ABSTRACT
Sin4p is a component of a mediator complex associated with the C-terminal domain of RNA polymerase II and
SIN4
is required for proper regulation of several genes in yeast, including the
HO
endonuclease gene, glucose repressible genes and MATa cell-specific genes. Previous studies indicated that
SIN4
may influence transcription through changes in the organization of chromatin.
We have examined a specific chromatin structure associated with MATa cell-specific repression in
sin4
MAT
[alpha]
cells to determine if
SIN4
is required for nucleosome positioning. Although the loss of
SIN4
has no effect on nucleosome location, we find that the sensitivity of bulk
chromatin from
sin4
cells to micrococcal nuclease digestion is strikingly increased relative to chromatin from isogenic wild-type cells. The nuclease hypersensitivity of chromatin from
sin4
cells is not related to gross alterations in histone gene expression or to bulk increases in histone modification. Our experiments suggest that
SIN4
directly or indirectly regulates a global aspect of chromatin accessibility, providing a molecular
basis for phenotypic similarities between
sin4
mutations and mutations in histones.
Genetic analysis in yeast has revealed two opposing sets of global
transcriptional regulators. The
SWI
/
SNF
genes encode positive regulators of diverse processes, including expression of
the HO endonuclease, suppression of Ty or solo [delta] insertion mutations and activation of glucose repressible genes (
1
). Conversely, the
SPT
/
SIN
genes generally encode negative regulators (
2
). Mutations in the
SPT
/
SIN
genes often suppress mutations in the positive
SWI
/
SNF
activators. Members of both families of regulators are thought to influence transcription through modulation of chromatin structure (
2
).
SIN4
is required for repression of the
HO
(
3
),
IME1
(
4
) and MAT
a
cell-specific genes (
5
). It is also required for full activation of genes such as
CTS1
and
HIS4
(
3
,
6
). Sin4p associates with Rgr1p
in vivo
(
6
,
7
) and mutations in either
SIN4
or
RGR1
exhibit similar phenotypes (
4
,
6
,
8
). The discovery that Sin4p and Rgr1p, together with Gal11p and p50, are
associated with polymerase II mediator complexes (
7
) indicates that these proteins directly affect the function of basal
transcription proteins
in vivo
. However, mutations in
SIN4
or
RGR1
exhibit pleiotropic phenotypes similar to mutation or depletion of histones,
including activation of UASless promoters, suggesting a role (direct or
indirect) for these gene products in the organization of chromatin (
3
,
9
). Mutations in
SIN4
also partially suppress defects in transcriptional activation of
ho
-
lacZ
caused by mutations in the
SWI2
gene, encoding an important subunit of the Swi/Snf chromatin remodeling complex, as well as other
SWI
genes (
8
,
10
,
11
). In addition, the superhelical density of plasmids is altered in
sin4
(
3
) and
rgr1
(
9
) cells and unique chromatin structures are associated with repression of several genes regulated by
SIN4
and
RGR1
, including the
GAL1-10
genes (
12
),
PHO5
(
13
) and the MAT
a
cell-specific genes (
14
,
15
).
To directly investigate whether mutations in
SIN4
alter chromatin structure, we have compared nucleosome organization and stability in isogenic wild-type and
sin4
cells. Our data indicate that
SIN4
modulates a global aspect of chromatin accessibility.
Saccharomyces cerevisiae
strains used in this study are listed in Table
1
. DY150, DY151, DY1704 and DY2693 are isogenic strains in the W303 background (
16
). DY882 and DY1720 are isogenic strains in the YPH499/YPH500 (S288C) background
(
17
). DY131 and DY1675 are isogenic strains in the K1107 background (
18
). Rich medium (YPD) and minimal medium (SD) supplemented with amino acids,
uracil or adenine (as appropriate) were prepared as described (
19
).
The methods for RNA blot hybridization and the
ACT1
probe have been described previously (
20
). Hybridization probes were prepared from plasmids generously provided by Mary
Ann Osley: the histone H2A probe was a 2.3 kb
Sac
I fragment from plasmid YCp50-TRT1 (
21
) containing the
HTA1
(and
PRT1
) gene. The histone H2B probe was a 1.6 kb
Bam
HI-
Sac
I fragment from plasmid YCp50-TRT1 containing the
HTB1
gene. The histone H3 probe was a 2.1 kb
Hin
dIII-
Bam
HI fragment from plasmid YCp50-H3-H4-copy2 containing the
HHT2
gene. The histone H4 probe was a 2.3 kb
Sma
I-
Bam
HI fragment from plasmid YCp50-H3-H4-copy1 containing the
HHF1
gene.
Table 1
Extracts were prepared from
sin4
or wild-type cells bearing the indicated reporter plasmids (
22
) and [beta]-galactosidase assays were performed as described (
19
).
Three different methods of chromatin preparation were used in these experiments
and in all cases chromatin from
sin4
cells exhibited increased nuclease sensitivity relative to wild-type cells. Chromatin for analysis of the histone genes was prepared after
micrococcal nuclease (MNase) digestion of NP-40 permeabilized spheroplasts basically as described by Kent
et al
. (
23
). To compare rates of nuclease digestion, wild-type or
sin4
cells were grown to equal densities (4 * 10
7
cells/ml) and then nuclei were prepared from each as previously described (
15
). In the `mixing' experiment shown in Figure
3
, nuclei were prepared from wild-type cells,
sin4
cells or an equal mixture of these cells. MNase digestion and purification of
the DNA were as previously described (
15
). Indirect end-labeling analysis and isolation of TALS chromatin was performed as
described (
14
). After MNase digestion, DNA was purified and digested with
Eco
RV and subjected to Southern blot analysis.
DNA samples were resolved by electrophoresis in 1.2% agarose gels and were
prepared for transfer as described (
24
). DNA was transferred by posiblot (Stratagene) onto GeneScreen+ (NEN) and then
UV crosslinked to the membrane. Prehybridization, hybridization and washes were
performed as described (
25
). Probes included an
Eco
RV-
Hin
dIII fragment of the TALS plasmid for the indirect end-labeling experiment in Figure
4
; a
Hin
dIII fragment (for H4) and a
Bam
HI-
Hin
dIII (for H3) fragment from plasmid YCp50-H3-H4-copy2 and a PCR-generated fragment containing sequences between 1535 and 1800 bp
of HML[alpha] for the blots shown in Figure
6
; an
Eco
RI-
Hpa
I fragment of the rRNA repeat for the blot shown in Figure
7
B.
The
SUP4-o
tRNA plasmid M3264 consists of a 0.6 kb
Pst
I-
Eco
RI fragment with ARS1 and a 0.4 kb
Xba
I-
Pst
I fragment with the
Sup4-o
tRNA gene cloned into the Bluescript KS+ vector. Two-dimensional electrophoresis of DNA in the presence of chloroquine was
performed as described previously (
3
). Topoisomers of M3264 were detected in the Southern blot using a
32
P-labeled ARS1 fragment.
Yeast histones were isolated as described previously (
26
-
28
). Histones were resolved by SDS-PAGE or acid urea-PAGE as described (
29
,
30
).
MNase is commonly used to probe chromatin structure because this enzyme
preferentially cleaves linker DNA located between nucleosomes. Partial
digestion of chromatin with this nuclease yields a `ladder' of oligonucleosome sized fragments. Examination of bulk chromatin isolated from isogenic wild-type or
sin4
cells after digestion with MNase reveals a striking and global increase in
nuclease sensitivity in the chromatin isolated from
sin4
cells (Fig.
1
A) relative to wild-type cells. For example, MNase `ladders' are evident in chromatin from
sin4
cells upon digestion with 12.5-25 U/ml of the enzyme, but ladders are not clearly evident in the
chromatin from wild-type cells until digestion with 100-200 U/ml MNase. Indeed, at every level of enzyme used, the
chromatin from
sin4
cells exhibits greater digestion than the wild-type chromatin. Importantly, the increased digestion observed in the absence of
Sin4
is not due to decreased recovery of chromatin from these cells, since similar
levels of chromatin were recovered from both types of cells.
SIN4
is required for repression of MAT
a
cell-specific genes in MAT[alpha] cells (
5
), which is mediated by [alpha]2/Mcm1p, in conjunction with Ssn6p and Tup1p (for a review see
30
). Binding of [alpha]2/Mcm1p to its operator recruits the Ssn6p-Tup1p complex (
31
,
32
), which then initiates a downstream organization of chromatin (
33
). Nucleosomes are precisely positioned adjacent to the [alpha]2 operator, probably through direct interactions between Tup1p and
histones H3 and H4 (
34
). To determine if
SIN4
also participates in organization of this chromatin, we examined nucleosome
positioning in wild-type and
sin4
MAT[alpha] cells using the plasmid TALS, which contains the [alpha]2/Mcm1 operator.
Indirect end-labeling experiments (Fig.
4
) indicate that the structure of TALS in the two strains is identical when
assayed with low levels of MNase. Nuclease hypersensitive sites flanking
nucleosome sized protected regions (relative to naked DNA digestion patterns)
are observed in both wild-type cells (+ lanes, Fig.
4
) and cells lacking
SIN4
(- lanes, Fig.
4
), matching the precise organization of nucleosomes evident in other MAT[alpha] cells (
14
,
35
).
SIN4
, then, is not required for localization of these nucleosomes.
The pleiotropic effects of
sin4
mutations on transcription and chromatin structure might be explained by
alterations in histone gene expression. For example,
sin4
mutations suppress loss of
HIS4
expression caused by insertion of a [delta] element of the Ty1 transposon into the
HIS4
promoter. Mutations in other genes which cause a Spt
-
phenotype, such as
SPT10
and
SPT21
, reduce expression of certain histone genes (
36
). Interestingly, deletion of one copy of the genes encoding histone H2A and H2B
([Delta]
hta1
-
htb1
) has previously been reported to cause local alterations in MNase digestion
patterns of specific genes (
37
), in contrast to the global affects on digestion we observe in the absence of
SIN4
.
The hypersensitivity of chromatin from
sin4
cells could result from a global activation of transcription. Conversely, this hypersensitivity could reflect an altered chromatin structure that plays a causal
role in gene activation. To distinguish these possibilities, we examined the
nuclease sensitivity of chromatin from loci which did not exhibit a change in
transcription in
sin4
cells relative to wild-type cells.
As shown above, histone gene expression is unaffected by
SIN4
loss (Fig.
5
). We analyzed MNase digestion patterns of chromatin associated with individual
histone genes by Southern blot to determine whether
SIN4
affects the structure of these loci. As for TALS and bulk chromatin, we
observed increased digestion of histone gene chromatin in the absence of
SIN4
. No digestion of chromosomal H4 (
HHF2
) DNA was observed in the absence of added nuclease (Fig.
6
, 0) and addition of MNase generated regularly spaced, nucleosome `ladders' in
both the wild-type and
sin4
chromatin. However, these ladders were evident at much lower concentrations of
nuclease in the chromatin from
sin4
cells. A similar increased sensitivity to nuclease was observed for chromatin
associated with the
HHT
2 gene, which encodes histone H3 (Fig.
6
B). Again, smaller digestion products were evident in the
sin4
chromatin at lower concentrations of nuclease than in the wild-type chromatin and larger molecular weight fragments were depleted more
readily. Chromatin associated with independent H2A and H2B loci also exhibited
hypersensitivity to MNase digestion in the
sin4
cells (data not shown).
We also examined the nuclease sensitivity of a gene which is not transcribed in
wild-type or
sin4
cells. Silencing of the
HML
mating locus is unaffected in
sin4
MAT
a
cells (data not shown), but loss of
SIN4
results in hypersensitivity of
HML
chromatin to nuclease digestion (Fig.
6
C), as for all other chromatin examined in these cells. Together with the above
results, our data indicate that transcription is not required for increased
nuclease sensitivity of
sin4
chromatin.
Since Sin4p associates with the RNA polymerase II holoenzyme complex (
7
), we wondered whether the effects of
sin4
mutations on chromatin are restricted to polymerase II transcription units. To
address this question, we examined the properties of chromatin associated with
the
SUP4-o
tRNA gene, which is transcribed by RNA polymerase III. This gene was inserted
into the plasmid Bluescript KS+, which lacks any known polymerase II
transcription units, together with an ARS (autonomously replicating sequence)
element. The
SUP4-o
tRNA suppresses ochre mutations such as that found in the
ade2-1
allele and thus yeast strains containing this plasmid can be selected for
adenine prototrophy. The linking number of the
SUP4-o
tRNA plasmid was determined after transformation into wild-type and
sin4
cells by two-dimensional electrophoresis on agarose gels containing chloroquine. Each
nucleosome induces a single superhelical turn in closed circular DNA.
Determination of superhelical density gives a relative measure of nucleosome
content. The two-dimensional gel resolves topoisomers as an arc, with linking number
increasing in a clockwise direction. The average distribution of topoisomers in
the
sin4
mutant is clearly shifted clockwise relative to the distribution from wild-type cells (Fig.
7
A), consistent with a change in chromatin structure, as previously reported for plasmids containing polymerase II transcription units.
The 35S precursor of rRNA is transcribed by RNA polymerase I from the cluster of tandemly repeated 9 kb rDNA genes on chromosome XII. To
determine whether a
sin4
mutation affected the chromatin structure of these RNA polymerase I
transcription units, we examined MNase sensitivity of the rDNA genes. Chromatin
isolated from
SIN4
and
sin4
strains was digested with MNase and a Southern blot was probed with rDNA-specific sequences. The rDNA chromatin from both
SIN4
and
sin4
cells exhibited a much less regular and more smeared pattern of digestion than
seen for the polymerase II genes examined above, perhaps reflecting the very
high transcriptional activity of these genes. However, like the polymerase II
genes above, the rDNA chromatin from
sin4
cells exhibited increased sensitivity to MNase digestion relative to the chromatin from
SIN4
cells (Fig.
7
B). We conclude that a
sin4
mutation has global effects on chromatin structure which are not limited to polymerase II transcribed genes.
Post-translational modification of histones, particularly acetylation of lysine residues, is often correlated with increased nuclease
sensitivity and transcriptional activation. To determine if the increased
nuclease sensitivity of
sin4
chromatin is associated with increased histone modification, histones were
isolated from wild-type and
sin4
cells. Differentially charged histone isoforms were first separated by acid
urea-PAGE and then subjected to SDS-PAGE. Histones were identified in the second dimension gels by mobility (Fig.
8
) and Western blot analysis (data not shown).
Previous studies indicated that changes in gene expression might be related to
changes in chromatin structure resulting from mutations in
SIN4
, as indicated by changes in linking number of nucleosomal plasmids (
3
,
6
,
9
). Our current studies have further defined these changes as a global increase
in nuclease sensitivity, reflecting an increased accessibility of chromatin in
sin4
cells.
Although we do not yet know the structural basis of this increased
accessibility, our findings help to explain other phenotypes reported for
sin4
cells. UASless promoters, for example, are activated in the absence of
SIN4
(
3
). These promoters are thought to be repressed by chromatin in wild-type cells, since mutation of genes encoding histones H2A, H2B and H3 or
depletion of histone H4 relieves this repression (
38
,
39
). The increased nuclease sensitivity of chromatin we observe in
sin4
cells may reflect a more open structure which mimics that resulting from
histone mutation, allowing activation of UASless promoters. Similarly,
mutations in histone genes (
2
) or
SIN4
(
3
,
9
) restore activity to promoters disrupted by Ty insertions. This Spt
-
phenotype is again consistent with a loosening of a repressive chromatin
structure (
2
). Finally,
sin4
mutations partially suppress transcriptional defects at the
ho
-
lacZ
and
SUC2
loci caused by mutations in the
SWI2
gene (
6
,
9
,
40
), which encodes a central component of the Swi/Snf chromatin remodeling machine
(
2
). The increased chromatin accessibility generated in the absence of
SIN4
may obviate the need for Swi/Snf function.
Several conditions could lead to the increased nuclease sensitivity we observe in the absence of
SIN4
. Obvious possibilities, including depletion of histones or bulk increases in histone acetylation,
however, were not observed. Our experiments, of course, do not rule out altered
usage of specific acetylation sites (or other modifications) in particular
histones in the absence of
SIN4
. Interestingly, we did not observe an increased suspectibility of
sin4
chromatin to digestion by DNase I, which might indicate that histone:DNA
contacts are not grossly altered in the absence of
SIN4
. A
sin4
mutation does cause changes in the supercoil density of plasmids (
3
,
9
; Fig.
7
), suggesting that the
sin4
mutation may change the number of nucleosomes loaded onto plasmid chromatin or
increase loss of nucleosomes from the plasmid.
SIN4
might also influence some aspect of higher order chromatin folding, either
through direct interactions with chromatin or through regulation of another
factor which affects chromatin accessibility (see below). Indeed, previous
observations indicate that the actions of
SIN4
in transcriptional repression are context dependent. For example,
sin4
mutations can suppress defects in the
swi5
activator to allow expression of
ho
-
lacZ
reporter genes but cannot suppress the
swi5
defect at the native
HO
locus (Jiang and Stillman, unpublished observations). Similarly, a
sin4
mutation causes derepression of a
PHO5
-
lacZ
reporter, but not of the native
PHO5
gene (Horz, personal communication). Moreover,
SIN4
negatively regulates
PHO5
transplaced into the
URA3
locus, but does not negatively regulate
PHO5
in its natural chromosomal location (
41
). These position effects could reflect a role for
SIN4
function in the organization of higher order structures.
Interestingly, the loss of repression of the MAT
a
cell-specific genes observed in
sin4
cells (
5
) is consistent with the increased chromatin accessibility we observe. Even
though nucleosomes are positioned properly in the absence of
SIN4
, alterations in nucleosome stability or in nucleosome-nucleosome contacts would provide a more open environment conducive to transcription.
Sin4p is associated with the RNA polymerase II holoenzyme as part of a mediator
complex (
7
). A putative subcomplex consisting of Sin4p, Rgr1p, Gal11p and an undefined 50
kDa peptide (p50) co-purifies with the mediator, raising the possibility that Sin4p plays a
direct role in transcriptional regulation. This idea is consistent with
previous findings that chimeric proteins consisting of Sin4p or Rgr1p fused to
the bacterial
lexA
DNA binding domain are able to activate
lexA
operator-containing reporters (
3
; unpublished observations). Additionally,
sin4
mutants are defective for the activation of specific genes, including MAT[alpha],
CTS1
and
HIS4
(
3
,
6
). Given the association of Sin4p and the mediator complex with polymerase II (
7
), it seems likely that
SIN4
participates in chromatin organization indirectly, through regulation of the
expression of a chromatin modifying activity or a non-histone structural component of chromatin. Identification of such a factor
or definition of direct interactions between Sin4p and chromatin may yield
further insights into the connections between chromatin structure and the regulation of gene expression.
We thank Robert T.Simpson, in whose laboratory the initial stages of this
project were done. We also thank members of the Simpson and Roth laboratories
for stimulating discussions and Diane Edmondson and Wolfram Horz for reading
the manuscript. We also thank Wolfram Horz for communication of results prior
to publication. We are grateful to Olga Yarygina for preparation of yeast
histones, to David Allis (University of Rochester) and Bryan Turner (University
of Birmingham) for H4-specific antibodies and to Mary Ann Osley (Sloan Kettering Cancer Center) for histone plasmids and [beta]-galactosidase reporter constructs. We thank Karen Hensley for help in preparing graphics. This
work was supported by grants from the NIH (GM51189) and the Texas Higher
Education Coordinating board (ARP#095) to S.Y.R. and NIH grant GM39067 to D.S.
*To whom correspondence should be addressed. Tel: +1 713 794 4908; Fax: +1 713
790 0329; Email: sroth@utmdacc.mda.uth.tmc.edu
Strain
Genotype
DY150
MAT
a
ade2 can1 his3 leu2 trp1 ura3
DY151
MAT[alpha]
ade2 can1 his3 leu2 trp1 ura3
DY1704
MAT[alpha]
sin4
::
URA3 ade2 can1 his3 leu2 trp1 ura3
DY1702
MAT
a
sin4
::
URA3 ade2 can1 his3 leu2 trp1 ura3
DY2693
MAT
a
rgr1
n
2
::
LEU2 ade2 can1 his3 leu2 trp1 ura3
DY882
MAT[alpha]
ade2 his3 leu2 lys2 trp1 ura3
DY1720
MAT[alpha]
sin4
::
URA3 ade2 his leu2 lys2 trp1 ura3
DY131
MAT
a
HO-lacZ ade2 ade6 can1 his3 leu2 met trp1 ura3
DY1675
MAT
a
HO-lacZ sin4
::
TRP1ade2 ade6 can1his3 leu2 met trp1 ura3
REFERENCES
Return