could be
rescued on media containing low thiamine concentrations. Examination of these
slow growing colonies revealed the presence of red sectors indicating that
overexpression of
cut7
+
and
nda3
+
decreased the stability of the marker minichromosome. This also indicates that
overexpression from
nmt
can be set (at least in some cells in the colony) to a level high enough to
induce chromosome loss but below the lethal threshold since the cell survived
and formed the red lineage. It is unclear how such intermediate levels of
expression can occur on low thiamine concentrations. Some cells may be
restricted for thiamine availability owing to their position in the colony and
have their internal thiamine pool oscillating around the threshold for
nmt
activation.
Screening procedure
The steps of the screen are depicted in Figure
2
. The cDNA library was introduced into the tester strain by transformation.
Colonies were replica plated twice at 24 h interval onto medium lacking
thiamine to turn on expression from the
nmt
promoter and clones not able to form colonies upon full induction were
selected. The rationale behind this was as follows; first, the experiment
described above showed that overexpression of known components of the spindle
inhibited cell growth and second, since
S.pombe
is highly sensitive to aneuploidy (
21
), it is likely that any cDNA interfering with chromosome segregation will be
lethal upon prolonged full induction of the expression. Therefore, this primary screen for lethality should provide a substantial enrichment in clones bearing cDNAs of interest. From ~15 000 transformants, 133 clones displayed overexpression induced growth
arrest and were selected. This collection of strains was subsequently screened
for mitotic defects induced by overexpression. The chromosome loss assay was
used to identify cDNAs that decrease chromosome stability upon sublethal conditions of overexpression and finally, cytological observation was used to look for mitotic defects upon
overexpression induced lethality. In addition to this main screening strategy,
the collection of conditional lethal strains was also screened for
cytologically visible mitotic defects upon full overexpression (Fig.
2
, dotted arrow). This additional step was devised to uncover cDNA clones which
when overexpressed cause a complete failure to segregate chromosomes. Such a defect is expected to
be lethal and therefore should escape the chromosome loss screen but should be
readily detected by the cytological examination of overexpressing cells.
cDNAs that decrease chromosome stability when overexpressed
The 133 clones were tested individually for minichromosome stability by spotting
serial dilutions of cells onto media containing 0, 0.01, 0.05 and 20 [mu]M thiamine. As expected, normal sized colonies developed on 20 [mu]M thiamine plates, colonies of reduced size formed on plates containing
0.01 and 0.05 [mu]M thiamine, and no growth or very limited growth (microcolonies) was
observed on plates lacking thiamine. Nineteen transformants were selected that
displayed a thiamine suppressible sectoring phenotype. To test whether the
sectoring phenotype was plasmid dependent, total DNA was prepared from each
transformant, the plasmids were isolated by transformation of
E.coli
and then reassayed for the sectoring phenotype by transforming the tester
strain. Seventeen plasmids were rescued successfully but restriction analysis
revealed that four of them had undergone structural rearrangements. From the
remaining 13 plasmids, two failed to reproduce the original phenotype but 11
induced sectoring at about the same frequency as displayed by the original transformants. An example
of the sectoring phenotype is shown in Figure
3
, illustrating phenotypes ranging from rare sectoring to extremely frequent
sectoring. In each case, the sectoring phenotype was suppressed by an excess of
thiamine, very rare sectored colonies being detectable on medium containing 20 [mu]M thiamine. To see if these cDNAs were derived from known genes, the cDNA
inserts from these 11 clones were partially sequenced from the 5' end and converted by translation into amino acid sequences. In all
cases, a start codon followed by an uninterrupted open reading frame was found
and the corresponding predicted peptides were used to search for similarities
in data bases using the BLAST program (
22
). Of 11 cDNAs, five were derived from previously characterized
S.pombe
genes:
nda3
+
, encoding [beta]-tubulin (
2
),
act1
, encoding actin (
23
), and
fib
, encoding fibrillarin (
24
). Among them, [beta]-tubulin was recovered three times while actin and fibrillarin only
once. These cDNAs were not further studied.
Phenotypes upon overexpression induced lethality
Our primary interest is in genes required for mitotic chromosome segregation. A
number of cDNAs derived from known and unknown
S.pombe
genes have been identified using chromosome loss as a primary phenotype.
However, defects in many cellular processes can lead to chromosome loss. These
include DNA replication and repair, chromatin assembly, recombination, all of these processes are distinct from mitotic specific events. Therefore, secondary screens are required to focus on genes with a mitosis specific
function. The isolated cDNAs decrease the stability of the marker minichromosome upon sublethal conditions of
overexpression and arrest cell proliferation upon full induction of the
nmt
promoter. If both phenotypes are due to a defect in chromosome segregation,
cytological observation of arrested cells should reveal a cell cycle block at a
particular mitotic stage. Therefore, we used the DNA stain DAPI (4', 6-diamidino-2-phenylindole dihydrochloride hydrate) to look at
overexpression induced mitotic defects. The cDNAs bearing strains were grown to
early log phase in liquid medium containing 20 [mu]M thiamine, cells were washed in minimal medium, and resuspended in fresh
minimal medium alone (induced conditions) or in minimal medium supplemented
with 20 [mu]M thiamine (uninduced control). It has been shown previously that
transcription from the
nmt
promoter starts ~10 h after transfer to medium lacking thiamine to reach full activity after
~16 h (
10
). Indeed, a preliminary time course experiment showed that the doubling time of
pREP
nda3
+
transformed cells started to decrease 10 h after transfer to minimal medium
lacking thiamine and these cells virtually stopped dividing after 15-17 h, indicating a good correlation between the rise in
nmt
trancriptional activity and the resulting effect on cell growth. Therefore, cytological observations were performed 17 h after transfer onto medium lacking thiamine. Cells were fixed with
glutaraldehyde, stained with DAPI and observed by fluorescence microscopy (Fig.
4
). Cells transformed with the control plasmid (pREP3X) grown with (Fig.
4
a) or without (Fig.
4
b) thiamine displayed patterns typical of wild type
S.pombe
cells. The population is composed of a majority of interphase cells with a
single nucleus displaying the hemispherical chromatin domain, ~10% of mitotic cells showing condensed chromatin or dividing nuclei and ~5% of septated cells having completed nuclear division. In contrast, a population of cells overexpressing [beta]-tubulin (Fig.
4
d) showed a decrease in interphase cells (18.5%), while cells displaying condensed chromatin (Fig.
4
d) increased up to 80%. Interestingly, this phenotype is similar to that of
nda3-KM311
cells bearing a cold sensitive mutation in the [beta]-tubulin gene. At the restrictive temperature,
nda3-KM311
cells accumulate with condensed chromosomes at a stage similar to the mitotic
prophase of higher eukaryotes (
2
). We conclude that overexpression of [beta]-tubulin has a similar effect, causing cells to arrest at an early
stage of mitosis.
Screening the collection of conditional lethal strains by fluorescence
microscopy
In order to identify genes involved in chromosome segregation the above
collection of thiamine dependent conditional lethal transformants was screened
for chromosome loss upon sublethal levels of expression. However, a complete
failure to segregate the whole set of chromosomes is not expected to be
detected by the chromosome loss assay and thus these cDNAs would have escaped
the above screen. This class of cDNAs should be recovered by visual screening
of overexpressing cells for the complete failure to segregate DNA using
fluorescence microscopy. From the collection of 133 thiamine dependent
conditional lethal strains, 20 were selected which showed the most severe
inhibition of growth upon induction of the
nmt
promoter. The strains were grown to early log phase in liquid medium containing
thiamine, cells were washed and resuspended in fresh minimal medium alone
(induced conditions) or in minimal medium containing 20 [mu]M thiamine (uninduced control). After 20 h, cells were fixed with
glutaraldehyde, stained with DAPI and observed by fluorescence microscopy. Two
strains were selected which displayed a strong, thiamine suppressible mitotic
defect. The chromosome loss assay performed upon sublethal levels of
overexpression (0.01 and 0.05 [mu]M thiamine) failed to reveal any increase in minichromosome loss rate.
Plasmid recovery followed by limited sequencing showed that cDNA50 and cDNA72
were identical. Database searches showed that they define a new fission yeast
gene which we named
mlo3
+
.
As shown in Figure
6
, cells transformed with pREP
mlo3
+
are very similar to
wt
cells when grown in the presence of thiamine (Fig.
6
a): the population is composed of a majority of interphase cells (81%), ~11% of mitotic cells displaying condensed chromatin or separating nuclei
and ~8% of septated cells. In contrast, the percentage of mitotic cells dropped
to <1% upon induction but the most striking phenotype is the accumulation of
septated cells in which one daughter cell is devoid of a nucleus (Fig.
6
b). Such cells represent ~29% of the population, completely anucleate cells ~16% while the fraction of apparently normal interphase cells dropped to ~46%. This phenotype is consistent with a complete failure to
segregate chromosomes, resulting in two daughter cells, one containing an
undivided nucleus, the other no nucleus at all. FACS analyses (Fig.
7
) indicated that cells had a normal G2 DNA content until 16.5 h after transfer
to medium lacking thiamine. At 19 h, two new populations of cells were
detected: one with <1C DNA content, likely to represent the fraction of anucleate cells, and one
with ~4C DNA content. The latter might correspond to cells where cytokinesis and
replication occurred in the absence of chromosome segregation giving rise to
daughter cells with increased and decreased DNA content.
The cDNA was fully sequenced (accession number L42551). The predicted 199 amino
acid peptide is very basic (pI: 10.91). Data base searches failed to identify
proteins with high similarity to Mlo3p.
DISCUSSION
In order to identify fission yeast genes involved in chromosome segregation we
designed a screen based on gene overexpression. A cDNA library under the
control of the inducible
nmt
promoter was screened for cDNAs that cause cell death upon full induction of
the expression and induce elevated rates of chromosome loss at sublethal levels
of expression. Eight genes were identified (Table
1
). Some of them like actin, fibrillarin or the
S.pombe
homologue of the
S.cerevisiae
shared RNA polymerase I and III subunit gene
RPC19
are obviously not good candidates for having a direct role in the control of
chromosome segregation. Their appearance in the screen may result from
pleiotropic effects of their overexpression rather than a specific alteration
of a particular stage in chromosome segregation. To focus the screen on genes
important for mitosis, microscopic observation of DAPI stained cells was used
to identify cytologically visible mitotic defects. Four genes survived this
last step of the screening process: the
nda3
+
gene, encoding [beta]-tubulin (
2
), the recently identified
S.pombe
ubc4
+
gene (
27
) and two new genes (
mlo2
+
and
mlo3
+
) which do not show any significant similarity to known gene products.
Overexpression of [beta]-tubulin is known to have deleterious effects in
S.cerevisiae
: transient overexpression induced a high rate of chromosome loss while
continuous overexpression was lethal, causing cells to arrest in the G
2
stage of the cell cycle, a phenotype very similar to the phenotype of [beta]-tubulin deficiency (
31
). It was suspected that [beta]-tubulin overexpression might have the same effect in
S.pombe
(
2
). Indeed, we found that sublethal levels of overexpression caused an elevated
rate of chromosome loss and full induction of the expression was lethal,
causing cells to arrest with condensed, unsegregated chromosomes. This
phenotype is similar to the prometaphase-like arrest seen in cells bearing the cold sensitive
nda3-KM311
mutation in the [beta]-tubulin gene when shifted to the restrictive temperature (
2
). Therefore, overexpression or deficiency of the [beta]-tubulin gene has a very similar phenotype in both budding and
fission yeast.
Overexpression of
ubc4+
was found to induce a high rate of chromosome loss upon sublethal levels of
expression and to cause cell death upon full induction. Cells undergoing the
lethal event had a G2 DNA content and a high proportion showed condensed,
unsegregated chromosomes or the `
cut
' phenotype, demonstrating that overexpression of
ubc4
+
prevents chromosome segregation. The ubiquitin system is responsible for the
proteolysis of many critical regulatory proteins that must be rapidly
destroyed. Proteins are marked for degradation by the covalent attachment of
multiple molecules of the polypeptide ubiquitin. A complex multistep pathway
leads to ubiquitination. Ubiquitin is first activated by thioester formation
with E1, the ubiquitin-activating enzyme. Subsequently, E1 transfers ubiquitin to a family of
ubiquitin conjugating enzymes (E2s), again forming thioester intermediates.
Ubiquitination of the substrate usually requires a third activity, the
ubiquitin protein ligase or E3 that can directly mediate substrate specificity
(reviewed in
32
-
33
). Proteolytic degradation of cyclin B, an event that triggers exit from
mitosis, is mediated by the ubiquitin pathway. In addition to this role, recent
evidence suggests that the same machinery triggers the metaphase-anaphase transition by the proteolysis of a factor that prevents sister
chromatids separation (
34
-
37
). Because this proteolytic machinery is required for anaphase in yeast and
mammalian cells, it was called APC for anaphase-promoting complex (
38
). Genetics and biochemical evidence have shown that UBC4 is part of this
complex (
36
-
38
). Our results demonstrate that overexpression of
ubc4
+
causes
S.pombe
cells to fail to segregate their chromosomes at mitosis. It is therefore likely
that Ubc4p is a component of
S.pombe
APC. An excess of Ubc4p could interfere with APC function either by preventing
its assembly or by depleting other components of the ubiquitin pathway
resulting in a failure to dissolve the attachments between sister chromatids.
The present study identified two new
S.pombe
genes,
mlo2
+
and
mlo3
+
that interfere with mitotic chromosome segregation when overexpressed. Sequence
analysis did not give any insight into the function of the corresponding genes. Therefore, additional experiments will be needed to address their cellular functions.
Compared with similar screens performed in budding yeast, it appears that many
more fission yeast genes interfere with chromosome segregation and/or cause
cell death when overexpressed. The overexpression screen performed by Meeks-Wagner
et al
. (
6
) using a genomic library identified only two sequences and only a very limited
number of sequences in a GAL1-regulated cDNA library showed overexpression induced lethality (
39
). This observation, in addition to the fact that
S.pombe
is much more sensitive to aneuploidy (
21
) than
S.cerevisiae
suggests that gene dosage is more critical for cell viability in fission yeast.
Clearly, the screen described could be increased in scale to identify less
abundant cDNAs or alternatively, normalised cDNA libraries could be utilised to
try and enrich for rare cDNAs.
ACKNOWLEDGEMENTS
We are grateful to K. Ekwall, C. Gordon and N. Hastie for critical reading of
the manuscript, N. Davidson, S. Bruce and D. Stewart for photographic and art
work, C. J. Norbury and B. Edgar for providing the cDNA library and I. Hagan
for giving the pREP
cut7
+
plasmid. This work was supported by the Medical Research Council of Great
Britain and the french Centre National de la Recherche Scientifique. J.P.J. was
supported by Wellcome and European Union Human Capital and Mobility Program
fellowships, respectively.
REFERENCES
1 Toda, T., Adachi, Y., Hiraoka, Y. and Yanagida, M. (1984) Cell, 37, 233-242.MEDLINE Abstract
2 Hiraoka, Y., Toda, T. and Yanagida, M. (1984) Cell, 39,349-358.
3 Hagan, I. and Yanagida, M. (1990) Nature, 347,563-566.
4 Hagan, I. and Yanagida, M. (1992) Nature, 356,74-76.
5 Takahashi, K., Yamada, H. and Yanagida, M. (1994) Mol. Biol. Cell, 5, 1145-1158.MEDLINE Abstract
6 Meeks-Wagner, D., Wood, J. S., Garvik, B. and Hartwell, L. H. (1986) Cell, 44, 53-63.
7 Meluh, P. B. and Koshland, D. (1995) Mol. Biol. Cell, 6, 793-807.
