Cell, Vol. 121, 873–885, June 17, 2005, Copyright ©2005 by Elsevier Inc. DOI 10.1016/j.cell.2005.04.031
Physical Association and Coordinate Function
of the H3 K4 Methyltransferase MLL1
and the H4 K16 Acetyltransferase MOF
Yali Dou,1Thomas A. Milne,2,4Alan J. Tackett,3
Edwin R. Smith,2Aya Fukuda,1Joanna Wysocka,2
C. David Allis,2Brian T. Chait,3Jay L. Hess,4
and Robert G. Roeder1,*
1Laboratory of Biochemistry and Molecular Biology
2Laboratory of Chromatin Biology
3Laboratory of Mass Spectrometry and Gaseous
The Rockefeller University
New York, New York 10021
4Department of Pathology and Laboratory Medicine
University of Pennsylvania School of Medicine
Philadelphia, Pennsylvania 19104
The mechanistic basis for this was revealed by the dis-
covery that many transcriptional coactivators, such as
GCN5, p300/CBP, and MOF, proved to be histone ace-
tyltransferases (Carrozza et al., 2003). The effects of
histone acetylation are probably additive, given the lack
of site specificity for most HATs, with the H4 K16 acety-
lation catalyzed by MOF and its homologs being among
the exceptions (Carrozza et al., 2003; Dion et al., 2005).
Recent studies show that acetylation of H4 K16 is criti-
cal for chromatin decondensation (Corona et al., 2002)
and gene activation in the X chromosome of male Dro-
sophila (Smith et al., 2001). There also is accumulating
evidence that H4 K16 acetylation may be the founding
acetylation event on histone H4 (Smith et al., 2003;
Turner et al., 1992; Zhang et al., 2002).
Another histone modification strongly correlated with
transcription activation in a wide variety of eukaryotic
systems is H3 K4 methylation, especially the trimethy-
lated state (Santos-Rosa et al., 2002; Strahl et al.,
1999). Histone acetylation and H3 K4 methylation not
only are functionally correlated but also physically
linked, as demonstrated by the preferential localization
of trimethylated K4 residues in hyperacetylated H3
(Zhang et al., 2004). This strongly suggests that the en-
zymatic machineries that add these two marks may
function in a synergistic manner and, potentially, in-
teract with each other. This notion is supported by
demonstrations of physical interactions between Tri-
thorax (TRX) and CBP in Drosophila (Petruk et al., 2001)
and a transient interaction between MLL1 and CBP in
human cells (Ernst et al., 2001). The mechanism un-
derlying the synergism between histone acetylation
and H3 K4 methylation, and its functional implications,
remain to be elucidated.
Studies to date have described a number of mamma-
lian H3 K4 methyltransferases (Sims et al., 2003). MLL1,
of interest here, is the product of a protooncogene that
was first detected through chromosomal translocations
directly associated with aggressive lymphoid and my-
eloid acute leukemias, especially among infants (Hess,
2004). MLL1 has been implicated in Hox gene regula-
tion in early embryogenesis (Yu et al., 1995) and, more
recently, in the regulation of non-Hox genes such as
p27 (Kip1) and p18 (Ink4C) (Milne et al., 2005). Consis-
tent with its proposed role in transcriptional regulation
through modification of histones, chromatin immuno-
precipitation (ChIP) assays have shown accumulation
of MLL1 and corresponding H3 K4 methylation marks
on active target genes (Milne et al., 2002).
Among the known H3 K4 methyltransferases in mam-
mals, several (hSET1 MLL1, MLL2) have been found in
large complexes (Goo et al., 2003; Hughes et al., 2004;
Nakamura et al., 2002; Wysocka et al., 2003; Yokoyama
et al., 2004). While sharing some subunits (e.g., Ash2L
and WDR5), these complexes nevertheless contain
unique sets of proteins that suggest nonoverlapping
functions. In the case of MLL1, one study reported that
proteins from six different complexes associated with
MLL1 to form a single stable “supercomplex” (Naka-
mura et al., 2002). This complex was shown to have
A stable complex containing MLL1 and MOF has been
immunoaffinity purified from a human cell line that
stably expresses an epitope-tagged WDR5 subunit.
Stable interactions between MLL1 and MOF were con-
firmed by reciprocal immunoprecipitation, cosedi-
mentation, and cotransfection analyses, and interac-
tion sites were mapped to MLL1 C-terminal and MOF
zinc finger domains. The purified complex has a ro-
bust MLL1-mediated histone methyltransferase activ-
ity that can effect mono-, di-, and trimethylation of H3
K4 and a MOF-mediated histone acetyltransferase ac-
tivity that is specific for H4 K16. Importantly, both
activities are required for optimal transcription activa-
tion on a chromatin template in vitro and on an en-
dogenous MLL1 target gene, Hox a9, in vivo. These
results indicate an activator-based mechanism for
joint MLL1 and MOF recruitment and targeted methyl-
ation and acetylation and provide a molecular expla-
nation for the closely correlated distribution of H3 K4
methylation and H4 K16 acetylation on active genes.
Many of the changes in chromatin structure induced by
transcription factors involve complex patterns of his-
tone modifications by enzymes such as histone acetyl-
transferases (HATs), histone methyltransferases (HMTs),
and kinases. Plasticity in transcription regulation and
coupled biological processes can be achieved by dy-
namic regulation of these histone modifications. The
varying format of chromatin modifications has led to
the hypothesis of a histone code, which suggests that
specific combinations of histone modifications dictate
specific transcriptional responses and cellular func-
tions (Strahl and Allis, 2000; Turner, 2002). Thus, it is
well established that transcriptionally active, euchro-
matic regions of the eukaryotic genomes are marked
by hyperacetylation of all four core histones, while
gene-poor, transcriptionally inactive heterochromatin
regions exhibit hypoacetylation (Vaquero et al., 2003).
a K4 methyltransferase activity, but which was much
weaker than that of the recombinant MLL1 SET domain
and unable to use dimethyl H3 K4 as substrate. A sub-
sequent study reported a much smaller MLL1-HCF
complex (six components in addition to MLL) with only
minimal overlap with the supercomplex but failed to
show any enzymatic activity of the complex (Yokoyama
et al., 2004).
To understand the function and mechanism of MLL1
H3 K4 methyltransferase activity in transcriptional regu-
lation, we set out to purify an MLL1-containing complex
for functional analysis in our chromatin-based in vitro
transcription assays (An et al., 2002). Given the pres-
ence in the MLL1 supercomplex of an ortholog (WDR5)
to a component in the yeast SET1 complex (Krogan et
al., 2002), we developed an immunoaffinity purification
procedure based on expression in HeLa cells of an epi-
tope-tagged WDR5. Surprisingly, our studies have re-
sulted in isolation of a complex that contains not only
MLL1, WDR5, and other expected components, but
newly isolated components that include, most notably,
MOF, the MYST family HAT that specifically acetylates
H4 K16. This has led us to investigate and to document
a functional connection between the MLL HMT and the
MOF HAT activities both in vitro and in vivo.
noblot (data not shown). The most notable new proteins
in our MLL1-WDR5 preparation include TAF compo-
nents of TFIID, components of the E2F6 subcomplex,
and MOF, a MYST family HAT. hSET1 and MLL2 were
not detected by mass spectrometry. CBP, which is
known to transiently interact with MLL1 (Ernst et al.,
2001), also was not detected.
The mass spectrometry results were confirmed by
immunoblot using available antibodies. In all cases
tested, proteins that were identified by mass spectrom-
etry were confirmed (Figure 2A). Menin, which was
found in both the MLL1-HCF complex and the MLL2
complex, was detected by immunoblot even though it
was not detected by mass spectrometry. The human
ortholog (hMSL1) of Drosophila MSL1, which is known
to interact with Drosophila MOF in the evolutionarily
conserved dosage compensation complex (Smith et al.,
2000), was not present in the MLL1-WDR5 complex
(Figure 2B). Human TIP60, another MYST family HAT
that shares sequence similarity with hMOF, also was
not detected. Extremely low levels of hSET1 and MLL2
were detected by immunoblot, but, consistent with the
separation of most SET1 and MLL2 from MLL1 in the
ion exchange steps (above), we were unable to detect
any unique components (e.g., Sin3A) of the hSET1
complex (Figure 2B).
Since the finding of hMOF in the MLL1-WDR5 com-
plex has important functional implications, our subse-
quent studies (described herein) were focused on in-
teractions of MOF with the MLL1-WDR5 complex.
Functional characterization of other identified proteins
is ongoing (see also below). In further confirmation of
the stable association of MOF with the MLL1-WDR5
complex, MOF was found to cosediment on a sucrose
gradient with other components (MLLC, Ash2L, RbBP5,
and Ring2) of the three step purified MLL1-WDR5 com-
plex (Figure 2C). The apparent size of the MOF-contain-
ing MLL1-WDR5 complex is around 1.5 MDa. Recipro-
cal immunoprecipitation of the three step purified
MLL1-WDR5 complex with anti-MOF antibody also re-
vealed coimmunoprecipitation of MLL1 with other
tested components (f-WDR5, RbBP5, and Ring2) of the
MLL1-WDR5 complex (Figure 2D). Importantly, TAFs,
hSET1, and MLL2 were not detected in the anti-MOF
immunoprecipitate. This suggests, for the M2 agarose
preparation, that the interactions between MLL1-WDR5
and MOF are specific, that trace levels of hSET1 and
MLL2 may reflect contaminating f-WDR5-containing
complexes, and that TAFs (presumably as TFIID) are
associated with a distinct (MOF-deficient) subfraction
of WDR5-containing complex(es). The copurification of
MOF with Flag-tagged RbBP5, another component of
the MLL1-WDR5 complex, provides further support for
a stable association of MOF with the MLL1-WDR5 com-
plex (Figure 2E).
Purification and Characterization
of an MLL1-WDR5 Complex
After establishment of a HeLa cell line that stably ex-
presses Flag-tagged WDR5 (f-WDR5), derived nuclear
extracts were subjected to a three step purification
(with MLL1 monitored by immunoblot) that involved
conventional column chromatography on phosphocel-
lulose (P11) and SP Sepharose, followed by immunoaf-
finity purification on anti-Flag antibody (M2 agarose)
beads (Figure 1A). Although WDR5 is common to MLL1
(Nakamura et al., 2002; Yokoyama et al., 2004), MLL2
(Hughes et al., 2004), and hSET1/HCF (Wysocka et al.,
2003) complexes, the first two steps efficiently separate
MLL1 from hSET1 (mostly in the P11 BC300 and SP
BC300 fractions) and MLL2 (mostly in P11 BC300 and
SP BC300 fractions) and remove most of the unincor-
porated f-WDR5 protein.
Mass spectrometric analysis of the purified f-WDR5-
containing MLL1 complex (designated MLL1-WDR5)
revealed 29 proteins (listed alongside the zinc stained
polyacrylamide gel, Figure 1B), including the tagged
WDR5, that were not present in the parallel mock-puri-
fied preparation derived from a conventional HeLa
nuclear extract (data not shown). Five of these proteins
are uncharacterized proteins in the database. Of the
remaining 24 proteins, six were recently reported in the
MLL1-HCF complex (Yokoyama et al., 2004). They are
the proteolytically derived MLLNand MLLCcompo-
nents of MLL1, ASH2L, RbBP5, WDR5, and HCF1. They
appear to represent the most tightly associated (“core”)
components of the MLL1 complex. With the notable ex-
ceptions of RbBP5, WDR5, and some TAFs, most of the
proteins that were reported in the earlier MLL1 complex
preparation (Nakamura et al., 2002) were not detected
in our preparation by mass spectrometry or immu-
MLL1 and MOF Interact Both In Vivo and In Vitro
To further substantiate intracellular MOF interactions
with the MLL1-WDR5 complex, 293T cells were trans-
fected with vectors expressing HA-tagged MLL1 (HA-
MLL1) and histidine-tagged MOF (His-MOF). Analysis
of derived cell lysates by Ni-NTA chromatography and
immunoblot revealed retention of MLL1-HA along with
MLL1 and MOF Interactions
Figure 1. Purification and Mass Spectrometric Analysis of the MLL1-WDR5 Complex
(A) Scheme for three step purification of the MLL1-WDR5 complex.
(B) Zinc-stained 4%–20% gradient gel of proteins eluted from M2 agarose beads. Proteins identified by MALDI mass spectrometry are
indicated. Several proteins were recovered from more than one slice.
His-MOF when both were coexpressed but not when
either was expressed alone (Figure 3A and data not
shown). Similar experiments with HA-MLL1 and His-
Ring2, another newly identified member of the MLL1-
WDR5 complex, and with HA-MOF and His-Ring2 re-
vealed MLL1-HA binding (Figure 3B) and MOF-HA
binding (Figure 3C) in conjunction with His-Ring2. Alto-
gether, these results provide further support for intra-
cellular interactions between MLL1, MOF, and Ring2
and their presence within a common complex.
Incubation of nuclear extract with GST-fused WDR5,
RbBP5, and MOF proteins also led to binding of endog-
enous MLL1 (data not shown), suggesting that MLL1
might serve as a scaffold for direct interaction of these
proteins and prompting an analysis of MLL1 domains
involved in the interactions. To this end, six MLL1 frag-
ments that cover both the MLLNregion (M1–M4) and
the MLLCregion (M5, M6) were generated by in vitro
transcription/translation and used in binding assays
(Figure 4C). MOF showed the strongest interaction with
the M5 fragment that contains the CID domain and a
weaker interaction with the M6 fragment that contains
the SET domain (Figure 4C). WDR5 and RbBP5 showed
similar strong and weak interactions with the M5 and
M6 fragments, respectively, but also showed significant
interactions with the M2 fragment that contains the
DNA methyltransferase homologous region (DNMT). A
further analysis with MLL1 M5 subfragments indicated
that a region (residues 3100–3300) C-terminal to the
CID domain is required for mediating the interaction
with MOF (Figure 4D). These results extend a previous
study, involving different assays, which suggested that
WDR5 and RbBP5 interactions with MLL1 are depen-
dent on (but not necessarily restricted to) the SET do-
main (Yokoyama et al., 2004).
A similar scheme for mapping MLL1 interacting do-
mains in MOF employed GST fusion proteins contain-
ing full length MOF, MOF (1–235) (with deletion of HAT
domain), MOF (1–171) (with deletion of both the HAT
domain and the zinc finger). Both full-length MOF and
MOF (1–235), but not MOF (1–171) lacking the zinc fin-
ger, bound the M5 fragment, suggesting that the zinc
finger domain is required for the MLL1 interaction (Fig-
ure 4E). Similar results were obtained using purified
recombinant MLLCprotein for the binding assay, indi-
cating that MOF directly interacts with MLLCin the
MLL1-WDR5 complex (Figure 4E).
The Purified MLL1-WDR5 Complex Has Both HMT
and HAT Activities
HMT and HAT activities of the three step purified MLL1-
WDR5 complex were assayed using both free histone
octamer and either native HeLa or recombinant nucleo-
somes as substrates. The purified MLL1-WDR5 com-
plex showed robust H3-specific HMT activities on all of
these substrates (Figure 5A). When assayed with a free
Figure 2. Immunoblot Confirmation of Com-
ponents of the Purified MLL1-WDR5 Complex
(A) Immunoblot of selected proteins iden-
tified by mass spectrometry. In, input, SP
Sepharose BC500 fraction; FT, flowthrough;
E, elution. Antibodies used in this and other
(B–E) blots are indicated on the left.
(B) Immunoblot analysis of select compo-
nents previously identified in other com-
plexes as controls.
(C) Immunoblot analysis of 3-step purified
complex following fractionation by sucrose
gradient sedimentation. Sedimentation posi-
tions of molecular weight markers are indi-
cated at the top.
(D) Immunoprecipitation (IP) of the three step
purified complex by anti-MOF antibodies.
Mouse IgG was used for control IP.
(E) Immunoblot analysis of three step puri-
fied complex through Flag-RbBP5.
recombinant histone octamer carrying the H3 K4Q mu-
tation, there was no detectable H3 methylation but, in-
stead, a low level of H2B methylation. In contrast, with
the more physiological (recombinant) nucleosome sub-
strate carrying the H3 K4Q mutation, all histone methyl-
ation was lost (Figure 5A). Hence, the HMT activity is
specific for H3 lysine 4 (K4).
To test the substrate specificity of the MLL1-WDR5
complex, H3 peptides that were either non-, mono-, di-,
or trimethylated on K4 were used in HMT assays (Figure
5B). All except the trimethylated H3 K4 peptide served
as substrates, although the dimethylated H3 K4 peptide
was a better substrate than unmodified and mono-
methylated K4 (Figure 5B). This result differs from that
reported with the recombinant SET domain of MLL1,
which failed to methylate dimethylated K4 peptides
(Milne et al., 2002; Nakamura et al., 2002). The failure
of the trimethylated H3 K4 peptide to be further methyl-
ated, even though it has unmethylated K9, further de-
monstrates the K4 specificity of the MLL1-WDR5
Consistent with the presence of MOF, the purified
MLL1-WDR5 complex also has a strong HAT activity.
With the free histone octamer substrate, the complex
shows a clear substrate preference for histone H4 but
also acetylates H3 and H2A (Figure 5C). In contrast, the
MLL1-WDR5 complex only acetylates H4 in nucleoso-
mal substrates, thus showing a much greater specific-
ity with the more physiological substrate. The same
change in acetylation pattern with different substrates
was previously observed for MOF in Drosophila (Akhtar
and Becker, 2000; Smith et al., 2000). Since MOF in
Drosophila was shown to specifically acetylate H4 ly-
sine 16 (Akhtar and Becker, 2000; Smith et al., 2000),
we also assayed the utilization of this site by the MLL1-
WDR5 complex. Using recombinant nucleosomes re-
constituted with H4 containing a K16 mutation, acetyla-
tion of H4 was dramatically decreased when compared
to H4 acetylation on a wild-type H4 nucleosome (Fig-
HMT and HAT Activities of the MLL1-WDR5 Complex
Act Coordinately in Transcriptional Regulation
Both H3 K4 trimethylation and H4 K16 acetylation are
marks for active transcription (see Introduction). To test
the function of the HMT and HAT activities of the MLL1-
MLL1 and MOF Interactions
Figure 3. Stable Intracellular Interactions of MLL1 and MOF
293T cells were transiently transfected with expression vectors encoding His-MOF and MLL1-HA (A), His-Ring2 and MLL1-HA (B), or His-
Ring2 and MOF-HA (C). For each analysis, proteins eluted from Ni-NTA were monitored by immunoblot with either anti-His or anti-HA antibody.
WDR5 complex in transcription regulation, we turned
to our established in vitro transcription system using a
recombinant chromatin template assembled with the
Acf-1/ISWI/NAP1 system (Figure 6B; An et al., 2004). In
this system, activator-mediated transcription is strongly
dependent upon recruitment of cofactors that effect
histone modifications (An et al., 2004). Our previous
demonstration of p53-induced accumulation of H3 K4
methylation on an endogenous p53 target gene (An et
al., 2004) raised the possibility that this might be due
to MLL1 recruitment by p53. Consistent with this possi-
bility, GST-p53, but not GST alone, was found to effi-
Figure 4. Mapping of MLL1 and MOF Interaction Domains
(A) Schematic representation of MLL1 and derived fragments (see Supplemental Data).
(B) SDS-PAGE analysis of purified recombinant proteins used either for in vitro binding experiments or in the chromatin assembly reaction
(see Figure 6B).
(C) Interaction of in vitro-translated,35S-methionine-labeled MLL1 fragments (indicated on the left) with GST-fusion proteins (indicated at
(D) Interactions of in vitro-translated,35S-methionine-labeled M5-derived fragments with GST-MOF. Full-size translated proteins in the inputs
are indicated by arrowheads.
(E) Interactions of MOF deletion mutants with MLL1 fragments. Indicated GST-MOF fusion proteins were incubated either with an in vitro
translated,35S-methionine-labeled M5 fragment or with a purified recombinant Flag-MLLCprotein.
Figure 5. The MLL1-WDR5 Complex Has
HMT and HAT Activities
(A) Methylation of H3 K4 by the MLL1-WDR5
complex on free histone octamers or on na-
tive or recombinant nucleosomes with wt or
mutated (K4Q) H3. Addition of purified
MLL1-WDR5 complex is indicated.
(B) Methylation by the MLL1-WDR5 complex
of H3 peptides with different levels of K4
(C) Acetylation of histones on free histone
octamers or nucleosomes with wt or mu-
tated (K16Q) H4. Addition of purified MLL1-
WDR5 complex is indicated.
ciently bind MLL1, MOF, and other components of the
MLL1-WDR5 complex from HeLa nuclear extracts (Fig-
ure 6A). This interaction allowed us to use p53 as the
activator for our in vitro transcription assay. When as-
sayed according to the scheme in Figure 6B, p53-
dependent transcription was significantly enhanced by
the MLL1-WDR5 complex when added with S-adenosyl
methionine (SAM; allowing methylation) or acetyl-CoA
(allowing acetylation). This activation is a direct result
of HMT and HAT activities of the MLL1-WDR5 complex
since adding the MLL1-WDR5 complex without cofac-
tors or adding cofactors without the MLL1-WDR5 com-
plex had no effect (Figure 6C). Higher levels of activity
that were at least equal to the sum of the individual
SAM and acetyl-CoA activities were observed when
SAM and acetyl-CoA were added together with the
MLL1-WDR5 complex, indicating a moderate coopera-
tivity between MLL1 and MOF (Figure 6C, lane 4). This
transcription activation is activator dependent, since
the transcription activity is greatly reduced in the ab-
sence of p53 (Figure 6C, lane 5). The results of several
independent transcription experiments are quantified
and summarized in Figure 6D. Further mechanistic
analysis showed p53-dependent H3 methylation and
H4 acetylation of the chromatin template by the MLL1-
MOF-WDR5 complex (Figure 6E). These results are
consistent with the results of the p53 interaction and
transcriptional activation studies and indicate that co-
activation by MLL1 and MOF involves targeted (p53-
dependent) histone acetylation and methylation.
MLL1 Targets Both H3 K4 Methylation and H4 K16
Acetylation Activities to the Hoxa9 Locus
Given the physical and functional interactions of MLL1
and MOF, we next examined H4 K16 acetylation and
H3 K4 trimethylation on well-established MLL1 target
genes in vivo. Three cell lines have been used: Mll(+/+)
and Mll(−/−)fibroblast lines established from day 10.5
embryos and an Mll(−/−)+ MLL1 line established by sta-
bly transfecting Mll(−/−)cells with a human Flag-tagged
MLL1 expression construct (Milne et al., 2002). Hoxa9
expression measured by quantitative RT-PCR showed
a greater than 5-fold difference in Mll(+/+)cells com-
pared with Mll(−/−)cells. However, ectopic expression
of MLL1 in Mll(−/−)cells restored much of the Hoxa9
expression (Figure 7A). To detect MLL1 binding at the
Hoxa9 locus, ChIP assays for four Hoxa9 regions were
performed (Figure 7B). These regions were located
w200–300 bp upstream of the first exon (1), within the
first exon (2,3) and within second exon that contains
the homeodomain (4) of Hoxa9 (Figure 7C). Using an
antibody against MLLCin the CHIP assay, and as ex-
pected, MLL1 binding was detected in all four regions
in wild-type Mll(+/+)cells and in Mll(−/−)+ MLL1 cells
(Figure 7D). As a control, binding at the Gapdh locus
was at a background level similar to that seen in Mll(−/−)
cells. Consistent with the binding of MLL1 to the Hoxa9
locus, changes of H3 K4 dimethylation and trimethyla-
tion tightly correlated with MLL1 binding to this locus
(regions 1-4). Fold differences between the di- and tri-
methylation levels of H3 K4 in Mll(+/+)and Mll(−/−)+
MLL1 and MOF Interactions
Figure 6. The HMT and HAT Activities of the MLL1-MOF Complex Modify Histones in an Activator-Dependent Manner and Act Coordinately
to Effect p53-Dependent Transcription on Chromatin
(A) Binding of the MLL1-WDR5 complex (from nuclear extract) to GST-p53. Bound proteins (indicated at left) were analyzed by immunoblot.
(B) Schematic of in vitro transcription assay using recombinant chromatin templates (An and Roeder, 2004).
(C) Chromatin-templated transcription assays with p53, purified MLL1-WDR5 complex, acetyl-CoA, and SAM addition as indicated and
according to the scheme in (B).
(D) Quantitation by phosphoimager of data as in (C) (lanes 1–5) is shown. y axis indicates transcription levels relative to lane 5 (p53-indepen-
dent basal activity). Average and standard deviations (error bars) from three autoradiographic analyses are indicated at the bottom.
(E) Histone modification assays on chromatin templates with p53, purified MLL1-WDR5 complex, acetyl-CoA, and SAM additions as indicated.
MLL1 cells compared with that in Mll(−/−)cells are
shown in Figures 7E and 7F. As a control, no change in
di- or trimethylation was seen at the Gapdh locus,
which does not have bound MLL1.
We next examined H4 acetylation using antibodies
against specific lysine residues, namely K5, K8, K12,
and K16, across the Hoxa9 locus in the three cell lines.
These four acetylation sites showed very different re-
sponses in relation to bound MLL1. There was no
change in K5 acetylation and only small increases in K8
and K12 acetylation at the Hoxa9 locus in Mll(+/+)cells
and in Mll(−/−)+ MLL1 cells when compared with
Mll(−/−)cells (Figures 7G, 7H, and 7I). In contrast, the
K16 acetylation level increased dramatically across the
Hoxa9 locus in both Mll(+/+)and Mll(−/−)+ MLL1 cells
(Figure 7J). Comparing the changes in H3 K4 methyla-
tion and in H4 K16 acetylation within the Hoxa9 locus,
a close correlation of these two marks was also ob-
served. There was gradual increase of H3 K4 methyla-
tion from regions 1 to 4, and a similar pattern was de-
Figure 7. Specific Increase of H4 K16 Acety-
lation upon MLL1 Binding at the Hoxa9 Pro-
(A) Reduced Hoxa9 expression in Mll(−/−)
cells and restoration by ectopic MLL1.
Hoxa9 expression in wt cells (dark gray) is
arbitrarily set as 100%. Standard deviations
in this and other experiments (D–J) are indi-
cated by error bars.
(B) Legend for (D)–(H). Dark gray, Mll(+/+)
cells; white, Mll(−/−)cells; light gray, Mll(−/−)+
ectopic MLL1-expressing cells.
(C) Four probe sets in the Hoxa9 locus used
for real-time PCR quantification of ChIP.
Black bars indicate CpG-rich regions of the
locus; arrows indicate two different tran-
scription start sites (Fujimoto et al., 1998);
putative TATA boxes are shown (Nakamura
et al., 2002), and the first and second exons
are shown as gray boxes. The homeodomain
(HD) in exon 2 is indicated by crosshatching.
(D–J) ChIP experiments in Mll(+/+), Mll(−/−),
and Mll(−/−)+ MLL1 cells using the indicated
antibodies. ChIP was quantified using the
Taqman primer/probe sets indicated in (C).
Signals in Mll(−/−)cells were set to 1, and sig-
nals in Mll(+/+)and Mll(−/−)+ MLL1 cells were
expressed as fold differences relative to
Mll(−/−)cells. (D) MLLC, (E) H3 di-MeK4, (F) H3
tri-MeK4, (G) H4 AcK5, (H) H4 AcK12, (I) H4
AcK8, (J) H4 AcK16.
tected for H4 K16 acetylation. Since, in all cases, we
scored the fold difference in histone modifications at
the Hoxa9 locus in Mll(+/+)and Mll(−/−)+ MLL1 cells to
that in Mll(−/−)cells, the changes in K4 methylation and
H4 K16 acetylation reported here are MLL1 dependent.
No MLL1-dependent changes in H3 K4 methylation or
H4 K16 acetylation were seen at the Gapdh locus,
which is not regulated by MLL1, suggesting that
changes observed at the Hoxa9 locus require MLL1
binding. Thus, these analyses have established a clear
correlation between H3 K4 methylation and H4 K16
acetylation at the Hoxa9 locus and, most significantly,
a strong dependence of the H4 K16 acetylation on the
presence of MLL1.
ther showed that HOXA9 expression was significantly
downregulated (>50% decrease) in the MOF knock-
down cells compared to control siRNA-treated cells
(Figure 8B). The expression of GAPDH, which is not an
MLL1 target, was less affected (inconsistent small re-
ductions were observed in some cases) by MOF siRNA
treatment. In all cases, HOXA9 and GAPDH expression
levels were normalized to total input RNA.
To detect H4 K16 acetylation and H3 K4 methylation
at the HOXA9 locus, ChIP assays were directed toward
a HOXA9 TATA-containing region (1) located w200–300
bp upstream of the first exon and a region (3) within the
first exon (Figure 7C). MOF knockdown resulted in a
significant reduction of H4 K16 acetylation both around
the TATA region (w50% decrease) and in the coding
region (w70% decrease). Consistent with the expres-
sion results, the K16 acetylation level on GAPDH was
less (and inconsistently) affected by MOF siRNA. No
significant changes in trimethylation at the HOXA9 lo-
cus were observed in MOF siRNA-treated cells (Figure
8D). This indicates that whereas H4 K16 acetylation by
MOF is dependent upon MLL1 (Figure 7), H3 K4 methyl-
ation by MLL1 can occur independently of MOF. Con-
sidering that MOF stably associates with MLL1, that
MLL1 and MOF coordinately activate transcription in
vitro, and that K16 acetylation at the HOXA9 locus is
MOF Is Required for the Expression of HOXA9
The in vivo function of MOF on the well-established
MLL1 target gene HOXA9 was further explored in HeLa
cells. MOF expression was effectively eliminated by
siRNA techniques (Figure 8A). As revealed by immu-
noblot using acetylated H3 as an internal control, MOF
siRNA resulted in a significant reduction of the global
H4 K16 acetylation level (w30%) relative to that ob-
served with control siRNA-treated cells. This indicates
that MOF is the major histone acetyltransferase for this
site in vivo. Quantitative real time RT-PCR analysis fur-
MLL1 and MOF Interactions
Figure 8. Loss of MOF Protein Results in a
Decrease of Histone H4 Lysine 16 Acetyla-
tion and HOXA9 Expression
(A) MOF siRNAs, M1 and M2, decrease MOF
protein and global H4 K16 acetylation levels.
Acetylated H3 was used as internal loading
(B) MOF siRNA treatment reduces HOXA9
HOXA9 and GAPDH in MOF siRNA-treated
cells (M1 and M2) are normalized to levels in
control cells (C). Standard deviations in (B)–
(C) are indicated by error bars.
(C) Histone H4 K16 acetylation at HOXA9
and GAPDH loci in MOF (M1 and M2) and
control (C) siRNA-treated cells. The K16
acetylation level in MOF siRNA-treated cells
is normalized to the level in control cells (C),
which is arbitrarily set as 1.
(D) Histone H3 K4 methylation at the HOXA9
locus in MOF (M1) and control (C) siRNA-
treated cells. H3 K4 methylation in control
cells (C) is arbitrarily set as 1.
MLL1 dependent, it is highly probable that downregula-
tion of HOXA9 by MOF siRNA treatment is specific
This idea is supported by our observation of two MLL1-
containing peaks following the fractionation of nuclear
extract by gel filtration (data not shown) and by previ-
ous indications of heterogeneity in the MLL1-HCF prep-
aration (Yokoyama et al., 2004).
That MOF is a bona fide component of an MLL1-con-
taining complex is indicated by (1) copurification
through ion-exchange chromatography and affinity pu-
rification steps (involving epitope-tagged WDR5 or
RbBP5), (2) cosedimentation on sucrose gradients of
MOF with MLL1 and other core components (WDR5,
Ash2L, RbBP5) of the purified MLL1-WDR5 complex,
(3) coimmunoprecipitation of MLL1 and select compo-
nents (including Ash2L, RbBP5, Ring2, and WDR5) of
the MLL1-WDR5 preparation using anti-MOF antibody,
(4) coimmunoprecipitation of MLL1, MOF, and Ring2
following pairwise expression in transfected cells, and
(5) direct in vitro interactions between MLL1 and MOF,
through mapped subdomains. The possibility that MOF
might also be associated with contaminating hSET1
and MLL2 complexes, which also contain WDR5, is
ruled out by the failure of anti-MOF antibodies to coim-
munoprecipitate the residual amounts of hSET1 or
MLL2 in the MLL1-WDR5 preparation and by MOF
coimmunoprecipitation with f-RbBP5 (which is not a
component of the hSET1 complex). This is consistent
with the greater abundance of MOF (detected by mass
spectrometry) relative to hSET1 or MLL2 (not detected
by mass spectrometry) in the MLL1-WDR5 preparation.
Although well-studied in Drosophila, very little is
known about MOF in mammalian cells. The Drosophila
ortholog (dMOF) of human MOF is important for dosage
compensation of the male X chromosome and acts
through an H4 K16 acetylation mechanism that effects
a 2-fold activation of transcription (Smith et al., 2000).
In Drosophila, dMOF function in dosage compensation
depends on its integration into a complex with male-
specific MSL1, MSL2, and other components (Morales
et al., 2004). It has been suggested that a similar dos-
age compensation complex also exists in mammals
Studies to date indicate that the chromatin modifica-
tions associated with gene activation are diverse and
involve the action of combinations of histone modifying
enzymes/coactivators (Fischle et al., 2003; Turner,
2002). In studies directed toward the function of human
MLL1 in transcription activation through H3 K4 methyl-
ation, we have identified a stable complex (MLL1-
WDR5) containing both MLL1 and the MYST family
histone acetyltransferase MOF. The relevance of the
demonstrated ability of this complex to effect both H3
K4 methylation (mono-, di-, tri-) and H4 K16 acetylation,
to interact with a transcriptional activator, and to stimu-
late activator-dependent transcription through the resi-
dent HMT and HAT activities in vitro is underscored by
our concomitant demonstration of MLL1-dependent H3
K4 methylation and H4 K16 acetylation events on a
known MLL1 target gene. Altogether, these results indi-
cate a coordinate function of MLL1 and MOF, through
a physical association, in gene activation events. Along
with complementary studies (Wysocka et. al., 2005 [this
issue of Cell]) indicating direct interactions of WDR5,
an MLL1-interacting component of the MLL1-MOF
complex, with methylated H3 K4 residues, these results
also lead to a model for both the establishment and
spreading of K4 methylation, perhaps in conjunction
with H4 K16 acetylation, in transcriptionally active chro-
MLL1 and MOF Can Form a Stable Complex In Vivo
The MLL1-WDR5 complex purified on the basis of an
intrinsic epitope-tagged subunit (WDR5) represents a
unique subpopulation of cellular MLL1 complexes with
novel components (notably MOF; see Table S1 in the
Supplemental Data available with this article online).
(Neal et al., 2000), but its function remains enigmatic
since mammals use a totally different mechanism for
dosage compensation. The presence in the MLL1-
WDR5 complex of MOF suggests that, in mammals,
MOF probably plays an alternative role in transcription
regulation that is independent of the dosage compen-
chromosome (Morales et al., 2004), analogous interac-
tions of hMOF with MLL1 through the same zinc finger
region may imply new functions for MOF, possibly by
targeting different sets of genes through its interaction
HMT and HAT Activities of the MLL1-WDR5 Complex
Can Act Coordinately in Transcriptional Regulation
The use of a chromatin-templated assay in which tran-
scription is dependent upon a transcriptional activator
and interacting chromatin modifying cofactors has al-
lowed us to document in vitro functions of the MLL1
HMT and MOF HAT activities. Thus, in the absence of
other factors (p300, CARM1, PRMT1) previously shown
to serve as coactivators in this assay (An et al., 2004),
p53-mediated transcription is dependent upon addition
of the MLL1-MOF-WDR5 complex and either SAM or
acetyl-CoA. The dependence on these cofactors indi-
cates that the coactivator functions can be attributed
to the histone methyltransferase and histone acetyl-
transferase activities of MLL1 and MOF, respectively.
The inability of p53 to enhance transcription upon addi-
tion of SAM and acetyl-CoA without the MLL1-WDR5
complex indicates that the observed activation by p53
is not due to endogenous MLL1 or MOF. Reciprocally,
the lack of an effect of the MLL1-WDR5 complex (with
SAM and acetyl-CoA) in the absence of p53 indicates
that transcription activation does not result from gen-
eral (nontargeted) modifications of the chromatin tem-
plate. This is further indicated by our demonstration of
p53-dependent (targeted) methylation and acetylation
of chromatin template histones by MLL1 and MOF. The
fact that histone modifications by the MLL1-MOF-
WDR5 complex are p53 dependent and correlate with
transcriptional activation indicates a direct involvement
of MLL1 and MOF in p53-dependent transcription. This
is consistent with our demonstrated interaction of the
MLL1-WDR5 complex with p53, as well as prior indi-
cations of p53-induced accumulation of methylated H3
K4 on a p53 target gene (An et al., 2004). The possible
involvement of the MLL1-WDR5 complex in transcrip-
tion initiation is consistent with the reported enrichment
of di- and trimethylated H3 K4 at active promoters in
higher eukaryotes (Santos-Rosa et al., 2002; Schneider
et al., 2004) and is further supported by studies indicat-
ing enhanced transcription of a reporter with Gal4 bind-
ing sites following expression (and artificial recruitment)
of a Gal4-WDR5 fusion protein (Wysocka et al., 2005).
The coordinate action of MLL1 HMT and MOF HAT
activities in the in vitro assays is in agreement with the
close correlation between H3 K4 methylation and his-
tone acetylation marks in vivo but describes coopera-
tion that is site specific and involves only H3 K4 and
H4 K16. In contrast to other acetylated residues on H4,
acetylated K16 is known to play a pivotal role in deter-
mining the potential of coding DNA for expression or
silencing. Thus, (1) only the H4 K16 mutation has spe-
cific transcription consequences independent of the
mutational state of the other lysines (Dion et al., 2005),
(2) in yeast, H4 K16 is the only residue whose acetyla-
tion can, on its own, prevent silencing of the mating
type genes (Johnson et al., 1990; Megee et al., 1990),
and (3) MOF can activate transcription in yeast when
The MLL1-WDR5 Complex Is Enzymatically Active
Although lysine residues may be mono-, di-, or trimeth-
ylated in vivo, the trimethyl H3 K4 is preferentially and
strongly associated with the transcribed regions of
active genes from yeast to higher eukaryotes (Krogan
et al., 2002; Ng et al., 2003; Santos-Rosa et al., 2002;
Schneider et al., 2004). In yeast, SET1 is the only H3 K4
methyltransferase and can methylate H3 K4 to all three
levels. In contrast, higher eukaryotes contain several
H3 K4 methyltransferases with various specificities,
suggesting a more complex picture for the regulation
and function of H3 K4 methylation in higher eukaryotes
(Sims et al., 2003).
The purified MLL1-WDR5 complex shows a robust
H3 K4 methyltransferase activity on H3 peptide, free
histone octamer, and nucleosomal substrates. Impor-
tantly, like SET1 in yeast, the MLL1-WDR5 complex is
active with non-, mono-, and dimethylated substrates
and, in particular, efficiently converts dimethylated H3
K4 residues (the preferred substrate) to trimethylated
residues. This result is consistent with the view that
MLL1 is directly involved in transcription activation of
Hox genes as a consequence of the enhanced levels of
H3 K4 trimethylation associated with MLL1 binding to
Hox gene promoters in vivo. This result contrasts with
the reported inability of the MLL1 supercomplex (weak
HMT activity) or the MLL1 SET domain (much stronger
HMT activity) to effect H3 K4 trimethylation. Thus, and
consistent with its distinct composition relative to that
of the first reported MLL1 complex (Nakamura et al.,
2002), the MLL1-WDR5 complex described here is
more likely to be the functionally active MLL1 complex.
Since a recombinant MLLCfragment (180 kDa) shows
extremely weak activity compared with the MLL1-
WDR5 complex (data not shown), it appears that MLL1,
like yeast SET1, must be associated with other compo-
nents for full activity. With respect to the MLL1-associ-
ated HAT activity of MOF, Drosophila MOF has an un-
usually narrow substrate specificity and only acetylates
histone H4 K16 on a nucleosomal template (Akhtar and
Becker, 2000; Smith et al., 2000). Moreover, faithful and
efficient acetylation of nucleosomal histone H4 by MOF
is observed only upon interaction with MSL1 and MSL3
(Morales et al., 2004). MOF in the MLL1-WDR5 complex
shows a similar preference for H4 K16 on nucleosomes.
In fact, the observed reduction of global H4 K16 acety-
lation in MOF knockdown cells suggests that MOF is
one of the major histone acetyltransferases involved in
H4 K16 acetylation and may play important roles in
functions other than dosage compensation.
Analysis of the MOF and MLL1 interaction domains
has further shown that MLL1, via a C-terminal domain,
interacts with the zinc finger of MOF. Given that an in-
teraction of dMOF with MSL1 through its zinc finger is
essential for correct targeting of MOF to the male X
MLL1 and MOF Interactions
according to the scheme in Figure 1A. For sucrose gradient sedi-
mentation, 500 ?l purified MLL1-WDR5 complex (from 50 ml
nuclear extract) was loaded onto an 11 ml 10% to 40% (w/v) su-
crose gradient. After centrifugation for 16 hr at 23,000 rpm (SW41
rotor), 0.5 ml fractions were collected.
tethered to a promoter via a DNA binding domain, sug-
gesting that a rather local acetylation may lead to a
significant stimulation of transcription (Akhtar and
Becker, 2000). The potentially cooperative in vivo func-
tions between H3 K4 methylation and H4 K16 acetyla-
tion have clearly been demonstrated by the downregu-
lation of a well-established MLL1 target gene, Hoxa9,
in MOF siRNA-treated cells and by a specific increase
in H4 K16 acetylation, which is dependent on MLL1 and
closely correlated with the MLL1-dependent increase
in H3 K4 methylation, in the same locus. Our hypothesis
of coordinate MLL1 and MOF function is further sup-
ported by our demonstration that MOF, like other (e.g.,
RbBP5) components of MLL1-WDR5 complex, can
bind to methylated H3 K4 through WDR5, arguing for
the joint recruitment of HMT and HAT activities to the
same targets (Figure S2).
Immunoaffinity Purification of the MLL1-WDR5 Complex
and Mass Spectrometry
The SP Sepharose BC500 fraction was incubated with M2 agarose
in BC300, 0.05%NP40 at 4°C for 4 hr and extensively washed with
BC500, 0.05% NP40. The complex was eluted with 0.25 mg/ml Flag
peptide in BC100. The eluted MLL1-WDR5 complex was resolved
by SDS-PAGE and visualized by zinc staining. The entire gel lane
was sliced into 2 mm bands and proteins were subjected to MALDI
mass spectrometry (Krutchinsky et al., 2000; Krutchinsky et al.,
2001). Proteins were identified by XProteo (Chao Zhang; http://
www.xproteo.com). Common background proteins (i.e., keratins,
tubulins, ribosomal proteins) were excluded from the list of in-
teracting proteins. We also note that WDR5-Flag was identified in
multiple bands, which may be due to breakdown and/or over-
Model for Transcription Activation
Our demonstration of transcriptional activation through
the HMT (MLL1) and HAT (MOF) activities in the MLL1-
WDR5 complex, direct interaction of the complex with
a DNA binding transcriptional activator, and binding of
a resident subunit (WDR5) to methylated H3 K4 (Wy-
socka et al., 2005) lead to a model for both the estab-
lishment and spreading of an active chromatin struc-
ture. The model invokes activator binding to DNA
regulatory elements, activator-mediated recruitment of
the MLL1-WDR5 complex, consequent methylation and
acetylation, and either the recruitment of additional
MLL1-WDR5 complexes or internucleosomal transfer
of primary MLL1-WDR5 complexes through binding of
the WDR5 subunit to methylated H3 K4 or through
binding of the MLL1 bromodomain to acetylated H4
K16 residue. This model is related, in part, to that pro-
posed for propagation of heterochromatin by H3 K9
methylation and subsequent HP1 binding (Lachner et
Previously established H3 K4 methylation marks can
also be recognized by other effectors, such as CHD1 in
the SAGA complex (Pray-Grant et al., 2005) and com-
ponents in ATP-dependent chromatin remodeling com-
plexes (Beisel et al., 2002; Santos-Rosa et al., 2003).
Similarly, acetylation marks can also be recognized by
various bromodomains (Hassan et al., 2001) that are
common structural features of many histone acetyl-
transferases (Turner, 2002). Recruitment of other HAT
complexes and ATP-remodeling complexes by mole-
cules that directly or indirectly interact with methyl- or
acetylmarks may further facilitate the spreading of
“open chromatin” and enable RNA polymerase to effi-
ciently move through the chromatin template for
Anti-MLLCantibody was generated using the previously described
epitope (Hsieh et al., 2003). Other antibodies were obtained com-
mercially as follows: anti-MLL-N, hSET1, MLL2, menin, RbBP5, and
Ash2 (Bethyl Laboratory); anti-TIP60 (Santa Cruz); mouse IgG, M2,
and M2 agarose (Sigma); anti-His (Qiagen); anti-HA (Roche); anti-
H3 dimethyl K4 and trimethyl K4 (Abcam); anti-H4 acetyl K5, acetyl
K8, acetyl K12, and acetyl K16 (Upstate Biotechnology). Anti-MOF
and anti-MSL1 antibodies will be described elsewhere (E.R.S., C.
Cayrou, R. Huang, W.S. Lane, J. Cote, and J.C. Lucchesi, unpub-
Plasmids and Expression Vectors
Bacterial vectors for core histone expression and purification were
as described (Luger et al., 1999). Mutations (H3 K4 and H4 K16)
were introduced by PCR-based site-directed mutagenesis. Flag-
tagged p53 and the DNA template for transcription were as de-
scribed (An et al., 2004). GST and GST-tagged WDR5, RbBP5, MOF
full length, MOF (1–171), and MOF (1–235) proteins were expressed
in bacteria from the pGEX4T-1 vector (Amersham). A Flag-tagged
MLLCcDNA was inserted into pVL1392 baculovirus vector (Bac-
vector 3000, Novagen). For transfection assays, cDNAs encoding
tagged MLL1, MOF, and Ring2 proteins were inserted into the CMV-
driven expression vector pIRESneo (Clontech). Manufacturer’s pro-
tocols for the TNT Quick coupled transcription/translation system
(Promega) were used for in vitro translation. Information for in vitro
translation templates is in the Supplemental Data.
GST Pull-Down Assay
For GST pull-down assays, 4 ?g GST-tagged protein and 200 ?l
HeLa NE, 10 ?l of the TNT translation reaction or 2 ?g purified
recombinant MLLCwere used in each binding assay. Reactions
were carried out at 4°C for 4 hr and beads were washed three times
Histone Modification Assays
Unmodified, mono-, di-, and trimethylated H3 K4 peptides were
from Upstate Biotechnology. HeLa nucleosomes were purified as
described (Owen-Hughes et al., 1999). Recombinant nucleosomes
were prepared by salt dialysis using 5S array DNA from G5ML
(Kundu et al., 2000) and recombinant histone octamers were recon-
stituted as described (Luger et al., 1999). For each HMT or HAT
assay, 5 ?g peptide, 2 ?g HeLa nucleosomes, or 2 ?g recombinant
nucleosomes were used. Reactions were carried out at 30°C for 1
hr in the presence of [3H]-SAM (S-adenosyl-L-[methyl-3H] methio-
nine) or [3H]-acetyl-CoA.
Mll(+/+), Mll(−/−), and Mll(−/−)+ MLL cell lines have been described
previously (Milne et al., 2002). The f-WDR5 and f-RbBP5 cell lines
were made by transfecting HeLa S3 cells with either an f-WDR5-
pIRESneo vector or an f-RbBP5-pIRESneo vector.
Chromatin Modifications and Transcription Assays
Chromatin assembly and histone modification reactions were car-
ried out with coactivators essentially as described (An and Roeder,
Fractionation of Nuclear Extracts
Nuclear extracts were obtained from f-WDR5 or f-RbBP5 cells by a
modified Dignam procedure (Dignam et al., 1983) and fractionated
2004). Transcription assays included 40 ng p53 and about 100 ng
purified MLL1-WDR5 for each reaction.
transcription initiation by RNA polymerase II in a soluble extract
from isolated mammalian nuclei. Nucleic Acids Res. 11, 1475–1489.
Dion, M.F., Altschuler, S.J., Wu, L.F., and Rando, O.J. (2005). From
the cover: genomic characterization reveals a simple histone H4
acetylation code. Proc. Natl. Acad. Sci. USA 102, 5501–5506.
Ernst, P., Wang, J., Huang, M., Goodman, R.H., and Korsmeyer, S.J.
(2001). MLL and CREB bind cooperatively to the nuclear coactiva-
tor CREB-binding protein. Mol. Cell. Biol. 21, 2249–2258.
Fischle, W., Wang, Y., and Allis, C.D. (2003). Histone and chromatin
cross-talk. Curr. Opin. Cell Biol. 15, 172–183.
Fujimoto, S., Araki, K., Chisaka, O., Araki, M., Takagi, K., and Yama-
mura, K. (1998). Analysis of the murine Hoxa-9 cDNA: an alterna-
tively spliced transcript encodes a truncated protein lacking the
homeodomain. Gene 209, 77–85.
Goo, Y.H., Sohn, Y.C., Kim, D.H., Kim, S.W., Kang, M.J., Jung, D.J.,
Kwak, E., Barlev, N.A., Berger, S.L., Chow, V.T., et al. (2003). Activa-
ting signal cointegrator 2 belongs to a novel steady-state complex
that contains a subset of trithorax group proteins. Mol. Cell. Biol.
Hassan, A.H., Neely, K.E., and Workman, J.L. (2001). Histone ace-
tyltransferase complexes stabilize SWI/SNF binding to promoter
nucleosomes. Cell 104, 817–827.
Hess, J.L. (2004). MLL: a histone methyltransferase disrupted in
leukemia. Trends Mol. Med. 10, 500–507.
Hsieh, J.J., Ernst, P., Erdjument-Bromage, H., Tempst, P., and Kors-
meyer, S.J. (2003). Proteolytic cleavage of MLL generates a com-
plex of N- and C-terminal fragments that confers protein stability
and subnuclear localization. Mol. Cell. Biol. 23, 186–194.
Hughes, C.M., Rozenblatt-Rosen, O., Milne, T.A., Copeland, T.D.,
Levine, S.S., Lee, J.C., Hayes, D.N., Shanmugam, K.S., Bhat-
tacharjee, A., Biondi, C.A., et al. (2004). Menin associates with a
trithorax family histone methyltransferase complex and with the
hoxc8 locus. Mol. Cell 13, 587–597.
Johnson, L.M., Kayne, P.S., Kahn, E.S., and Grunstein, M. (1990).
Genetic evidence for an interaction between SIR3 and histone H4
in the repression of the silent mating loci in Saccharomyces cere-
visiae. Proc. Natl. Acad. Sci. USA 87, 6286–6290.
Krogan, N.J., Dover, J., Khorrami, S., Greenblatt, J.F., Schneider, J.,
Johnston, M., and Shilatifard, A. (2002). COMPASS, a histone H3
(Lysine 4) methyltransferase required for telomeric silencing of
gene expression. J. Biol. Chem. 277, 10753–10755.
Krutchinsky, A.N., Zhang, W., and Chait, B.T. (2000). Rapidly switch-
able matrix-assisted laser desorption/ionization and electrospray
quadrupole-time-of-flight mass spectrometry for protein identifica-
tion. J. Am. Soc. Mass Spectrom. 11, 493–504.
Krutchinsky, A.N., Kalkum, M., and Chait, B.T. (2001). Automatic
identification of proteins with a MALDI-quadrupole ion trap mass
spectrometer. Anal. Chem. 73, 5066–5077.
Kundu, T.K., Palhan, V.B., Wang, Z.X., An, W.J., Cole, P.A., and
Roeder, R.G. (2000). Activator-dependent transcription from chro-
matin in vitro involving targeted histone acetylation by p300. Mol.
Cell 6, 551–561.
Lachner, M., O’Carroll, N., Rea, S., Mechtler, K., and Jenuwein, T.
(2001). Methylation of histone H3 lysine 9 creates a binding site for
HP1 proteins. Nature 410, 116–120.
Luger, K., Rechsteiner, T.J., and Richmond, T.J. (1999). Preparation
of nucleosome core particle from recombinant histones. Methods
Enzymol. 304, 3–19.
Megee, P.C., Morgan, B.A., Mittman, B.A., and Smith, M.M. (1990).
Genetic analysis of histone H4: essential role of lysines subject to
reversible acetylation. Science 247, 841–845.
Milne, T.A., Briggs, S.D., Brock, H.W., Martin, M.E., Gibbs, D., Allis,
C.D., and Hess, J.L. (2002). MLL targets SET domain methyl-
transferase activity to Hox gene promoters. Mol. Cell 10, 1107–
Milne, T.A., Hughes, C.M., Lloyd, R., Yang, Z., Rozenblatt-Rosen,
O., Dou, Y., Schnepp, R.W., Krankel, C., Livolsi, V.A., Gibbs, D., et al.
(2005). Menin and MLL cooperatively regulate expression of cyclin-
Chromatin Immunoprecipitation (ChIP) and Q-PCR Reactions
Chromatin immunoprecipitations were performed using the Chro-
matin Immunoprecipitation Assay Kit (Upstate, Lake Placid, New
York) and protocols recommended by the manufacturer. Real-time
PCR quantitation of ChIP was performed in triplicate using Taqman
probes and an ABI Prism 7700 (Applied Biosystems) and both the
relative quantification and percent input methods as outlined in
Milne et al. (2002). Real-time PCR quantitations of Hoxa9 and
Gapdh expression in MOF siRNA experiments were normalized
against input total RNA.
HeLa cells were transfected with siRNA duplexes (200 pmol; Dhar-
macon) using Oligofectamine (Invitrogen) according to the manu-
facturer’s instructions. One additional round of transfection was
performed under identical conditions 48 hr after the initial transfec-
tion. Cells were harvested 48 hr after the second transfection. The
sequences for MOF small interfering RNAs (siRNAs) used in the
assays are available upon request.
Supplemental Data include two figures, one table, and Supplemen-
tal Experimental Procedures and can be found with this article on-
line at http://www.cell.com/cgi/content/full/121/6/873/DC1/.
We are grateful to Drs. S. Korsmeyer, W. Herr, T. Tamura, and M.
Vidal for antibodies and J. Kim and Drs. Q. Yang and S. Malik in
the Roeder lab for technical advice. Y.D. is a fellow of the Irvington
Institute for Immunological Research; J.W. is a fellow of the Damon
Runyon Cancer Research Foundation, and A.F. is a fellow of the
Japan Society for the Promotion of Science. This work was sup-
ported by NIH grants (to J.L.H., B.T.C., A.J.T., and C.D.A), by a Leu-
kemia and Lymphoma Society of America SCOR grant (to J.L.H.),
and by funds from the Rockefeller University (to R.G.R.).
Received: January 26, 2005
Revised: March 25, 2005
Accepted: April 28, 2005
Published: June 16, 2005
Akhtar, A., and Becker, P.B. (2000). Activation of transcription
through histone H4 acetylation by MOF, an acetyltransferase
essential for dosage compensation in Drosophila. Mol. Cell 5,
An, W., and Roeder, R.G. (2004). Reconstitution and transcriptional
analysis of chromatin in vitro. Methods Enzymol. 377, 460–474.
An, W., Palhan, V.B., Karymov, M.A., Leuba, S.H., and Roeder, R.G.
(2002). Selective requirements for histone H3 and H4 N termini in
p300-dependent transcriptional activation from chromatin. Mol.
Cell 9, 811–821.
An, W., Kim, J., and Roeder, R.G. (2004). Ordered cooperative func-
tions of PRMT1, p300, and CARM1 in transcriptional activation by
p53. Cell 117, 735–748.
Beisel, C., Imhof, A., Greene, J., Kremmer, E., and Sauer, F. (2002).
Histone methylation by the Drosophila epigenetic transcriptional
regulator Ash1. Nature 419, 857–862.
Carrozza, M.J., Utley, R.T., Workman, J.L., and Cote, J. (2003). The
diverse functions of histone acetyltransferase complexes. Trends
Genet. 19, 321–329.
Corona, D.F., Clapier, C.R., Becker, P.B., and Tamkun, J.W. (2002).
Modulation of ISWI function by site-specific histone acetylation.
EMBO Rep. 3, 242–247.
Dignam, J.D., Lebovitz, R.M., and Roeder, R.G. (1983). Accurate
MLL1 and MOF Interactions Download full-text
dependent kinase inhibitors. Proc. Natl. Acad. Sci. USA 102, 749–
Morales, V., Straub, T., Neumann, M.F., Mengus, G., Akhtar, A., and
Becker, P.B. (2004). Functional integration of the histone acetyl-
transferase MOF into the dosage compensation complex. EMBO J.
Nakamura, T., Mori, T., Tada, S., Krajewski, W., Rozovskaia, T., Was-
sell, R., Dubois, G., Mazo, A., Croce, C.M., and Canaani, E. (2002).
ALL-1 is a histone methyltransferase that assembles a supercom-
plex of proteins involved in transcriptional regulation. Mol. Cell 10,
Neal, K.C., Pannuti, A., Smith, E.R., and Lucchesi, J.C. (2000). A
new human member of the MYST family of histone acetyl transfer-
ases with high sequence similarity to Drosophila MOF. Biochim.
Biophys. Acta 1490, 170–174.
Ng, H.H., Robert, F., Young, R.A., and Struhl, K. (2003). Targeted
recruitment of Set1 histone methylase by elongating Pol II provides
a localized mark and memory of recent transcriptional activity. Mol.
Cell 11, 709–719.
Owen-Hughes, T., Utley, R.T., Steger, D.J., West, J.M., John, S.,
Cote, J., Havas, K.M., and Workman, J.L. (1999). Analysis of nu-
cleosome disruption by ATP-driven chromatin remodeling com-
plexes. Methods Mol. Biol. 119, 319–331.
Petruk, S., Sedkov, Y., Smith, S., Tillib, S., Kraevski, V., Nakamura,
T., Canaani, E., Croce, C.M., and Mazo, A. (2001). Trithorax and
dCBP acting in a complex to maintain expression of a homeotic
gene. Science 294, 1331–1334.
Pray-Grant, M.G., Daniel, J.A., Schieltz, D., Yates, J.R., 3rd, and
Grant, P.A. (2005). Chd1 chromodomain links histone H3 methyla-
tion with SAGA- and SLIK-dependent acetylation. Nature 433,
Santos-Rosa, H., Schneider, R., Bannister, A.J., Sherriff, J., Bern-
stein, B.E., Emre, N.C., Schreiber, S.L., Mellor, J., and Kouzarides,
T. (2002). Active genes are tri-methylated at K4 of histone H3. Na-
ture 419, 407–411.
Santos-Rosa, H., Schneider, R., Bernstein, B.E., Karabetsou, N.,
Morillon, A., Weise, C., Schreiber, S.L., Mellor, J., and Kouzarides,
T. (2003). Methylation of histone H3 K4 mediates association of the
Isw1p ATPase with chromatin. Mol. Cell 12, 1325–1332.
Schneider, R., Bannister, A.J., Myers, F.A., Thorne, A.W., Crane-
Robinson, C., and Kouzarides, T. (2004). Histone H3 lysine 4 meth-
ylation patterns in higher eukaryotic genes. Nat. Cell Biol. 6, 73–77.
Sims, R.J., 3rd, Nishioka, K., and Reinberg, D. (2003). Histone lysine
methylation: a signature for chromatin function. Trends Genet. 19,
Smith, E.R., Pannuti, A., Gu, W.G., Steurnagel, A., Cook, R.G., Allis,
C.D., and Lucchesi, J.C. (2000). The Drosophila MSL complex acet-
ylates histone h4 at lysine 16, a chromatin modification linked to
dosage compensation. Mol. Cell. Biol. 20, 312–318.
Smith, E.R., Allis, C.D., and Lucchesi, J.C. (2001). Linking global
histone acetylation to the transcription enhancement of X-chromo-
somal genes in Drosophila males. J. Biol. Chem. 276, 31483–31486.
Smith, C.M., Gafken, P.R., Zhang, Z., Gottschling, D.E., Smith, J.B.,
and Smith, D.L. (2003). Mass spectrometric quantification of acety-
lation at specific lysines within the amino-terminal tail of histone
H4. Anal. Biochem. 316, 23–33.
Strahl, B.D., and Allis, C.D. (2000). The language of covalent histone
modifications. Nature 403, 41–45.
Strahl, B.D., Ohba, R., Cook, R.G., and Allis, C.D. (1999). Methyla-
tion of histone H3 at lysine 4 is highly conserved and correlates
with transcriptionally active nuclei in tetrahymena. Proc. Natl.
Acad. Sci. USA 96, 14967–14972.
Turner, B.M. (2002). Cellular memory and the histone code. Cell
Turner, B.M., Birley, A.J., and Lavender, J. (1992). Histone H4 iso-
forms acetylated at specific lysine residues define individual chro-
mosomes and chromatin domains in Drosophila polytene nuclei.
Cell 69, 375–384.
Vaquero, A., Loyola, A., and Reinberg, D. (2003). The constantly
changing face of chromatin. Sci. Aging Knowledge Environ. 2003,
Wysocka, J., Myers, M.P., Laherty, C.D., Eisenman, R.N., and Herr,
W. (2003). Human Sin3 deacetylase and trithorax-related Set1/Ash2
histone H3-K4 methyltransferase are tethered together selectively
by the cell-proliferation factor HCF-1. Genes Dev. 17, 896–911.
Wysocka, J., Swigut, T., Milne, T.A., Dou, Y., Zhang, X., Burlingame,
A.L., Roeder, R.G., Brivanlou, A.H., and Allis, C.D. (2005). WDR5
associates with histone H3 methylated at K4 and is essential for
H3 K4 methylation and vertebrate development. Cell 121, this is-
Yokoyama, A., Wang, Z., Wysocka, J., Sanyal, M., Aufiero, D.J., Ki-
tabayashi, I., Herr, W., and Cleary, M.L. (2004). Leukemia proto-on-
coprotein MLL forms a SET1-like histone methyltransferase com-
plex with menin to regulate Hox gene expression. Mol. Cell. Biol.
Yu, B.D., Hess, J.L., Horning, S.E., Brown, G.A., and Korsmeyer,
S.J. (1995). Altered Hox expression and segmental identity in Mll-
mutant mice. Nature 378, 505–508.
Zhang, K., Siino, J.S., Jones, P.R., Yau, P.M., and Bradbury, E.M.
(2004). A mass spectrometric “Western blot” to evaluate the corre-
lations between histone methylation and histone acetylation. Pro-
teomics 4, 3765–3775.
Zhang, K., Williams, K.E., Huang, L., Yau, P., Siino, J.S., Bradbury,
E.M., Jones, P.R., Minch, M.J., and Burlingame, A.L. (2002). Histone
acetylation and deacetylation: identification of acetylation and
methylation sites of HeLa histone H4 by mass spectrometry. Mol.
Cell. Proteomics 1, 500–508.