Copyright 0 1994 by the Genetics Society of America
Synergistic Release From Glucose Repression by migl and ssn Mutations in
Saccharomyces cereuisiae
Laura G. Vallier' and Marian Carlson
Department of Genetics and Development and Institute of Cancer Research, College of Physicians and Surgeons,
Columbia University, New York, New York 10032
Manuscript received November 24, 1993
Accepted for publication January 24, 1994
ABSTRACT
Saccharomyces cerevisiae, glucose repressionof SUC2 transcription requires the SSNGTUPI
In the yeast
repressor complex. It has been proposed that the DNA-binding proteinMIGl secures SSNGTUPI to the
SUC2 promoter. Herewe show that a migl deletion does not cause nearly as dramatic
a loss of repression
as ssnb: glucosegrown migl mutants display 20-fold lower SUCB expression than ssn6 mutants. Thus,
repression by SSN6TUPl is not mediated solelyby MIGl, but also involves MIGl-independent mechanisms. We report thatmigl partially restoresSUCB expression in mutants lacking the
SNFl protein kinase
and show that migl is allelic to s s n l , a mutation selected as a suppressor of s n f l . Other SSN genes
role in repression of SUC2. We show that migl
identified in this selection were therefore candidates afor
acts synergistically with ssn2 through ssn5, ssn 7, and ssn8 to relieve glucose repressionof SUCB and to
suppress the requirement for
SNFl. These findings indicate that SSN
the proteins contribute
to repression
of SUC2, and the pleiotropic phenotypesof the ssn mutants suggest global roles in repression. Finally,
the regulated SUC2 expression observed in snfl migl mutants indicates that signals regarding glucose
availability can be transmitted independently of the SNFl protein kinase.
I
N the yeast Saccharomycescerevisiae, mutations in
SSN6 ( CYCS) or T U P l affect negative regulation of
gene expression and cause pleiotropic defects in a variety of cell
processes, including glucose repression, mating,
sporulation
and flocculation (for review, see
JOHNSTON and CARLSON 1992; TRUMBLY
1992). Genetic
and biochemical evidence suggests that SSNG and TUPl
proteins function together
as a general repressor of transcription. The two proteins arephysically associatedwith
one another (WILLIAMS
et al. 1991), and a LexASSN6
fusion protein, bound to DNA via a lexA operator, represses target gene expression by a TUPldependent
mechanism (KELEHER et al. 1992). SSNG and TUPlhave
no apparent DNA-binding domains (SCHULTZ
et al. 1990;
WILLIAMS
and TRUMBLY 1990),
and KELEHER et al. (1992)
postulated that the SSN6-TUP1 complex is tethered to
the promoter region of specific genes via interaction
with gene-specific DNA-binding proteins. They provided evidence that SSNG and TUPl are recruited tocell
type-specific genes by the al, 02 and MCMl DNAbinding proteins.
In this study, we address the mechanism by which
SSNG and TUPl effect glucose repression of the SUC2
(invertase) gene. KELEHER et al. (1992) suggested that
MIGl might tether the
SSNGTUPl complex to theregulatory regions of SUC2 and other glucose-repressible
genes. The MZGl gene was isolated as a multicopy inhibitor of the GALl promoter (NEHLINand RONNE
'
Current address: Department of Biology, Yale University, New Haven,
Connecticut 06511.
Genetics 137: 49-54 (May, 1994)
1990). MIGl encodes a zinc-finger protein thatbinds to
the SUC2, GAL4, and GALl promoters, and deletionof
the gene partially relieves glucose repression (NEHLIN
and RONNE
1990; NEHLIN
et al. 1991;F L I C K ~ ~ ~ J O H N S T O N
1992).
Here we examine the role of MIGl in mediating repression of SUC2 by SSN6-TUP1. Our genetic studies
are consistent with the postulated role of MIGl asa tethering protein. However, our data also suggest that repression bySSNG involves MIG1-independent mechanisms because the deletion of MZGl does not cause
nearly as dramatic a loss of repression of SUC2 as deletion of SSN6.
Genetic evidence suggested additionalcandidate
genes that might be involved in repression of SUC2. A
selection that yielded ssn6 mutations also identified
seven other loci: ssnl through ssn5, ssn7, and ssn8 ( s s n ,
for suppressor of s n f l ) (CARLSON et al. 1984). These
mutations suppress growth defects of mutants lacking
the SNFl (CATl/CCRl) protein kinase, which is required forrelease from glucose repression ( CELENZA
and
CARLSON 1986; SCHULLER
and ENTIAN1987). The ssn mutations alleviate the SUCB derepression defect of a snfl
mutant to varyingextents,and all except ssnl cause striking
clumpiness. None is allelic to t u p 1 (L. NEIGEBORN
and
M. CARLSON, unpublished results).
In this study we show that a m i g l A mutation partially
suppresses s n f l A mutant defects and that M I G l is the
same gene as SSNl. We present evidence that m i g l acts
synergistically withssn2 through ssn5, ssn 7 and ssn8 to
50
L. G . Vallier and M. Carlson
TABLE 1
Strains used in this study
Straina
MCY363
MCY438
MCY442
MCY447
"456
MCY479
Genotype
M A T a snfl-28ssn3-1ade2-101gal2f
SUC2
M A T a snfl-28ssn5-4ura3-52his4-539gal29
M A T a snfl-28ssn4-lade2-101gal2f
SUC2
M A T a snfl-28ssn2-4ade2-101ura3-52his4-539
SUCP
M A T a snfl-28ssn7-1ade2-101his4-539gal2
M A T a snfl-28ssn8-1ade2-101lys2-801ga12?
SUC2
SUC2
suc2
MCY888
MCYl094
MCYl253
MCYl594
MCYl595
MCYl705
MCYl752
MCYl773
MCYl920
MCYl921
MCYl941
MCYl943
MCYi958
MCY3308
MCl3311
CSH91L
YM3920b
M A T a snf4Alura3-52lys2-801
ga12S SUC2
M A T a ura3-52
ade2-101 SL'C2
MATa ssnl-6lys2-801ade2-101
SUC2
M A T a snflA3 leu2::HIS3ura3-52lys2-801ade2-101
suc2
M A T a snflA3ura3-52his4-539lys2-801his3A200
suc2
MATa ssn6A5::URA3ura3-52his4-539lys2-801
suc2
MATa snflA3ura3-52ade2-101his4-539gal2
suc2
M A T a snflA3ssnl-6ura3-52lys2-801
SUCZ
M A T a migIAZ::LL?U2 ura3-52lys2-801leu2::HIS3
SUCP
MATa miglAP::LEU2ura3-52his4-539 SUCP
M A T a snflA3miglA2::LEU2ura3-52his4-539
lys2-801 ade2-101 ga12f SUC2
MATa snflA3miglA2::LEU2ura3-52his4-539
gal2S SUC2
MATa ssn6A9ura3-52his4-539lys2-801leu2-3,112
suc2
MATa snflA3 or -28ssn3-1miglA2::LEU2ura3-52
his4-539 gal2? SUC2
M A T a snflA3 or -28ssn4-1miglAP::LEU2ura3-52
his4-539 lys2-801? ga12S SUCP
MATa ura3adel his1 leu2 lys7 met? trp5 SUC+
M A T a miglAP::URA3ura3-52his3A200ade2-101
lys2-801 ga180-538 SUC2
All strains have the S288C genetic background except CSH91L.
Obtained from MARK JOHNSTON
relieve glucose repression of SUC2and to suppressthe
requirement for SNF1.
MATERIALSANDMETHODS
Yeast strains: All strains have the S288C genetic background except where noted (Table 1). The MZGl locus was
disrupted by integration of the miglA2 allele (NEHLINand
RONNE
1990)which replaces codons 67-460 with LEU2 (called
here miglA2::LEU2) or URA3(miglA2::URA3). Disrup
tions were verified by Southern analysis. For ssn mutations
other than ssn6, triple mutants (snfl migl ssn) and double
mutants (snfl migl; snflssn) were segregants from crosses of
MCY447, MCY363, MCY442, MCY438, MCY456 or MCY479 to
MCYl943 carrying pCE9 or were derived from such segregants by further crossing. pCE9 contains SiVFI cloned in
YEp24, which carries URA3 (CELENZA and
CARLSON 1989).
The desired strains were initially identified by2-deoxyglucose (2DG) resistance and clumpy growth (migl and ssn
phenotypes, respectively), and genotypes were confirmed by
complementation after plasmid loss. To select for loss of
pCE9, spore clones were patched onto WD and replica
plated or streaked onto medium containing 5-fluoroorotic
acid (5-FOA); lossof the plasmid was confirmed by testing
on SD-Ura medium.
Strains of genotype snfl migl, migl ssn6, snfl ssn6or snfl
migl ssn6 were recovered from a cross between M a 9 4 1
( s n f l A 3 m i g l A 2 : : L E U Z )and MCYl958 ( s s n 6 A 9 ) .The snfl
mutation was followed by scoringgrowth on glycerol or
galactose. Strains were transformed with pCE9 for assays in
Figure 2.
Genetic methods and media: Standard methods were used
for genetic analysis and transformation (ROSEet al. 1990).Utilization of sugars was scored by anaerobic growth on solid rich
medium (YEP) containing 2% glucose (WD), sucrose (WS),
or raffinose (YPR).Utilizationofglycerolwas
scored on
YEP-3% glycerol. Strains with plasmids were scored
on selective
synthetic mediumcontaining 2% glucose (SD)or sucrose (SS)
or selective syntheticcomplete medium with 2% glucose (SC)
,
sucrose (SCS) or raffinose (SCR).Cell suspensions were spotted onto solid medium and incubated at 30". Clumpy strains
were suspended in 20 mM EDTA prior to spotting to disperse
the clumps. Resistance to
2DG was scored by anaerobic growth
on WS or SCS medium containing 200 pg/ml2DG.
Invertase assays: Glucose repressed and derepressed cells
were prepared from exponentially growing cultures as deand CARISON 1991). For derepresscribed previously (VALLIER
sion, cells were shifted to medium containing 0.05% glucose
for 2.5 hr. Strains containing plasmids were grownin SGUra.
For clumpy strains, cell density
was measured after addition of
EDTA to 20 mM. Secreted invertase was assayed in whole cells,
and activity isexpressed as micromoles of glucose releasedper
minute per 100 mg (dry weight) of cells.
RESULTS
MIG1-independent mechanisms contribute to repression of SUC2 by SSNG: KELEHER et al. (1992) suggested
that the MIGl protein mightserve to tether the SSN6TUPl complex tothe SUC2 regulatory region to effect
transcriptional repression. This simple model predicts
that a migl mutation should relieve glucose repression
of SUC2 to the same extent asan ssn6 mutation.To test
this prediction, we compared the effects of miglA and
ssn6A mutationsinthe
S288C geneticbackground.
miglA mutants grown in 2% glucose produced about
10%of the wild-type derepressed invertaseactivity. NEHLIN and RONNE(1990) also reported
a partial release
from glucose repression in migl mutants of the W303
strain background.In contrast, glucose-repressedssn6A
mutants produced greater
activity than derepressedwild
type (Table 2). Thus, migl is much less effective than
ssn6 inrelieving glucose repression, indicating that the
MIGl protein does not mediate the major repressive
data do not
effect ofSSNG o n SUC2 transcription. These
by SSN6-TUP1 is meexclude the model that repression
diated partly via MIG1.
migl partially suppresses defects of snflA mutants:
Mutations in SSNG or TUPl allow SUC2 expression in
a snfl mutant, suggesting that the SNFl kinase activity
is required for release of repression by the SSN6-TUP1
proteins (SCHULTZand CARLSON
1987; WILLIAMS
and
TRUMBLY
1990). If MIGl is involved in this repression
mechanism, a m i g l mutation would be predicted to
function asa suppressor of snfl.
To test this idea,
we carried out tetradanalysis of a cross heterozygous forboth
miglA2::LEU2andsnflA (MCYl752 X MCYl920).The
of
Synergy
migl and ssn Mutations
TABLE 2
Strain
51
Plasmid
glucose
sucrose
migl is less effective than ssn6 in relieving glucose
repression of invertase
Invertase activity"
Relevant genotype
Repressed
Derepressed
Wild type
297
miglA2::URA3
miglA2::LEUZ
ssn6A5::URA3
ssn6A9 1650
ssn6A9 miglA2::LEUZ
1350
1
20
30
390
493
414
284
406
1280
" Repressed and derepressed cultures were grown in rich medium
as described in MATERIALS AND METHODS. Values are averages of at least
two assays, usually of two or more different strains. Standard errors
were less than 25%.
TABLE 3
FIGURE1.-pMIG1 complements the ssnl mutation. Two
representative transformants of strains MCYl094, MCYl943
and MCYl773 carrying pMIGl or YEp24 were spotted on SDUra (glucose) andSSUra
(sucrose). Plates were phot@
graphed after 48 hr at 30". WT, wild type.
fects of a snfl point mutant and
showed varyingdegrees
ofefficacy
in restoring invertase derepression. All
caused striking clumpiness except ssn 1 , which was thereInvertase activity"
fore themost likely candidate for allelism with migl. To
assess the relationship between migl and s s n l , we first
Relevant genotype
Derepressed
Repressed
determined whether ssnl suppresses snfl and snf4 null
1
284
Wild type
mutations by constructing double mutants. The s s n l - 6
<1
<1
snflA3
allele suppressed the growth defects caused by snflA
<1
4
snf4A1
snflA3 miglA2::LEUZ
4
54
and sn f4A,conferring a Suc+ (sucrose) and Raf+ (rafsnf4Al miglA2::LEU2
3
112
finose)
phenotype, and restored derepression of inver4
35
snflA3 ssnl-6
tase to 12% and 60%of the wild-type level, respectively
snf4Al ssnl-6
7
161
ssn 1-6
7
283
(Table 3). Thus, s s n l - 6 and miglA show similar s u p
pressor phenotypes.
Values are averages of at least two assays, usually of two or more
different strains. Cultures were grown inrich medium. Standard
To test migl and ssnl for complementation, we
errors were less than 25%.
crossed snflA3ssnl-6 and snflA3miglA2::LEU2
strains (MCYl773 X MCYl943). The diploid exhibited
a SUC' Raf+ phenotype, indicating that the mutations
double mutantsegregants grew on sucrose and raffinose
fail to complement. Control s n f l / s n f l diploids het(substrates of invertase), and to some extent on galactose and glycerol. The miglA2::LEU2 mutation reerozygous for ssnl or migl wereSuc- Raf-. We also
stored invertase derepression to 20% of the wild-type
tested the cloned MIGI gene for complementation of
s s n l . Transformation of MCYl773 ( s n f l A 3 s s n l - 6 )
levelin s n f l A mutants (Table 3). In addition,the
with plasmid pMIGl (NEHLIN and
RONNE1990) caused
double mutants producedlow levels of invertaseunder
a Suc-phenotype (Figure 1) and reduced invertase derepressing conditions. In an analogous cross (MCY888
X MCYl921), we tested the effects of migl in a mutant
repression more than tenfold (data not shown).
We then tested the linkage of ssnl to trp5, which is5.4
lacking the SNF4 (CATS) protein, which is required as
a positive effector of the SNFl kinase activity (CELENZA cM from migl on chromosome VZZ (NEHLINand RONNE
and CARLSON 1989; CELENZA
et al. 1989; SCH~LLER and 1990), by tetrad analysis ofthe cross MCYl773 ( s n f l A 3
s s n l - 6 ) by CSH9lL (trp5).The snfl segregants were
ENTVW1988). The m i g l A 2 mutation was found to rescored for the ssnl-6 suppressor. Of the 38 Suc+ ( s n f l
store raffhose growth and derepression of invertase (to
ssnl ) segregants, 35 were Trp'; of the 29 Suc- ( s n f l
40% of the wild-type level) in snf4A strains (Table 3).
SSNl ) segregants, 27 were Trp-. Thus, ssnl is tightly
These data areconsistent with the view that MIGl funclinked to trp5.
tions with the SSNG-TUP1 proteins, downstream of
We next tested directly for allelism betweenmigl and
SNFl.
It is noteworthy that invertase expression was regus s n l . We crossed MCYl943(snflA miglA2::LEU2)carlated 1 4 and 37-fold in s n f l A m i g l A and snf4A miglA
ryingpCE9, a multicopy SNFl plasmid marked with
URA3 (CELENZAand CARLSON1989),to
MCYl773
mutants, respectively, inresponse to glucose availability.
( s n f l A 3 s s n l - 6and
) examined 31 tetrads in which auxMZGl is the same gene as SSNl: The suppression of
otrophic markers segregated 2+:2-. Segregants lacking
snfl mutant defects by a migl mutation raised the posthe plasmid were recovered by selection on 5-FOA. For
sibility that migl is allelic to one of the previously identified suppressors of snfl, ssnl through ssn8 (CARLSON
27 tetrads, growth on sucrose segregated 4+:0-. In four
tetrads, one segregant initially scored Suc- but upon reet al. 1984). These ssn mutations remedied growth demigl partially suppresses snfl and snf4 invertase defects
52
and
L. G. Vallier
testing freshly grown cells, displayeda SUC+ phenotype;
we have inother instances observed similar instabilityof
the ssnl suppression phenotype (unpublished results).
Therefore, all segregants in the 31 tetrads appeared to
carry a suppressor of snfl. Taken together, thecomplementation and linkage data indicate that M I G l and
SSNl are the same gene.
migl and other ssn mutations act synergistically to
relieve glucose repression: Because migl mutations relieve glucoserepression much less effectivelythan ssn6,
we considered the possibility that other proteins
besides
MIGl mediate effects ofSSNG on SUC2 expression. Mutations in the six other SSN loci were recovered in the
same selection for suppressors of snfl that yielded migl
and ssn6. Thus, thesix other SSN proteins seemed likely
candidates for such a role.
To test the idea that other SSN proteins function together with MIGl in repression, we constructed double
mutants carrying migl and each of the mutations ssn2,
ssn3, ssn4, ssn5, ssn 7 and ssn8, henceforth referredto
collectively as ssn mutations (ssn6 is considered separately below). We crossed each of the snfl ssn strains
(see MATERIALSAND METHODS) to asnflA3 miglA2::LEUZ
strain carrying pCE9, which providesthe SNFl function
required for sporulation. Segregants carrying pCE9
were assayed for invertase activity under repressing conditions (Figure 2). Strains carrying a single migl or ssn
mutation showed low level or no constitutive invertase
activity (ranging from undetectable for ssn 7 to 10 units
for ssn8 and m i g l ) . Strains containing migl in combination with an ssn mutation gave strikingly higher invertase activity than migl strains (4-11-fold higher) or
thecorresponding
ssn strain (10-100-fold higher).
Thus, m i g l actssynergisticallywith ssn2,ssn3, ssn4,
ssn5, ssn7 and ssn8 to relieve glucose repression of
invertase.
migl and other ssn mutations act synergistically in
suppression of snfl: To determine whether rnigl and
the other ssn mutations also act synergistically as suppressors of snf I , we selected for loss of the pCE9 plasmid
from the above strains using 5-FOA. Under derepressing
conditions, allof thesnfl
migl ssn triple mutants
showed much greater invertase activity than any of the
snfl ssn doublemutants andand at least fourfold
greater activity than snfl rnigl strains (Table 4). Under
glucose repressing conditions, constitutive invertase activity in the triple mutants ranged from 4 to 16-fold
greater than that of snfl migl strains. Thus, migl and
each of the ssn mutations function synergistically insuppression of the snfl invertase defect.
Interaction of the migl and ssnb mutations: If MIGl
contributes to repression of SUC2 by tetheringthe
SSN6-TUP1 complex, then mutations in rnigl and ssnb
should not show synergyfor relief of repression. To examine the relationshipbetween ssn6 andmigl, we compared ssn6migl strains and thessnbparent. Thedouble
M. Carlson
TABLE 4
Invertase activity of snfl migl ssn triple mutants
Invertase activityb
s n f l ssn
genotype"
snfl
snfl
snfl
snfl
snfl
snfl
snfl
snfl
SSN
ssn2-4
ssn3-l
ssn4-1
ssn5-4
ssn6A9
ssn7-I
ssn8-1
MIGI
Repressed
<1
<1
2
<1*
<1
366
1
2
Derepressed
58
<1
1
2
517 3 d
286 1
1000
1
3
miglA2::LEUZ'
Repressed
6
50
69
96
37
439
67
22
Derepressed
300
332
1110
275
25 1
snfl allele is either snfl-28 or snflA3.
Values are the average of 2-1 1 assays of one to three strains with
standard errors usually less than 20%. Cultures were grown in rich
medium (YEP).
'The snfl ssn miglA strains are the same strains used in Figure 2,
except they lack pCE9. Cultures used for assays were tested on SD-Ura
to confirm absence of the plasmid.
Data taken from WON
et al. (1984).
''
mutants showed constitutive invertase expression comparable to, but not greater than, ssnb single mutants
(Table 2). Comparison of snfl ssn6 andsnfl ssn6 migl
strains showed no substantial differences in invertase expression (Table 4), indicating that the combination of
migl and ssnb does not
result in a significantly enhanced suppression phenotype. In assays of snfl ssnb
and snfl ssn6 migl mutants carrying pCE9, the migl
strains showed somewhat greater invertase activity, but
the values were only 1.5-fold higher (Figure 2). These
results are consistent with the model that MIGl functions as a DNA-binding protein that associates SSN6TUPl with the SUC2 regulatory region.
DISCUSSION
Glucose repression of SUC2 transcription is dependent on the SSNG and TUPl proteins. KELEHER et al.
(1992) have proposed that the SSN6-TUP1 complex is
tethered to the SUC2 promoter by the DNA-binding
protein MIG1. Our data are consistent with this model
in two respects. First, both migl and ssn6 are suppressors of snfl, suggesting that bothMIGl and SSNG could
function together downstream of the SNFl kinase. Second, migl and ssn6 mutations do not show synergistic
loss of repression. Additional support for this model
comes from evidence that a LexA-MIG1 fusion protein
represses target gene transcription in an SSN6- and
TUPl-dependent manner (M. TREITEL
and M. CARLSON,
unpublished results).
However, our findings do not support the idea that
the major repressive effects of SSN6-TUP1
on SUC2 are
achieved via the tethering function
of MIG1.Disruption
of M I G l causes onlypartial release from glucose repres
sion while mutations in SSN6 or TUPl cause dramatic
loss of repression. The simple interpretation of these
Synergy of migl and ssn Mutations
c
53
et al. 1984). We showed that MIGl is identical to SSNI.
MZGl is also the same as CA T4 (H.-J. SCH~~I.I.EK,
personal
Single mutant
communication), which was identified by suppression
Double mutant
of s n f l / c a t l and snf4/cat3 mutations (SCH~U.I.ER and
ENTI
AN 1991) .
We then present evidence that other SSN genes also
contribute to repression of SUC2. As is the case for migl,
the mutations ssn2 through ssn5, ssn7 and ssn8 cause
at best a slight release of repression for SUC2 in wild type
and only weak suppression of the snfl defect in derepression. However, migl acts synergistically witheach of
these ssn mutations to relieve glucose repression. This
synergy also extends to the suppression of the snfl defect in SUC2 derepression.
Several models could explain the synergistic interactions between migl and the ssn mutations. First, some
of the SSN proteins may contribute to the bindingof the
SSN6-TUP1 complex to the SUC2 locus. One or more
SSN proteins may function redundantly with MIGl, or
alternatively, another SSN may assist MIGl in tethering
the SSN6-TUP1 complex to the SUC2 promoter. In ei1
<
j
b
,<l
ther case, the SSN6-TUP1 complex would no longer be
0
1
+ - + - + - + - + - + - + - 4"
MIGI:
efficiently secured to the promoter region when both
SSN:
+ + sm2 sm3 sm4 sm5 ssn6 ssn7 ssn8
M I G l and SSNgenesare mutant,
resulting in synergistic
GMOTYPE
loss of repression. It is also possiblethat otherSSN proFIGURE
P.-migl and ssn mutations act synergisticallyto reteins recognize sites at theSUC2 locus distinct from the
lieve glucose repression of invertase. Bars represent
the averMIGl
sites and mediate repression, which may or may
grown under repressing conage invertase activity for cultures
not
depend
on SSNGTUPl. If repression of SUC2 inditions in SC-Ura. All strains assayed carry the snfl mutant
volves a transcriptional cascade, perhaps similar to that
allele at the chromosomal
locus and the wild-type ShFl allele
on pCE9. Relevant genotypes
are indicated.+, wild-type allele;
of the GAL regulatory pathway, SSN proteins could me-, mutant allele. Blackbars represent single mutants and
diate repression of regulatory genes within the cascade.
shaded bars represent double mutants. Standard errors were
These models are not mutually exclusive, and different
less than 20%.
SSN proteins may contribute to repression by different
mechanisms.
results is that the repression of SUC2 by SSN6-TUP1 is
We have demonstrated a role for theSSN proteins in
not mediated solely by MIGl.Amore
complicated
glucose
repression. Evidence suggeststhat like SSNG and
model, that loss of MIGl abolishes activation as well as
TUPl,
the
other proteins in the group SSN2 through
repression, is suggested by evidence that Led-MIGI
SSN8
may
also
participate in global repression mechafunctions as an activator in the absence of SSNG (M.
nisms.
These
ssn
mutations confer aflocculent phenoTREITEL
and M. CARLSON, unpublished results);however,
type,
and
several
cause slow growth and/or partial dethe high level SUC2 expression observed in a migl ssn6
fects
in
an
a-factor
halo assay (data not shown). Further
double mutant argues against this model.
insight
into
the
functions
of these SSN proteins will reWe favor the interpretation that other proteins bequire
molecular
genetic
characterization.
sides MIGl contributeto repression by SSN6-TUPl.
Since the isolation of ssnl through ssn8, other similar
One possibility isthat otherproteins assist MIGl in bindmutations
have been described. For example, urrl afing the SSN6-TUP1 repressor complex to the SUC2
fects
glucose
repression, confers clumpiness, and sup
regulatory region. Another possibility is
that SSNG exerts
presses
s
n
f
l
,
suggesting
possible allelism with
an ssn muits repressive action on SUC2 via a regulatory cascade
tation
(FLICK
andJoHNsToN
1990).
The
skol
disruption
involving other DNA-binding proteins, as well as MIG1.
affects glucose repression of SUC2 and interacts with
For example, glucose repression of GAL genes involves
m i g l , but does not cause the clumpiness characteristic
a transcriptional cascade, although in this case MIGl
acts at multiple steps (FLICK
andJoHNsToN 1990; GRlGGs
of the ssn mutants (NEHLINet al. 1992). The tsJ3/sin4
and othertsf mutations affect repression and cause flocand JOHNSTON 1991; NEHLINet al. 1991; FLICKand
culence (JIANC and STILLMAN
1992; CHENet al. 1998a,b),
JOHNSTON 1992).
but differ from the ssn alleles in causing temperatureBecause migl was found to suppress s n f l , we tested
sensitive lethality.
for allelism to other mutations in thegroup ssnl
An unexpected finding in this study was that SUC2
through ssn8, identified as suppressors of snfl (CARLSON
a
, i,;
L. G. Vallier and M. Carlson
54
expression is significantly regulated in snfl A m i g l A and
snf4A miglAmutants in response to glucose availability
( 1 4 and 37-fold, respectively). Thus, signals regarding
glucose availability can be transmitted to theSUCZ promoter in the absence of the SNFl protein kinase. In
addition, these results suggest that some of the SSN proteins could function in a novel signalling pathway. It will
be important to identify components of this pathway in
future studies.
We thank MARKJOHNSTON and HANS RONNEfor strains and plasmids.
We thank PACO ESTRUCH
and JANET SCHULTZ for constructing strains.
This workwas supported by National Institutes of Health grant
GM34095 to M.C.
LITERATURE CITED
CUUSON,M.,B.C. OSMOND,
L. NEICEBORN
and D. BOTSTEIN,
1984 A
suppressor of snfl mutations causes constitutive high-level invertase synthesis in yeast. Genetics 107: 19-32.
CELENZA,
J. L., and M. CARLSON,1986 A yeast gene that is essential for
release from glucose repression encodes a protein kinase. Science
233:1175-1180.
CELENZA,
J. L., and M. CARLSON, 1989 Mutational analysisof the
Saccharomyces cerevisiae SNFl protein kinase and evidence for
functional interaction with the SNF4 protein. Mol. Cell. Biol. 9:
5034-5044.
J. L., F. J. ENGand M. CARUON,
1989 Molecular analysis of
CELENZA,
the SNF4 gene of Saccharomyces cerevisiae: evidence for physical
association of the SNF4 protein with the SNFl protein kinase.
Mol. Cell. Biol 9: 5045-5054.
CHEN,
S., R.J. WEST, JR.,
J. MA, S.JOHNSON,
H. CANSet al., 1993a TSFl
to TSFG: required for silencing the Saccharomyces cerevisiae GAL
genes, are global regulatory genes. Genetics 134: 701-716.
H. CANS,B. KRUCER et al.,
CHEN,S., R.W.WEST, JR., S. L. JOHNSON,
1993b TSF3, a global regulatory protein that silences transcription of yeastGAL genes, also mediates repression by a2 repressor
and is identical to SIN4. Mol. Cell. Biol. 13: 831-840.
FLICK,
J. S., and M.JOHNSTON, 1990 Two systems of glucose repression
of the G A L l promoter in Saccharomycescerevisiae. Mol. Cell.
Biol. 10: 4757-4769.
FLICK,
J. S., and M. JOHNSTON,
1992 Analysis of URS,-Mediated Glucose Repression of the G A L l promoter of Saccharomyces cerevisiae. Genetics 130: 295-304.
GRIGGS,
D. W., and M. JOHNSTON,
1991 Regulated expression of the
GAL4 activator gene in yeast provides a sensitive genetic switch
for glucose repression. Proc. Natl. Acad. Sci.USA 88: 8597-8601.
Y. W., and D. STILLMAN,
1992 Involvement of the SIN4 global
.JIANG,
transcriptional regulator in the chromatin structure of Saccha-
romyces cereuisiae. Mol. Cell. Biol. 12: 4503-4514.
JOHNSTON,
M., and M. CARLSON, 1992 Regulation of carbon and phos
phate utilization, pp. 193-281 in The Molecular and Cellular Biology of the Yeast Saccharomyces: Gene Expression. Cold Spring
Harbor Laboratory, Cold Spring Harbor, N.Y.
KELEHER, C. A.,M. J. REDD, J. SCHULTZ,
M. CARLSON and A. D. JOHNSON,
1992 Ssn6Tupl is a general repressor of transcription in yeast.
Cell 68: 709-719.
NEHLIN,
J. O., and H. RONNE,
1990 Yeast MIGl repressor is related to
the mammalian early growthresponse and Wilms’ tumour finger
proteins. EMBO J. 9: 2891-2898.
NEHLIN,
J. O., M. CARLBERG and H. RONNE,
1991 Control ofyeast GAL
genes by MIGl repressor: A transcriptional cascade inthe glucose
response. EMBO J. 10: 3373-3377.
NEHLIN, O.,
J. K. CARLBERC
and H. RONNE,
1992 Yeast SKOl gene
encodes a bZIP protein that binds to the CRE motif and acts
as a repressor of transcription. Nucleic Acids Res. 20: 52715278.
ROSE,M. D., F. WINSTONand P. HIETER,
1990 MethodsinYeast
Genetics:ALaboratoryCourseManual.
Cold Spring Harbor
Laboratory, Cold Spring Harbor, N.Y.
%HULLER,
H.-J., and IC-D. ENTIAN,1987 Isolation and expression
analysis oftwo yeast regulatory genes involved inthe derepression
of glucose-repressible enzymes. Mol. Gen. Genet. 209: 366-373.
SCHULLER,
H.-J., and R-D. ENTIAN, 1988 Molecular
characterization of
yeast regulatory gene CAT3 necessary for glucose derepression
and nuclear localization of its product. Gene 67: 247-257.
%HULLER,H.-J.,and IC-D. ENTIAN,
1991 Extragenic suppressors of
yeast glucose derepression mutants leading to constitutive synthesis ofseveral glucose-repressible enzymes. J. Bacteriol. 173:
2045-2052.
SCHULTZ,
J., and M. CARLSON,1987 Molecular analysis of SSN6, a gene
functionally related to the SNFl protein kinase of Saccharomyces
cerevisiae. Mol. Cell. Biol 7: 3637-3645.
SCHULTZ, J., L. MARSHALL-CARLSON
and M. CARISON,
1990 The
N-terminal TPR region is the functional domain ofSSN6, a
nuclear phosphoprotein of Saccharomyces cereuisiae. Mol. Cell.
Biol. 10: 4744-4756.
TRUMBLY,
R. J., 1992 Glucose repression in the yeast Saccharomyces
cerevisiae. Mol. Microbiol. 6: 15-21.
L. G., and M.CARLSON, 1991 New SNF genes, GAL11 and
VALLIER,
G R R l affect SUCP expression in Saccharomyces
cerevisiae.
Genetics 129: 675-684.
1990 Characterization of T U P l :
WILLIAMS,
F. E., and R. J. TRUMBLY,
a mediator ofglucose repression in Saccharomyces cerevisiae.Mol.
Cell. Biol. 10: 6500-6511.
WILLIAMS,
F. E., U. VARANASI
and R. J. TRUMBLY,
1991 The CYC8 and
TUPl proteins involved in glucose repression in Saccharomyces
cerevisiae are associated in a protein complex. Mol. Cell. Biol. 11:
3307-3316.
Communicating editor: M. JOHNSTON