Content uploaded by Michel Werner
Author content
All content in this area was uploaded by Michel Werner
Content may be subject to copyright.
Tandem bromodomains in the chromatin
remodeler RSC recognize acetylated histone
H3 Lys14
Margaret Kasten
1
, Heather Szerlong
1
,
Hediye Erdjument-Bromage
2
,
Paul Tempst
2
, Michel Werner
3
and
Bradley R Cairns
1,
*
1
Howard Hughes Medical Institute and Department of Oncological
Sciences, Huntsman Cancer Institute, University of Utah School of
Medicine, Salt Lake City, UT, USA,
2
Molecular Biology Program,
Memorial Sloan-Kettering Cancer Center, New York, NY, USA and
3
Service de Biochimie et Ge
´
ne
´
tique Moleculaire, Ba
ˆ
timent 44, CEA/
Saclay, Gif-Sir-Yvette, France
The coordination of chromatin remodeling with chromatin
modification is a central topic in gene regulation. The yeast
chromatin remodeling complex RSC bears multiple bromo-
domains, motifs for acetyl-lysine and histone tail interac-
tion. Here, we identify and characterize Rsc4 and show
that it bears tandem essential bromodomains. Conditional
rsc4 bromodomain mutations were isolated, and were
lethal in combination with gcn5D, whereas combinations
with esa1 grew well. Replacements involving Lys14 of
histone H3 (the main target of Gcn5), but not other H3 or
H4 lysine residues, also conferred severe growth defects to
rsc4 mutant strains. Importantly, wild-type Rsc4 bound an
H3 tail peptide acetylated at Lys14, whereas a bromodo-
main mutant derivative did not. Loss of particular histone
deacetylases suppressed rsc4 bromodomain mutations,
suggesting that Rsc4 promotes gene activation. Further-
more, rsc4 mutants displayed defects in the activation of
genes involved in nicotinic acid biosynthesis, cell wall
integrity, and other pathways. Taken together, Rsc4 bears
essential tandem bromodomains that rely on H3 Lys14
acetylation to assist RSC complex for gene activation.
The EMBO Journal (2004) 23, 1348–1359. doi:10.1038/
sj.emboj.7600143; Published online 4 March 2004
Subject Categories: chromatin & transcription
Keywords: bromodomain; chromatin remodeling; histone
acetylation; RSC
Introduction
Chromatin structural changes play central roles in controlling
gene expression. Activation is often associated with the
mobilization and modification of nucleosomes, the basic
repeating unit of chromatin (Vignali et al, 2000).
Nucleosomes are covalently modified by histone acetyltrans-
ferases (HATs), deacetylases (HDACs), and methyltrans-
ferases (HMTs), which target specific lysine residues in the
histone subunits (Narlikar et al, 2002). Histone acetylation is
generally associated with transcription activation, while de-
acetylation generally correlates with repression (Grunstein,
1997). Histone modifications help recruit additional factors
including chromatin remodeling complexes, which utilize the
energy of ATP to mobilize nucleosomes and render chromatin
accessible to transcription factors (Havas et al, 2001). Two
such remodeling complexes in Saccharomyces cerevisiae are
the SWI/SNF complex and the highly related and essential
RSC (
Remodels the Structure of Chromatin) complex, both of
which have close counterparts in higher eucaryotes.
One conserved motif found in all SWI/SNF family of
remodelers is the bromodomain, a 110-amino-acid domain
also found in HATs, TFIID components, and other remodeler
complexes. In vitro, bromodomains bind to the amino-term-
inal tails of histones H3 and H4 (Ornaghi et al, 1999;
Ladurner et al, 2003; Matangkasombut and Buratowski,
2003), and acetylation of lysine residues on these tails
improves binding (Dhalluin et al, 1999; Hudson et al, 2000;
Jacobson et al, 2000; Owen et al, 2000; Ladurner et al, 2003;
Matangkasombut and Buratowski, 2003). These observations
suggest that acetylated lysines on histone tails provide a
platform for the recruitment of bromodomain-containing
transcriptional regulators (Jenuwein and Allis, 2001; Hassan
et al, 2002).
RSC subunits contain eight of the 15 bromodomains in
yeast, suggesting that the recognition of chromatin modifica-
tions is an important aspect of RSC function. Characterized
subunits include Sth1, the catalytic ATPase subunit of RSC
(Du et al, 1998), and the double bromodomain proteins Rsc1
and Rsc2 (Cairns et al, 1999). Rsc1 and Rsc2 are similar in
domain structure (multiple bromodomains, BAH domain,
HMG domain) to the polybromo/BAF180 subunit of human
hSWI/SNF-B (PBAF) complex, which functions as a cofactor
for ligand-activated transcription on a chromatin template
in vitro (Lemon et al, 2001). Taken together, bromodomain
proteins are emerging as important regulators of chromatin
remodeling and modifying complexes, but much remains to
be learned about their binding determinants and their utiliza-
tion in transcriptional regulation. Of particular interest is
whether the tandem arrangement of bromodomains in cer-
tain proteins might enable combinatorial recognition of his-
tone/factor modifications. Here, we identify and characterize
the tandem double bromodomain protein Rsc4, and develop
its connections to histone tail recognition, acetylation, and
gene expression.
Received: 28 August 2003; accepted: 3 February 2004; published
online: 4 March 2004
*Corresponding author. Howard Hughes Medical Institute and
Department of Oncological Sciences, Huntsman Cancer Institute,
University of Utah School of Medicine, Room 4362, 2000 Circle of Hope,
Salt Lake City, UT 84112, USA. Tel.: þ 1 801 585 1822;
Fax: þ 1 801 585 6410; E-mail: brad.cairns@hci.utah.edu
The EMBO Journal (2004) 23, 1348–1359
|
&
2004 European Molecular Biology Organization
|
All Rights Reserved 0261-4189/04
www.embojournal.org
The EMBO Journal VOL 23
|
NO 6
|
2004 & 2004 European Molecular Biology Organization
EMBO
THE
EMBO
JOURNAL
THE
EMBO
JOURNAL
1348
Results
Rsc4 identification and structure
RSC was purified to homogeneity from yeast cellular extracts
as described previously (Cairns et al, 1996; Figure 1A), and
peptides from Rsc4 were isolated and analyzed by MALDI-
TOF mass spectrometry. Mass fingerprinting from Rsc4 uni-
quely identified the open reading frame (ORF) YKR008W.
This result supports recent proteomic approaches for identi-
fying protein complexes, which found Ykr008w in associa-
tion with many other proteins, among them certain RSC
components (Gavin et al, 2002; Sanders et al, 2002).
Sequence comparisons using the CLUSTALW algorithm
(Higgins et al, 1996) reveal two bromodomains (BD1 and
BD2) separated by 36 amino acids (aa) (Figure 1B). This
proximity raises the possibility that the two bromodomains
pack against each other in a manner similar to the double
bromodomains of TAF1 (Jacobson et al, 2000). RSC4 is an
essential gene, as dissection of a sporulated RSC4/rsc4D
heterozygous diploid yielded only two viable spores, and as
a rsc4D strain was unable to lose a URA3-marked plasmid
bearing wild-type (WT) RSC4 on medium containing 5-FOA
(Figure 2B and data not shown), which prevents the growth
of URA3
þ
strains.
Immunoprecipitation of 6xMyc-Rsc4 efficiently co-precipi-
tates Sth1, the catalytic subunit of RSC, and this interaction is
stable at high stringency (Figure 2C, lane 2, and data not
shown), verifying the association of Rsc4 with RSC (and not
just fortuitous co-purification). To determine the oligomeric
state of Rsc4 in the RSC complex, we tested for co-immuno-
precipitation of WT Rsc4 with 6xMyc-Rsc4 utilizing a
polyclonal antibody that we raised against purified recom-
binant full-length Rsc4. This 6xMyc-Rsc4 derivative fully
complements rsc4D (Figure 2B and data not shown).
Immunoprecipitation of 6xMyc-Rsc4 with the anti-Myc anti-
body does not co-precipitate untagged Rsc4 (Figure 1C, lane
7) suggesting one copy of Rsc4 per RSC complex.
Both Rsc4 bromodomains are required for viability,
and the C-terminus is essential for assembly into RSC
For structure–function analysis, we prepared plasmids en-
coding Rsc4 derivatives lacking either BD1, BD2, both BD1
and BD2, or a small portion (23 aa) of the C-terminus, each
tagged with six copies of the myc epitope (Figure 2A).
Derivatives were tested for complementation of a rsc4D
mutation by assessing their ability to support growth follow-
ing the loss of a WT RSC4-URA3 plasmid on medium contain-
ing 5-FOA. Whereas full-length 6xMyc-Rsc4 complements
rsc4D, derivatives lacking BD1, BD2, or the C-terminus fail
to complement, demonstrating that these domains are essen-
tial for Rsc4 function (Figure 2B). Derivative assembly into
RSC complex was determined by co-precipitation of Sth1,
using strains bearing a WT untagged copy of RSC4 to support
viability. Whereas bromodomain deletion derivatives co-pre-
cipitated Sth1, the C-terminal truncation did not (Figure 2C).
Taken together, the Rsc4 C-terminus mediates assembly into
the RSC complex whereas the bromodomains perform an
alternative essential function.
Figure 1 Identification of Rsc4. (A) Purified RSC complex stained with Coomassie dye (from Cairns et al, 1996). (B) Domain structure of Rsc4
and alignment of Rsc4 bromodomains with the bromodomains of yeast Gcn5 and P/CAF. Regions of identity are highlighted in gray. Triangles
(.) mark residues mutated in rsc4-2. Rectangles (
) mark paired residues mutated by site directed mutagenesis. (C) Stoichiometry of Rsc4 in
RSC. Western analysis using anti-Rsc4 antiserum of whole-cell extracts (WCEs) and anti-Myc precipitations from three strains bearing different
tagged RSC4 alleles. Extracts were prepared from YBC627 transformed with either 6xMyc-Rsc4 (p603) (lanes 3, 4, 6, and 7) or the empty vector
(p415.MET25, p520) (lanes 2 and 5). Cells used in lanes 3 and 6 lost the untagged Rsc4 plasmid on SC þ 5-FOA.
Rsc4 recognizes acetylated H3 Lys14
M Kasten et al
& 2004 European Molecular Biology Organization The EMBO Journal VOL 23
|
NO 6
|
2004 1349
Site-directed mutagenesis of Rsc4 bromodomains
To better understand the role of the two bromodomains in
Rsc4 function, site-directed mutations (SDMs) were isolated
based on mutagenesis and structural studies of the Gcn5 and
P/CAF bromodomains (Dhalluin et al, 1999; Owen et al,
2000; Mujtaba et al, 2002). The bromodomain utilizes a
left-handed four-helix bundle (with helices termed Z, A, B,
and C) to form an architectural platform for two helix-
connecting loop regions (ZA and BC). These loops form a
significant portion of the binding pocket for acetyl-lysine
recognition, and also recognize residues flanking the acetyl-
lysine, which make important contributions to binding spe-
cificity. Residues in the ZA loop are predicted to make
essential acetyl-lysine contacts (Y364 of Gcn5 and Y760 of
P/CAF) (Dhalluin et al, 1999; Owen et al, 2000; Mujtaba et al,
2002). The BC loop residue N407 of Gcn5 (N134 in BD1 and
N268 in BD2 of Rsc4) coordinates a network of water-
mediated hydrogen bonds, or ‘water ring,’ with a carbonyl
group on the acetyl and is also critical for acetyl-lysine
binding (Owen et al, 2000). In addition, the conserved
adjacent aromatic residue (Y406 of Gcn5 and Y802 of P/
CAF) interacts directly with residues flanking the acetyl-
lysine, and is important for tail recognition (Dhalluin et al,
1999; Owen et al, 2000; Mujtaba et al, 2002).
Converting the conserved Y92 and Y93 residues (ZA loop)
to alanines in BD1 did not affect Rsc4 function, nor did
independently changing the analogous Y225 and Y226 resi-
dues in BD2 (Table I, SDM1 and SDM2; and see Figure 1B).
However, combining these mutations in BD1 and BD2 con-
ferred lethality (Table I, SDM3) without significantly affecting
protein stability or assembly into RSC (see Supplementary
Figure 1). Thus, the bromodomains appear partially redun-
dant; a reduction in the function of one bromodomain can
make viability reliant on full function of the other tandem
bromodomain partner. Conversion of the conserved Y133 and
N134 residues to alanines in BD1 conferred slow growth at
381C (Table I, SDM4; and Figure 1B), whereas the analogous
F267 N268 alanine substitutions (in BD2) conferred no
phenotype (Table I, SDM5). Thus, the two bromodomains
do not strictly rely on identical residue positions for substrate
recognition. However, we again observe greater than additive
defects in combination (Table I, SDM6). Similar relationships
are observed with replacements at other positions (Table I,
SDM7-11). Taken together, our targeted amino-acid replace-
ments identified residues important for Rsc4 bromodomain
function and provide evidence for partial redundancy,
although the complete loss of either bromodomain confers
lethality.
Isolation and characterization of conditional mutations
in Rsc4 bromodomains
For functional and suppression analysis, we sought tempera-
ture-sensitive (Ts
) rsc4 bromodomain alleles (rsc4 BD
Ts
)
with ideal properties: alleles that display WT growth at 281C
and inviability at 381C, encode stable proteins that assemble
well into RSC at the nonpermissive temperature, and encode
single amino-acid substitutions in each of the bromodomains.
To acquire such alleles, we targeted random mutations to the
region of RSC4 encoding both bromodomains using a plasmid
shuffle strategy (see Materials and methods). The screen
yielded 10 Ts
mutants (Figure 3A).
Figure 2 Deletion analysis of Rsc4 domains. (A) Diagram of Rsc4
bromodomain and C-terminal deletion constructs. (B) Rsc4 function
requires both bromodomains and the C-terminus. Strains were
incubated at 281C for 2 days. Strains (YBC627) bearing
pRS316.RSC4 and a LEU2-marked rsc4 derivative or control
(pRS415MET25, 6xMyc.RSC4 (p603), 6xMyc.rsc4 BD1D (p804),
6xMyc.rsc4 BD2D (p1172), 6xMyc.rsc4 BD1-2D (p805), or
6xMyc.rsc4 C-termD (p809)) were grown on selective media with
or without 5-FOA to enforce the loss of pRS316.RSC4.(C) Rsc4
mutant assembly into RSC. Extracts were prepared from strains
used in (B) grown in the presence of pRS316.RSC4. Immune
complexes were formed with anti-Myc antibody bound to beads,
washed, eluted, immunoblotted, and probed with anti-Sth1 or anti-
Myc antiserum.
Rsc4 recognizes acetylated H3 Lys14
M Kasten et al
The EMBO Journal VOL 23
|
NO 6
|
2004 & 2004 European Molecular Biology Organization1350
The majority of the alleles obtained bore mutations in both
bromodomains, again supporting the notion of partial redun-
dancy, whereas the few alleles obtained bearing a mutation in
a single bromodomain typically resulted in lower protein
stability or assembly (see below and Figure 3B). However,
as alleles bearing multiple mutations were not separated,
their individual contributions have not been determined. We
proceeded with three alleles: rsc4-2, rsc4-7, and rsc4-11. rsc4-2
has two mutations, one in each bromodomain (L120P,
Y275H; Figure 1B), and satisfies all the criteria listed above.
Therefore, rsc4-2 represents the most informative and useful
allele identified in our random screen, and all further studies
focused on this allele; rsc4-7 and rsc4-11 were chosen for
comparative purposes as they are stronger Ts
alleles (see
Materials and methods for all alleles/replacements). Most
rsc4 BD
Ts
alleles encoded stable Rsc4 derivatives, although
their mobility varied due to deletion of one or more of the six
Myc epitopes by spurious homologous recombination
(Figure 3B and data not shown). The rsc4 BD
Ts
derivatives
varied in their expression levels and competency for assem-
bly, as determined by co-precipitation of Sth1 (Figure 3B). We
note that reducing the abundance of Rsc4 protein 10-fold
below WT levels (through repression using a MET25 promo-
ter) has no phenotypic consequence (data not shown).
Certain RSC mutants show a cell wall defect at elevated
temperatures that can be suppressed by osmotic stabilizers
(Angus-Hill et al, 2001). Similarly, we found that growth on
1 M sorbitol, 200 mM CaCl
2
, or 100 mM MgCl
2
partially
suppressed the temperature-sensitive phenotype of the rsc4
BD
Ts
mutants (Figure 3A and data not shown), supporting a
role for Rsc4 in assisting RSC in maintaining cell wall
integrity. Certain rsc mutants (such as rsc3 or sth1) accumu-
late in G2/M at the nonpermissive temperature (Tsuchiya
et al, 1998; Angus-Hill et al, 2001) and may be linked to
defects in centromeric chromatin (Hsu et al, 2003), whereas
other rsc mutants lack this property. We observed little or no
cell cycle bias with rsc4 mutants (data not shown), suggesting
that Rsc4 function is not linked to centromere function.
Genetic cooperativity between Rsc4 bromodomains
and the Gcn5 histone H3 acetyltransferase
As bromodomains recognize acetylated histone tails, we
tested for genetic interactions between rsc4 bromodomain
mutations and mutations in the HAT complexes SAGA/ADA
and NuA4. Gcn5 is the catalytic subunit of the SAGA and ADA
complexes and primarily acetylates histone H3 at lysine 14
(Grant et al, 1998). Esa1 is the catalytic subunit of the NuA4
complex and primarily acetylates histone H4 at lysines 5, 8,
12, and 16 (Allard et al, 1999). ESA1 is an essential gene, and
Table I Impact of site-directed mutations on RSC4 function
Phenotype
Mutant BD1 mutation BD2 mutation 301C351C381C
SDM1 Y92A, Y93A – +++ +++ +++
SDM2 – Y225A, Y226A +++ +++ +++
SDM3 Y92A, Y93A Y225A, Y226A Inviable
SDM4 Y133A, N134A – ++ ++ +/
SDM5 – F267A, N268A +++ +++ +++
SDM6 Y133A, N134A F267A, N268A + +
SDM7 P99A, M100H – +++ +++ ++
SDM8 – P232A, M233H +++ +++ +++
SDM9 – P212A, F213A +++ +++ +++
SDM10 P99A, M100H P232A, M233H +++ +++ +++
SDM11 P99A, M100H P212A, F213A Inviable
Growth was assessed from 10-fold dilutions spotted to plates after 3 days at the indicated temperature and compared to WT (defined as +++).
‘++’ indicates slightly reduced viability (fewer colonies) but near WT colony size. ‘+’ indicates moderately reduced viability and smaller
colony size. ‘+/’ indicates severely reduced viability and extremely small colony size. ‘’ indicates no growth.
Figure 3 Conditional mutations in RSC4 bromodomains. (A)
Growth of rsc4 BD
Ts
mutants. Strains are transformants of
YBC627 where pRS316.RSC4 is exchanged for rsc4 BD
Ts
plasmids:
pRS314.RSC4 (p1060), pRS314.rsc4-2 (p1083), pRS314.rsc4-5
(p1084), pRS314.rsc4-6 (p1085), pRS314.rsc4-7 (p1086),
pRS314.rsc4-9 (p1087), pRS314.rsc4-10 (p1088), pRS314.rsc4-11
(p1089), pRS314.rsc4-12 (p1090), pRS314.rsc4-13 (p1091),
pRS314.rsc4-14 (p1092). (B) Assembly of rsc4 BD
Ts
mutants into
RSC. Extracts were prepared from strains transformed with plas-
mids bearing rsc4 BD
Ts
alleles fused to Myc epitopes grown in the
presence of pRS316.RSC4 at 351C. Immune complexes were formed
with anti-Myc antibody bound to beads, washed, eluted, and
immunoblotted with anti-Sth1 or anti-Myc antiserum. The migra-
tion of the Myc-Rsc4 derivatives deviates from WT due to deletion
of Myc epitopes by homologous recombination.
Rsc4 recognizes acetylated H3 Lys14
M Kasten et al
& 2004 European Molecular Biology Organization The EMBO Journal VOL 23
|
NO 6
|
2004 1351
the esa1-L327S allele is Ts
and displays a modest reduction
in H4 acetylation levels genome wide (Clarke et al, 1999), but
shows significant reductions at certain genes (Suka et al,
2002). esa1-L254P is a stronger Ts
allele and displays a
greater reduction in H4 acetylation (Clarke et al, 1999).
Interestingly, rsc4 BD
Ts
mutations were lethal in combi-
nation with a gcn5 deletion at 281C (Figure 4A), suggesting
that Rsc4 function is highly sensitive to the acetylation status
of histone H3. In striking contrast, combinations involving
the esa1-L327S and esa1-L254P alleles showed no growth
defects at 281C (Figure 4B and data not shown). A slight
lowering (by 21C) of the nonpermissive temperature was
observed with rsc4-2 esa1-L327S combinations (Figure 4C);
however, this only occurs over a very narrow temperature
range (32–341C). Taken together, rsc4 BD
Ts
mutants are
entirely reliant on proper H3 tail acetylation, whereas they
display only partial and conditional reliance on proper H4
acetylation.
Genetic interactions between Rsc4 and histone
N-terminal tails
To test whether rsc4-2 gcn5D lethality was due to defects in
acetylation of particular residues of histone tails, the rsc4-2
allele was combined with specific histone H3 and H4 tail
mutations and tested for phenotypes (see Materials and
methods). Individual replacements involving Lys5, 8, 12, or
16 of the H4 tail had, at most, a slight impact on viability and
no impact on growth (Zhang et al, 1998), and combinations
with rsc4-2 conferred little or no additional growth defect
(Table II). Certain replacements involving two H4 residues
slightly reduced viability, but combinations with rsc4-2 typi-
cally had little additional impact. In contrast, strong genetic
interactions were observed between rsc4-2 and replacements
involving Lys14 of H3. Substitution of glutamine, arginine, or
glycine for lysine at position 14 in isolation conferred almost
no growth defect, whereas combinations with rsc4-2 con-
ferred extremely slow growth. We note that the acetyl-lysine
binding pocket (for bromodomains characterized structu-
rally) utilizes conserved hydrophobic residues (and a coordi-
nated water ring) that contact the acetyl group itself.
Although glutamine (like acetyl-lysine) is uncharged, several
lines of evidence suggest that it should not fully mimic acetyl-
lysine: the acetyl moiety is absent in glutamine, and gluta-
mine is significantly shorter and contains different terminal
functional groups. In accordance with this reasoning, the
K14Q substitution enhanced rsc4-2 phenotypes rather than
suppressing them. In addition, enhancement of rsc4-2 phe-
notypes was specific for combinations involving Lys14 of H3,
as replacements involving Lys9 had no effect (Table II). Since
Lys14 of histone H3 is the preferred site of acetylation by
Gcn5 (Kuo et al, 1996), these results support the specificity
observed in our combinations with HAT mutations and point
to H3 Lys14 as a critical residue for Rsc4 bromodomain
function.
Suppression of rsc4 bromodomain mutations by
deletion of HDACs and components of repression
complexes
As loss of histone H3 acetylation exacerbates rsc4 BD
Ts
mutations, we tested whether increased acetylation levels
might suppress rsc4 BD
Ts
mutations through combinations
involving HDAC mutations. Sir2 is the founding member of
the NAD-dependent family of HDACs, and it deacetylates
lysine residues on histone H3 and H4 (Imai et al, 2000). In
addition, there are five different NAD-independent HDACs in
yeast: Hda1, Hos1, Hos2 (part of the Set3C complex), Hos3,
and Rpd3 (part of the Sin3 complex). These HDACs vary
widely in their substrate specificity (histone tail and residue)
Figure 4 Genetic interactions between Rsc4 bromodomains and
HATs. (A) rsc4 BD
Ts
mutants are lethal in combination with
gcn5D. Key indicates genotype of strains. YBC1622 was transformed
with rsc4 BD
Ts
plasmids or controls: pRS314.RSC4 (p1060),
pRS314, pRS314.rsc4-2 (p1083), pRS314.rsc4-7 (p1086), or
pRS314.rsc4-11 (p1089). Strains were grown at 281C on selective
media with or without 5-FOA to enforce the loss of pRS316.RSC4.
(B) rsc4 BD
Ts
mutants are viable in combination with esa1-L327S
at 281C. YBC1796 was transformed with the rsc4 BD
Ts
plasmids
listed above and plated as above. (C) rsc4-2 is lethal in combination
with esa1-L327S only at elevated temperatures. YBC627 and
YBC1796 were transformed with pRS314.RSC4 or pRS314.rsc4-2,
and plated to 5-FOA to enforce loss of pRS316.RSC4. The 10-fold
dilutions were spotted to YPD and grown for 2–3 days at the
indicated temperatures.
Rsc4 recognizes acetylated H3 Lys14
M Kasten et al
The EMBO Journal VOL 23
|
NO 6
|
2004 & 2004 European Molecular Biology Organization1352
and also in their impact on both genome-wide and locus-
specific acetylation levels.
To test for suppression of rsc4 BD
Ts
alleles, we crossed a
rsc4D strain bearing WT RSC4 on a URA3-marked plasmid to
strains containing deletions of the above-mentioned HDACs,
transformed in plasmids containing the rsc4 BD
Ts
mutations,
and selected for loss of the WT RSC4 on 5-FOA media. Cells
were then spotted to YPD plates, incubated at 28, 33, 35, or
381C and compared for growth. Interestingly, deletion of
RPD3 did not suppress rsc4 BD
Ts
mutations, whereas dele-
tion of HDA1, HOS1, HOS2, HOS3, or SIR2 conferred partial
suppression (Figure 5A and B and data not shown). Deletion
of any of these latter five HDACs suppressed the rsc4-2 allele
partially at 351C and weakly at 381C. Partial suppression of
Table II Genetic interactions between rsc4-2 and histone N-terminal tail mutations
Growth on 5-FOA at 331C
Plasmids H3 H4 RSC4 rsc4-2
pWZ414-F13 WT WT ++++ ++++
pWZ414-F30 K9Q WT ++++ ++++
pWZ414-F53 K9R WT ++++ ++++
pWZ414-F31 K14Q WT ++++ /+
pWZ414-F36 K14G WT ++++ /+
pWZ414-F43 K14R WT ++++ /+
pWZ414-F23 WT K5Q ++++ ++++
pWZ414-F22 WT K5R ++++ +++
pWZ414-F25 WT K16Q ++++ ++++
pWZ414-F26 WT K16G ++++ ++++
pWZ414-F24 WT K16R +++ ++
pWZ414-F51 WT K5,12Q +++ +++
pWZ414-F52 WT K5,12R ++++ ++
pWZ414-F47 WT K8,16Q ++++ +++
pWZ414-F49 WT K8,16R +++ ++
pWZ414-F48 K14Q K8,16Q ++++ /+
pWZ414-F50 K14R K8,16R +++
Strains WZY42 (RSC4) and YBC1931 (rsc4-2) carrying Ycp50-HHT2-HHF2 were transformed with plasmids bearing the indicated mutations
(Zhang et al, 1998) and then tested on medium containing 5-FOA for the ability to grow at 331C. ‘+++’ and ‘++’ indicate reduced viability
(fewer colonies), but near WT colony size.‘/+’ indicates severely reduced viability and extremely small colony size.‘’ indicates no growth.
Growth was assessed after 3–4 days and compared to WT (defined as ++++).
Figure 5 Genetic interactions between Rsc4 bromodomains and HDACs. (A) rsc4 BD
Ts
mutations are suppressed by specific HDAC deletions.
rsc4D (YBC627), rpd3D rsc4D (YBC1698), and hos3D rsc4D (YBC1789) were transformed with plasmids bearing WT RSC4 or rsc4 BD
Ts
alleles
and grown on 5-FOA media to enforce loss of pRS316.RSC4. The 10-fold dilutions were spotted to YPD and grown for 2–3 days at the indicated
temperatures. (B) rsc4-2 is suppressed at 35 and 381C by deletion of several HDACs, but not by others. rsc4D (YBC627), rpd3D rsc4D (YBC1698),
hda1D rsc4D (YBC1821), hos1D rsc4D (YBC1702), hos2D rsc4D (YBC1727), hos3D rsc4D (YBC1789), set3D rsc4D (YBC1882), and sir2D rsc4D
(YBC1892) were transformed with pRS314.rsc4-2 (p1083), and grown on 5-FOA media to enforce loss of pRS316.RSC4. The 10-fold dilutions
were spotted to YPD and grown for 2–3 days at the indicated temperatures.
Rsc4 recognizes acetylated H3 Lys14
M Kasten et al
& 2004 European Molecular Biology Organization The EMBO Journal VOL 23
|
NO 6
|
2004 1353
the stronger rsc4-7 and rsc4-11 alleles was also detected. Hos2
is part of the Set3C complex, and set3 null also partially
suppressed the rsc4 BD
Ts
alleles (Figure 5B and data not
shown), suggesting that suppression is a property of the
Set3C complex.
To compare their relative strength as rsc4 BD
Ts
suppres-
sors, we tested the entire panel of suppressor candidates
discussed above with the rsc4-2 allele on a single plate at
three temperatures. We found that deletion of HDA1, HOS1,
HOS2, HOS3, SIR2,orSET3 suppressed the rsc4-2 allele to a
similar extent at 35 and 381C (Figure 5B). Since we observed
only partial suppression by these HDAC deletions, and since
HDAC deletions cannot suppress bromodomain deletion mu-
tations (data not shown), HDAC deletions do not simply
bypass Rsc4 function. Rather, the presence of the acetyl
moiety on the histone tail may be important for a partially
crippled bromodomain to interact, and the loss of this
modification may drop binding below a crucial threshold
(see Discussion).
Rsc4 bromodomains bind to histone H3 peptides
acetylated at Lys14 in vitro
To determine the histone tail specificity of Rsc4, we tested for
interaction with a panel of biotinylated histone tail peptides
that either bear or lack acetylation at specific lysine residues.
Two Rsc4 derivatives were produced by in vitro transcription/
translation, WT and SDM3. SDM3 bears replacements pre-
dicted to impair acetyl-lysine recognition (see Table I) but still
produces a stable protein that assembles into RSC in vivo (see
Supplementary Figure 1). Equimolar amounts of each bioti-
nylated peptide (in vast excess over Rsc4 protein) were
bound separately to streptavidin agarose beads. Peptide
binding to the beads was estimated as 95–100% efficient
(see Supplementary Figure 2). The peptide beads were then
incubated with equal levels of Rsc4 protein. The bound
material was washed extensively with buffers of moderate
stringency, eluted in sample buffer, and tested for retention
by immunoblot analysis (see Materials and methods). We
find that WT Rsc4 protein, but not the SDM3 derivative,
shows preferential binding to the histone H3 peptide acety-
lated at Lys14 compared to the unmodified peptide
(Figure 6A, compare lanes 3 and 4 with lanes 9 and 10).
Furthermore, WT Rsc4 protein shows little binding to the
unacetylated or acetylated H4 tail peptides tested (Figure 6A,
lanes 5 and 6).
To determine whether the binding specificity was due sole-
ly to the tandem bromodomain region of Rsc4, we expressed
a recombinant version of the Rsc4
tandem bromodomain
Figure 6 Binding of Rsc4 to histone tail peptides. (A) Binding of WTand Rsc4 SDM3 (each bearing 2xMyc epitopes) to biotinylated N-terminal
histone tail peptides on streptavidin beads, assessed by Western blot. Peptides are labeled based on histone, amino-acid sequence, and
modification state. In all, 50% of the eluates and 5% of the input were resolved by SDS–PAGE, blotted, and analyzed using anti-Myc antiserum.
(B) Diagram of recombinant Rsc4 TBD construct. (C) Coomassie-stained gel of purified full-length recombinant Rsc4 TBD. (D) Slot blot
analysis of purified Rsc4 TBD following size exclusion chromatography. Blot was probed with anti-Rsc4 antiserum. Fraction peaks for three
molecular weight standards are indicated. (E) Western blot analysis of purified Rsc4 TBD bound to histone tail peptides. The H2B 7Kpos Ac
peptide is acetylated at seven lysines (Lys 6, 7, 11, 16, 17, 21, and 22) and H2B 10Kpos Ac is acetylated at all 10 lysines (Lys 3, 6, 7, 11, 16, 17,
21, 22, 30, and 31). In all, 50% of the eluates and 2% of the input were resolved by SDS–PAGE, blotted, and analyzed using anti-Rsc4
antiserum.
Rsc4 recognizes acetylated H3 Lys14
M Kasten et al
The EMBO Journal VOL 23
|
NO 6
|
2004 & 2004 European Molecular Biology Organization1354
region (Rsc4 TBD, residues 46–321) that bore a Flag epitope
at the amino terminus and a 10X-histidine tag at the carboxyl
terminus (Figure 6B). Full-length Rsc4 TBD was purified to
homogeneity by tandem affinity chromatography using nickel
agarose beads followed by anti-Flag agarose beads
(Figure 6C). Size exclusion chromatography of the purified
protein revealed a monomeric derivative of approximately
33 kDa (Figure 6D). Rsc4 TBD was then tested for peptide
binding specificity in the format described above (with addi-
tional peptide substrates), and showed clear preference for
H3 acetylated at Lys14 (Figure 6E, lane 4). Rsc4 TBD bound
poorly to unmodified and acetylated H4 peptides (Figure 6E,
lanes 5 and 6) and at essentially background levels to
unmodified or acetylated H2A and H2B peptides (Figure 6E,
lanes 7–10). Taken together, these data demonstrate that the
tandem bromodomain region of Rsc4 preferentially recog-
nizes histone H3 acetylated at Lys14.
Effect of Rsc4 bromodomain mutations on transcription
We then assessed the impact of rsc4 bromodomain mutations
on gene expression by performing transcriptional profiling
on rsc4-2, rsc4-7, and rsc4-11 strains at their permissive
(301C) and nonpermissive (35 or 371C) temperatures.
Different alleles and temperatures were employed to help
attribute changes to rsc4 as opposed to temperature shift
alone. Only a few genes were affected in the rsc4 BD
Ts
mutants at the permissive temperature: 6–11 genes were
downregulated more than two-fold, and 10–15 genes were
upregulated more than two-fold (see Supplementary Tables
S4, S5, S8, S9, S12 and S13). At nonpermissive temperatures,
between 47 and 111 genes were downregulated more than
two-fold and 123–194 genes were upregulated more than
two-fold. The transcription profiles of the mutants grown at
nonpermissive temperatures were remarkably similar, with
rsc4-7 and rsc4-11 affecting nearly identical sets of genes
(compare Supplementary Tables S6 with S10 and S7 with
S11). Common classes of affected genes included the upre-
gulation of genes that regulate cell wall integrity and the
response to cell stress, as well as genes involved in iron
homeostasis (see Supplementary Table S1). Misregulation of
genes involved in cell wall integrity has been observed with
other rsc mutants (Angus-Hill et al, 2001). Interestingly,
transcript levels of several genes involved in nicotinic acid
biosynthesis and transport were decreased in all three rsc4
mutants (see Supplementary Table S1). Additionally, genes
involved in mRNA splicing and mating were also decreased
(see Supplementary Table S1). To ensure that our microarray
results reported true differences in mRNA abundance, we
performed S1 nuclease analysis on several transcripts and
observed similar changes (see Supplementary Figure 3 and
data not shown). Expression profiles of rsc4-2 mutants are
similar to those of rsc4-7 and rsc4-11 mutants, but also
include additional gene classes. Most striking is the large
number (39) of ribosomal proteins and histone gene loci
(known targets of RSC) that are downregulated in rsc4-2 at
371C (see Supplementary Table S2).
Recently, a connection between RSC and Pol III transcrip-
tion has been demonstrated, as genome-wide localization
reveals RSC at virtually all Pol III targets (Ng et al, 2002).
Furthermore, protein–protein interactions occur between the
Pol III machinery and Rsc4, but this interaction does not
involve the tandem bromodomain region of Rsc4 (M Werner,
in preparation). Indeed, rsc4 BD
Ts
mutations do not signifi-
cantly affect the transcript levels of the Pol III targets tested,
including SCR1 or the unspliced version of either RPR1 or the
tRNA tF(GAA)P2 (data not shown). Thus, Rsc4 appears to
bear three distinct regions: one for interaction with Pol III
components, one for assembly into RSC, and tandem bromo-
domains for histone tail interaction and regulation of Pol II
transcription.
Discussion
Bromodomains are found in chromatin regulatory complexes
and have been characterized as histone tail and acetyl-lysine
recognition motifs in vitro, but how they are utilized in vivo is
poorly understood. RSC is an essential and abundant chro-
matin remodeling complex bearing eight of the 15 bromodo-
mains present in S. cerevisiae, making RSC an attractive
complex to study bromodomain function. Here, we identify
Rsc4 as an integral and essential member of RSC and show
that its tandem bromodomains are essential for viability,
cooperate with H3 acetylation in vivo, are utilized to recog-
nize H3 Lys14 acetylation in vitro, and are important for the
expression of certain genes.
Rsc4 bears essential tandem bromodomains
RSC appears to have two types of double bromodomain
proteins: Rsc4 with two essential tandem bromodomains,
and the paralogs Rsc1/Rsc2, which have a nonessential N-
terminal bromodomain separated from a second essential
bromodomain by a large intervening region (Cairns et al,
1999). Thus, the tight bromodomain spacing in Rsc4 better
resembles the tandem bromodomain protein TAF1 (a mem-
ber of the human TFIID complex) (Jacobson et al, 2000) and
its yeast counterparts, Bdf1 and Bdf2, alternative members of
yeast TFIID. Certain results with Rsc4 are similar concep-
tually to those obtained with Bdf1 (Matangkasombut and
Buratowski, 2003); strong phenotypes were observed only
when the proposed acetyl-lysine recognition sites are im-
paired in both bromodomains. This relationship suggests a
similar function and partial redundancy between the two
tandem bromodomains in these proteins. However, as em-
phasized below, Rsc4 and Bdf1 show essentially opposite tail
specificity and genetic interactions.
rsc4 bromodomain mutants require H3 Lys14 and
proper H3 acetylation for viability
Our results show a strong link between Rsc4 bromodomains
and histone H3 acetylation. First, deletion of the H3 HAT
Gcn5 is lethal in combination with a rsc4 BD
Ts
mutation.
Also compelling is the strength and specificity of the genetic
interactions between H3 Lys14 and rsc4 BD
Ts
mutations. As
all three amino-acid replacements (glutamine, glycine, or
arginine) at position 14 conferred (with rsc4-2) the same
strong phenotype, a lysine appears to be required. Rsc4
function depended only weakly on histone H4 acetylation
as a rsc4 BD
Ts
mutation only lowers the nonpermissive
temperature for esa1-L327S by 21C, does not appreciably
affect the growth of the stronger esa1-L254P allele, and only
very modestly affects the growth of several histone H4 lysine
mutants.
Rsc4 recognizes acetylated H3 Lys14
M Kasten et al
& 2004 European Molecular Biology Organization The EMBO Journal VOL 23
|
NO 6
|
2004 1355
Suppression of rsc4 bromodomain mutations by HDAC
deletions
Here we provided the first evidence that an HDAC deletion
can partially suppress a bromodomain mutation. However,
the breadth of hdac suppression was surprising: only rpd3D
failed to show partial suppression. These results may reflect
HDAC tail/residue specificity. For example, although Rpd3
can deacetylate all core histones, it prefers H4 tails in vivo
(Rundlett et al, 1998; Suka et al, 2001). As Rsc4 function
appears largely independent of H4 acetylation status, the lack
of suppression by rpd3D would appear consistent with our
results. Although we cannot provide a definitive explanation
for the breadth of hdac suppression, it seems reasonable to
propose that rsc4 BD
Ts
mutations may cause defects in the
activation at many genes, which together confer temperature
sensitivity. These gene targets may be repressed by different
HDACs, and, therefore, removing any single HDAC may
enable partial expression of only a subset of Rsc4 targets,
and thus confer only partial suppression of rsc4 BD
Ts
temperature sensitivity.
Rsc4 binds histone H3 tails acetylated at lysine 14
Our biochemical experiments involving histone tail peptides
reveal an interaction between the Rsc4 tandem bromodo-
mains and the histone H3 tail acetylated at lysine 14. Binding
by Rsc4 to H3 Lys14Ac utilizes a pair of conserved tyrosine
residues in the ZA loop, residues that are utilized by other
bromodomains for interaction with the acetyl moiety on the
lysine (Ornaghi et al, 1999; Owen et al, 2000). However, our
data present an apparent paradox: Rsc4 bromodomains are
essential whereas the acetylation of H3 Lys14 is not essential,
as yeast strains bearing substitutions at H3 Lys14 are viable.
Analysis of other bromodomain structures shows that the
bromodomain (and the ZA loop in particular) also interacts
with residues flanking the modified lysine residue; thus both
the acetyl-lysine and flanking residues likely contribute to the
binding energy. In keeping with this idea, the Rsc4 TBD
derivative interacted weakly with the unmodified H3 tail.
One explanation for the paradox is that Rsc4 bromodomain
mutants have moderately reduced interactions with both the
acetyl-lysine and the flanking residues. These impaired bro-
modomains may now require acetylation at H3 Lys14 to bind
nucleosomes above a critical threshold required for remodel-
ing RSC targets. An alternative (but related) explanation is
that Rsc4 bromodomains bind two different determinants on
the nucleosome, only one of which is acetylated H3 Lys14. By
this model, interaction with H3 Lys14Ac and the second
determinant together provide sufficient binding energy for
interaction and target remodeling. Here again, the impaired
bromodomains may rely on the presence of the acetylation at
H3 Lys14 to avoid falling below a critical binding threshold.
Tandem bromodomain utilization and specialization
Interestingly, rsc4 mutations and bdf1 mutations show pre-
cisely opposite genetic interactions: bdf1 alleles display
strong genetic interactions with H4 lysine mutations, lethality
in combination with esa1-L327S (the identical allele used in
our studies), and little effect in combination with gcn5
(Matangkasombut and Buratowski, 2003). These results
strongly suggest that the tandem bromodomains of Bdf1 are
functionally distinct from those in Rsc4. Indeed, we find that
a Rsc4 chimera bearing a precise replacement of its double
bromodomain region with that of Bdf1 fails to complement
rsc4D, even though the chimeric protein is expressed and
assembles well into the RSC complex (data not shown).
Taken together, we suggest that yeast cells have developed
two types of tandem bromodomain proteins, each displaying
very different tail-modification specificity, in order to broaden
their repertoire of histone tail recognition.
One interesting proposal for tandem bromodomain
function is their use in cooperative recognition of two mod-
ifications. Indeed, the presence of multiple tandem bromo-
domains in proteins such as polybromo/BAF180 (and other
orthologs) presents the opportunity for the simultaneous
recognition of multiple modifications. In TAF1 (TAF
II
250),
the tandem bromodomains pack together and utilize their ZA
loops as part of the interface (Jacobson et al, 2000), raising
the possibility that the binding of a ligand at one bromodo-
main may influence the binding properties of the other
bromodomain. Studies on TAF1 clearly show preferred inter-
action with diacetylated H4 tails as compared to nonacety-
lated H4 tails; however, cooperativity per se was not tested
(Jacobson et al, 2000). Thus, whether Rsc4, TAF1, or other
tandem bromodomains display cooperative binding remains
an important unresolved question. For Rsc4, approaching this
issue first requires determining whether the bromodomains
in Rsc4 both recognize H3 Lys14Ac (and whether this in-
volves two H3 tails on the same nucleosome), or whether H3
Lys14Ac recognition involves only one of the two bromodo-
mains, with the other bromodomain recognizing a second
uncharacterized determinant.
Impact of Rsc4 bromodomain mutations on gene
expression
RSC functions in both activation and repression of target
genes (Moreira and Holmberg, 1999; Angus-Hill et al, 2001;
Damelin et al, 2002; Ng et al, 2002). However, we find that
combinations of rsc4-2 with hat or hdac mutations show
opposite relationships, lethality versus suppression, strongly
suggesting that Rsc4 promotes activation. In keeping with
this notion, we observe downregulation of several classes of
genes at nonpermissive temperatures including several in the
nicotinic acid biosynthesis pathway. Although microarray
cannot determine whether an effect is direct, we note that
several genes downregulated in the rsc4 BD
Ts
microarray
profiles are occupied by RSC (Damelin et al, 2002; Ng et al,
2002). This set includes genes that demonstrate some of the
strongest downregulation, such as BNA6, TNA1, SMX3, and
YKR049C, suggesting that these may be direct targets of Rsc4.
Nicotinic acid pathway genes are unlikely to underlie the
temperature sensitivity of rsc4 BD
Ts
strains, however, as
nicotinic acid supplementation provides no suppression.
Although we also observe genes that are upregulated in the
mutants, few of these show significant occupancy by RSC.
Recently, bdf1D gene expression profiles have shown a sub-
telomeric bias of downregulated genes, suggesting that Bdf1
forms an antisilencing barrier at telomeres and other hetero-
chromatin–euchromatin boundaries (Ladurner et al, 2003).
However, genes regulated in rsc4 BD
Ts
mutants do not
cluster to telomeric or subtelomeric regions, further high-
lighting the differences between Rsc4 and Bdf1 function.
In conclusion, we show that Rsc4 bromodomains are
essential for viability, directly recognize acetylated H3
Lys14, cooperate with H3 Lys14 acetylation for their function,
Rsc4 recognizes acetylated H3 Lys14
M Kasten et al
The EMBO Journal VOL 23
|
NO 6
|
2004 & 2004 European Molecular Biology Organization1356
show suppression relationships with certain HDACs, and are
important for the expression of certain genes. These results
highlight the important role bromodomains play in the
coordination of chromatin remodeling with histone acetyla-
tion in transcriptional regulation, and reveal new properties
of tandem bromodomain function.
Materials and methods
Media, genetic methods, and strains
Standard procedures were used for media preparation, transforma-
tions, integrations, sporulation, and tetrad analysis. All strains are
derivatives of S288C. Strain genotypes are listed in Table III.
Plasmids
Details of plasmid constructions are available upon request. All
plasmid constructs generated through PCR were sequence verified.
The RSC4 gene was isolated from a genomic cosmid as a 4.2 kb
BamHI fragment and subcloned into pRS316 (URA3, CEN). PCR-
generated ORF of 6xMyc.RSC4 was subcloned into the methionine-
repressible p415.MET25 (Mumberg et al, 1994) to create p603.
6xMyc.rsc4 BD1D (p804) lacks aa 59–156, 6xMyc.rsc4 BD2D
(p1172) lacks aa 193–290, 6xMyc.rsc4 BD1-2 D (p805) lacks aa
59–290, and 6xMyc.rsc4 C-termD (p809) inserts a stop codon after
aa 602. The rsc4 BD
Ts
mutants were isolated as N-terminal Myc
fusions in p415MET25. The pRS314 series of rsc4 BD
Ts
mutant
plasmids were created by subcloning the indicated RSC4 derivative
without Myc tags into pRS314 (TRP1, CEN). DNA fragments
encoding WT Rsc4 (p1513) or Rsc4 SDM3 (p1516) tagged with
10X HIS and two Myc epitopes at the N-terminus were subcloned
into pCITE-4b (Novagen) for in vitro transcription and translation.
Isolation of rsc4 BD
Ts
mutants and site-directed mutants
The region encoding the double bromodomain (bp 92–982) of RSC4
was mutagenized utilizing the inherent error rate of Ta q in PCR
using oligonucleotides BC796 (5
0
GGAAAACATCCTAAAAACCAAG
3
0
) and BC797 (5
0
GATATAGCTAACGCACCTGCTG 3
0
). The PCR
product was co-transformed into YBC627 with 6xMyc.rsc4 BD1-2D
(p805) linearized with NotI. The PCR product was inserted into the
gapped plasmid by homologous recombination and cells bearing
the repaired plasmid were selected on medium lacking leucine and
pooled together to create a library (31000 colonies). Cells were
plated and 2500 colonies lacking pRS316.RSC4 were selected with 5-
FOA and tested for conditional growth at 381C via replica plating.
Plasmids were recovered, subcloned, and fully sequenced. The rsc4-
7 allele has a single mutation in BD2 (L241P), whereas rsc4-11 bears
a mutation in BD1 and in the linker between the two bromodomains
(C130S, L182P). Additional Ts
alleles used were as follows: rsc4-5
D122N, N286S; rsc4-6 D66V, Y133H; rsc4-9 Y38C, V147I, D277V,
L297S; rsc4-10 N59Y, V260A, I274V; rsc4-12 Q131R, L189E; rsc4-13
L120P; rsc4-14 M145V, M149V, T279P.
Plasmids bearing site-directed mutations in RSC4 were prepared
using the Quick Change Method (Stratagene), subcloned, and fully
sequenced.
Analysis of rsc4-2 with histone tail mutants
The rsc4-2 allele was exchanged for the WT allele at the
chromosomal RSC4 locus in WZY63 (gift from S Dent) by targeted
replacement to create strain YBC1931. YBC1931 was transformed
with a series of TRP1-marked plasmids containing mutant H3 alleles
(with WT H4), mutant H4 alleles (with WT H3), or mutations in
both histone genes (gifts from S Dent). Loss of the URA3-marked
plasmid bearing WT H3 and H4 was enforced through growth on 5-
FOA. Phenotypes of the rsc4-2 histone double mutations were
assessed by spot dilution analysis at 331C, the highest temperature
at which the rsc4-2 allele displayed WT growth.
Anti-Rsc4 antibody
Recombinant full-length Rsc4 with a 10X-histidine tag was
expressed in Escherichia coli BL21 Codon ( þ ) cells (Stratagene)
and purified by binding to nickel resin (Qiagen) under denaturing
conditions. A measure of 2 mg of purified protein was used for
injection into rabbits (Covance Inc.).
Extract preparation and immunoprecipitations
Whole-cell extracts were prepared as described previously (Cairns
et al, 1999). For RSC assembly assays, all Rsc4 derivatives were
tagged with Myc epitopes and analyzed in the presence of untagged
WT RSC4 (p164) for complementation. For the rsc4 BD
Ts
mutants,
cultures were grown at 351C in SC media lacking leucine and uracil.
Anti-Myc antibody (clone 9E10) was conjugated to protein G beads
and incubated with 400 mg extract for 3 h. Precipitates were
recovered, washed twice with IP wash buffer (50 mM Tris–Cl (pH
7.5), 250 mM NaCl, 1 mM EDTA, 10% glycerol, 0.05% Tween-20),
and eluted in 4 SDS sample buffer. SDS–PAGE gels were
immunoblotted to PVDF membrane and developed as indicated.
Rsc4 TBD protein purification
Recombinant Rsc4 bearing tandem bromodomains (Rsc4 TBD,
residues 46–321) and both a Flag and 10X-histidine tag (N- and C-
terminal, respectively) was expressed in BL21 DE3 E. coli and
purified using standard methods. The purified protein was analyzed
by size exclusion chromatography on a Superdex 200 column
(Amersham Pharmacia Biotech) followed by slot blot analysis of
odd fractions utilizing anti-Rsc4 antiserum.
Table III Yeast strains
Strains Mating
type
Genotype Source
YBC627 MAT a rsc4D::HIS3 [p164; RSC4; CEN URA3] leu2D1 ura3-52 trp1D63 his3D200 lys2-12d This work
YBC1622 MAT a gcn5D::LEU2 rsc4D::HIS3 [p164; RSC4; CEN URA3] leu2D1 ura3-52 trp1D63 his3D200 lys2-12d This work
YBC1796 MAT a esa1D::HIS3 esa1-L327S::LEU2 rsc4D::HIS3 [p164; RSC4; CEN URA3]
leu2 ura3-52 trp1 his3D200 lys2-12d
This work
YBC1947 MAT a esa1D::HIS3 esa1-L254P::LEU2 rsc4D::HIS3 [p164; RSC4; CEN URA3] leu2
ura3-52 trp1 his3D200 lys2-12d
This work
WZY63 MAT a hht1-hhf1::pWZ405-F2F9-LEU2 hht2-hhf2::pWZ403-F4F10-HIS3 [pWZ414-F13;
HHT2-HHF2; CEN TRP1] ura3-52 lys2-801 trp1D63 his3D200 leu2D1
Zhang et al (1998)
YBC1924 MAT a Same as WZY63 except rsc4-2 integrated into RSC4 locus This work
YBC1931 MAT a Same as YBC1924 except Ycp50-copyII (HHT2-HHF2) instead of pWZ414-F13 This work
WZY42 MAT a Same as WZY63 except Ycp50-copyII (HHT2-HHF2) instead of pWZ414-F13 Zhang et al (1998)
YBC1698 MAT a rpd3D::KanMX rsc4D::HIS3 [p164; RSC4; CEN URA3] leu2 ura3 trp1D63 his3 lys2 This work
YBC1821 MAT a hda1D::KanMX rsc4D::HIS3 [p164; RSC4; CEN URA3] leu2 ura3 trp1D63 his3 lys2 This work
YBC1702 MAT a hos1D::KanMX rsc4D::HIS3 [p164; RSC4; CEN URA3] leu2 ura3 trp1D63 his3 lys2 This work
YBC1727 MAT a hos2D::KanMX rsc4D::HIS3 [p164; RSC4; CEN URA3] leu2 ura3 trp1D63 his3 lys2 This work
YBC1789 MAT a hos3D::KanMX rsc4D::HIS3 [p164; RSC4; CEN URA3] leu2 ura3 trp1D63 his3 lys2 This work
YBC1882 MAT a set3D::KanMX rsc4D::HIS3 [p164; RSC4; CEN URA3] leu2 ura3 trp1D63 his3 lys2 This work
YBC1892 MAT a sir2D::KanMX rsc4D::HIS3 [p164; RSC4; CEN URA3] leu2D1 ura3-52 trp1D63 his3D200 lys2-12d This work
Rsc4 recognizes acetylated H3 Lys14
M Kasten et al
& 2004 European Molecular Biology Organization The EMBO Journal VOL 23
|
NO 6
|
2004 1357
Histone tail binding assay
Biotinylated histone tail peptides were bound to streptavidin beads
(Invitrogen) in a ratio of 20 nmol of peptide to 100 ml bed volume
of beads in peptide binding buffer (PBB) (20 mM Tris (pH 7.5),
150 mM NaCl, 5% glycerol, 0.05% Tween-20, 1 mM EDTA, 1 mM
b-mercaptoethanol, protease inhibitors) with 0.5 mg/ml protease-
free BSA. Beads were washed five times with PBB, blocked
with protease-free BSA, and resuspended in a 50% slurry with
PBB and 0.5 mg/ml protease-free BSA. WT and SDM3 mutant
10xHIS-2xMyc-Rsc4 were in vitro transcribed/translated using a
reticulocyte lysate system as described by the manufacturer
(Promega) with the addition of 10 mM KCl. The lysate (20 ml)
was incubated for 1.5 h with 20 ml of peptide-bound beads, washed
twice with PBB and twice with PBB containing 200 mM NaCl.
The Rsc4 protein was eluted by boiling in 4 SDS sample buffer
prior to Western blot analysis. For binding studies involving
the recombinant Rsc4 TBD, 15 ml of the bead slurry was rotated
at 41C for 3 h with 500 ng of purified Rsc4 TBD. The beads
were washed twice with PBB and twice with PBB containing
250 mM NaCl. The Rsc4 TBD protein was eluted as above and
analyzed by Western blot.
Supplementary data
Supplementary data are available at The EMBO Journal Online.
Acknowledgements
We thank Sharon Dent and Lorraine Pillus for generous gifts of
strains and/or plasmids. We are grateful to Douglas Roberts, Daniel
Richardson, and Matthew Gordon for assistance with microarray
analysis. We also thank Alisha Schlichter and Pierre Thuriaux for
comments on the manuscript. This work was supported by the
National Institutes of Health (GM60415 to BRC; CA24014 for core
facilities) the Human Frontier Science Program and the Howard
Hughes Medical Institute. Margaret Kasten is a Research Associate
and Brad Cairns is an Assistant Investigator with the Howard
Hughes Medical Institute.
References
Allard S, Utley RT, Savard J, Clarke A, Grant P, Brandl CJ, Pillus L,
Workman JL, Cote J (1999) NuA4, an essential transcription
adaptor/histone H4 acetyltransferase complex containing Esa1p
and the ATM-related cofactor Tra1p. EMBO J 18: 5108–5119
Angus-Hill ML, Schlichter A, Roberts D, Erdjument-Bromage H,
Tempst P, Cairns BR (2001) A Rsc3/Rsc30 zinc cluster dimer
reveals novel roles for the chromatin remodeler RSC in gene
expression and cell cycle control. Mol Cell 7: 741–751
Cairns BR, Lorch Y, Li Y, Zhang M, Lacomis L, Erdjument-Bromage
H, Tempst P, Du J, Laurent B, Kornberg RD (1996) RSC, an
essential, abundant chromatin-remodeling complex. Cell 87:
1249–1260
Cairns BR, Schlichter A, Erdjument-Bromage H, Tempst P, Kornberg
RD, Winston F (1999) Two functionally distinct forms of the RSC
nucleosome-remodeling complex, containing essential AT hook,
BAH, and bromodomains. Mol Cell 4: 715–723
Clarke AS, Lowell JE, Jacobson SJ, Pillus L (1999) Esa1p is an
essential histone acetyltransferase required for cell cycle progres-
sion. Mol Cell Biol 19: 2515–2526
Damelin M, Simon I, Moy TI, Wilson B, Komili S, Tempst P,
Roth FP, Young RA, Cairns BR, Silver PA (2002) The genome-
wide localization of Rsc9, a component of the RSC chromatin-
remodeling complex, changes in response to stress. Mol Cell 9:
563–573
Dhalluin C, Carlson JE, Zeng L, He C, Aggarwal AK, Zhou MM
(1999) Structure and ligand of a histone acetyltransferase bro-
modomain. Nature 399: 491–496
Du J, Nasir I, Benton BK, Kladde MP, Laurent BC (1998) Sth1p, a
Saccharomyces cerevisiae Snf2p/Swi2p homolog, is an essential
ATPase in RSC and differs from Snf/Swi in its interactions
with histones and chromatin-associated proteins. Genetics 150:
987–1005
Gavin AC, Bosche M, Krause R, Grandi P, Marzioch M, Bauer A,
Schultz J, Rick JM, Michon AM, Cruciat CM, Remor M, Hofert C,
Schelder M, Brajenovic M, Ruffner H, Merino A, Klein K, Hudak
M, Dickson D, Rudi T, Gnau V, Bauch A, Bastuck S, Huhse B,
Leutwein C, Heurtier MA, Copley RR, Edelmann A, Querfurth E,
Rybin V, Drewes G, Raida M, Bouwmeester T, Bork P, Seraphin B,
Kuster B, Neubauer G, Superti-Furga G (2002) Functional orga-
nization of the yeast proteome by systematic analysis of protein
complexes. Nature 415: 141–147
Grant PA, Schieltz D, Pray-Grant MG, Steger DJ, Reese JC, Yates III
JR, Workman JL (1998) A subset of TAF(II)s are integral compo-
nents of the SAGA complex required for nucleosome acetylation
and transcriptional stimulation. Cell 94: 45–53
Grunstein M (1997) Histone acetylation in chromatin structure and
transcription. Nature 389: 349–352
Hassan AH, Prochasson P, Neely KE, Galasinski SC, Chandy M,
Carrozza MJ, Workman JL (2002) Function and selectivity of
bromodomains in anchoring chromatin-modifying complexes to
promoter nucleosomes. Cell 111: 369–379
Havas K, Whitehouse I, Owen-Hughes T (2001) ATP-dependent
chromatin remodeling activities. Cell Mol Life Sci 58: 673–682
Higgins DG, Thompson JD, Gibson TJ (1996) Using CLUSTAL for
multiple sequence alignments. Methods Enzymol 266: 383–402
Hsu JM, Huang J, Meluh PB, Laurent BC (2003) The yeast
RSC chromatin-remodeling complex is required for kineto-
chore function in chromosome segregation. Mol Cell Biol 23:
3202–3215
Hudson BP, Martinez-Yamout MA, Dyson HJ, Wright PE (2000)
Solution structure and acetyl-lysine binding activity of the GCN5
bromodomain. J Mol Biol 304: 355–370
Imai S, Armstrong CM, Kaeberlein M, Guarente L (2000)
Transcriptional silencing and longevity protein Sir2 is an NAD-
dependent histone deacetylase. Nature 403: 795–800
Jacobson RH, Ladurner AG, King DS, Tjian R (2000) Structure and
function of a human TAFII250 double bromodomain module.
Science 288: 1422–1425
Jenuwein T, Allis CD (2001) Translating the histone code. Science
293: 1074–1080
Kuo MH, Brownell JE, Sobel RE, Ranalli TA, Cook RG, Edmondson
DG, Roth SY, Allis CD (1996) Transcription-linked acetylation
by Gcn5p of histones H3 and H4 at specific lysines. Nature 383:
269–272
Ladurner AG, Inouye C, Jain R, Tjian R (2003) Bromodomains
mediate an acetyl-histone encoded antisilencing function at het-
erochromatin boundaries. Mol Cell 11 : 365–376
Lemon B, Inouye C, King DS, Tjian R (2001) Selectivity of chroma-
tin-remodelling cofactors for ligand-activated transcription.
Nature 414: 924–928
Matangkasombut O, Buratowski S (2003) Different sensitivities of
bromodomain factors 1 and 2 to histone H4 acetylation. Mol Cell
11 : 353–363
Moreira JM, Holmberg S (1999) Transcriptional repression of the
yeast CHA1 gene requires the chromatin-remodeling complex
RSC. EMBO J 18: 2836–2844
Mujtaba S, He Y, Zeng L, Farooq A, Carlson JE, Ott M, Verdin E,
Zhou MM (2002) Structural basis of lysine-acetylated HIV-1 Tat
recognition by PCAF bromodomain. Mol Cell 9: 575–586
Mumberg D, Muller R, Funk M (1994) Regulatable promoters of
Saccharomyces cerevisiae: comparison of transcriptional activity
and their use for heterologous expression. Nucleic Acids Res 22:
5767–5768
Narlikar GJ, Fan HY, Kingston RE (2002) Cooperation between
complexes that regulate chromatin structure and transcription.
Cell 108: 475–487
Ng HH, Robert F, Young RA, Struhl K (2002) Genome-wide location
and regulated recruitment of the RSC nucleosome-remodeling
complex. Genes Dev 16: 806–819
Ornaghi P, Ballario P, Lena AM, Gonzalez A, Filetici P (1999) The
bromodomain of Gcn5p interacts in vitro with specific residues in
the N terminus of histone H4. J Mol Biol 287: 1–7
Owen DJ, Ornaghi P, Yang JC, Lowe N, Evans PR, Ballario P,
Neuhaus D, Filetici P, Travers AA (2000) The structural basis
for the recognition of acetylated histone H4 by the bromodomain
of histone acetyltransferase gcn5p. EMBO J 19: 6141–6149
Rsc4 recognizes acetylated H3 Lys14
M Kasten et al
The EMBO Journal VOL 23
|
NO 6
|
2004 & 2004 European Molecular Biology Organization1358
Rundlett SE, Carmen AA, Suka N, Turner BM, Grunstein M
(1998) Transcriptional repression by UME6 involves
deacetylation of lysine 5 of histone H4 by RPD3. Nature 392:
831–835
Sanders SL, Jennings J, Canutescu A, Link AJ, Weil PA (2002)
Proteomics of the eukaryotic transcription machinery: identifica-
tion of proteins associated with components of yeast TFIID
by multidimensional mass spectrometry. Mol Cell Biol 22:
4723–4738
Suka N, Luo K, Grunstein M (2002) Sir2p and Sas2p opposingly
regulate acetylation of yeast histone H4 lysine16 and spreading of
heterochromatin. Nat Genet 32: 378–383
Suka N, Suka Y, Carmen AA, Wu J, Grunstein M (2001)
Highly specific antibodies determine histone acetylation site
usage in yeast heterochromatin and euchromatin. Mol Cell 8:
473–479
Tsuchiya E, Hosotani T, Miyakawa T (1998) A mutation in NPS1/
STH1, an essential gene encoding a component of a novel
chromatin-remodeling complex RSC, alters the chromatin struc-
ture of Saccharomyces cerevisiae centromeres. Nucleic Acids Res
26: 3286–3292
Vignali M, Hassan AH, Neely KE, Workman JL (2000) ATP-depen-
dent chromatin-remodeling complexes. Mol Cell Biol 20:
1899–1910
Zhang W, Bone JR, Edmondson DG, Turner BM, Roth SY (1998)
Essential and redundant functions of histone acetylation revealed
by mutation of target lysines and loss of the Gcn5p acetyltrans-
ferase. EMBO J 17: 3155–3167
Rsc4 recognizes acetylated H3 Lys14
M Kasten et al
& 2004 European Molecular Biology Organization The EMBO Journal VOL 23
|
NO 6
|
2004 1359
