Current Biology, Vol. 12, 323–328, February 19, 2002, 2002 Elsevier Science Ltd. All rights reserved.PII S0960-9822(02)00681-4
Eco1 Is a Novel Acetyltransferase that Can
Acetylate Proteins Involved in Cohesion
ular domains (Figure 1A).
The Eco1 sequence was next compared with libraries
of known sequence domains and motifs. The sequence
segment 33–57 has a positive score (? 0.9; but statisti-
cally not significant E value ? 1) with the classical C2H2
Zn finger domain model of PFAM in the global domain
search mode [6, 9]. The Eco1 sequence differs from the
well-known PROSITE motif PS00028 “C-x(2,4)-C-x(3)-
in the loop spanning the cysteine and the histidine resi-
dues in the motif center (14 residues in loop 39–42 from
Eco1 instead of 12 in the classical motif).
We alsofound that thePFAM modelPF00583 “Acetyl-
transferase (GNAT) family” matches region 212–236 of
Eco1 in the local (domain fragment) search mode (score
12.1, E ? 0.009). Generally, such short fragment hits do
not implicate any relationship with the protein sequence
family represented by the HMM model, if they are not
substantiated by additional, nonstatistical arguments
(for example, structural considerations or biological
context) . In this case, the sequence piece is essen-
tially identical to the part of the acetyltransferase model
describing the motif A, the acetyl-coenzyme A binding
region of the GNAT superfamily, and critical residues
for ligand binding, such as R222, G225, and D232, are
well conserved [11, 12]. In addition to this functional
argument, we found highly similar secondary structure
predictions among Eco1 proteins (see below, Figure 1).
Third, we collected the family of sequences homolo-
gous to Eco1 with Blast/Psi-Blast [13, 14] and HMM 
searches and found representatives for a number of
eukaryotes, including all heavily sequenced species
among fungi, animals, and plants (Figure 1B and our
ECO1/). Both in the human and the mouse EST data-
bases, two distinct Eco1 variants with different tissue-
specific expression (as deduced from EST annotations)
All homologs of Eco1 in other species have also been
subjected to the same three-step analysis procedure.
The C2H2Zn finger-like motif of yeast Eco1 was found
in all apparently complete sequences, except for the
protozoan proteins from Trypanosoma brucei (embl
ceptual translation from genomic clone). Remarkably,
a Drosophila acetyltransferase MOF that belongs to a
domain is demonstrated to use the Zn finger to contact
the nucleosome and the histone H4 N-terminal tail. Mu-
tations in the Zn finger greatly diminish MOF activity in
Indications for an acetyltransferase domain, pre-
viously referred to as Eco1 domain , were also de-
tected in all homologs. The amino acid sequence region
945–1008 of Eco1 from Drosophila melanogaster was
the best hit in a domain fragment search with the PFAM
model PF00583 (score, 17.5; E value, 6.4e-05). In the
Dmitri Ivanov,2Alexander Schleiffer,2
Frank Eisenhaber, Karl Mechtler,
Christian H. Haering, and Kim Nasmyth1
Research Institute of Molecular Pathology
Dr. Bohr-Gasse 7
Cohesion between sister chromatids is established
during S phase and maintained through G2 phase until
it is resolved in anaphase (for review, see [1–3]). In
Saccharomyces cerevisiae, a complex consisting of
Scc1, Smc1, Smc3, and Scc3 proteins, called “co-
hesin,” mediates the connection between sister chro-
Eco1 is required for establishment of sister chromatid
cohesion during S phase but not for its further mainte-
complex onto DNA [6, 7]. We address the molecular
functions of Eco1 with sensitive sequence analytic
techniques, including hidden Markov model domain
fragment searches. We found a two-domain architec-
ture with an N-terminal C2H2Zn finger-like domain and
an ?150 residue C-terminal domain with an apparent
acetyl coenzyme A binding motif (http://mendel.imp.
univie.ac.at/SEQUENCES/ECO1/). Biochemical tests
confirm that Eco1 has the acetyltransferase activity in
vitro. In vitro Eco1 acetylates itself and components
of the cohesin complex but not histones. Thus, the
establishment of cohesion between sister chromatids
appears to be regulated, directly or indirectly, by a
Results and Discussion
Sequence Analysis of Eco1 Reveals
a Two-Domain Structure
The Eco1 sequence is evolutionary conserved through-
out eukaryotes,and ithas significantsimilarity tohuman
and mouse ESTs . However, no convincing sugges-
tion of molecular function has yet been identified [7, 8].
In an attempt to extend previous results , the Eco1
protein sequence of Saccharomyces cerevisiae was
subjected to a three-step sequence analysis procedure
with bioinformatic methods. We first studied the lexical
structure of its sequence, looking for compositionally
biased regions and other typically nonglobular protein
segments. We identified two polar regions of low com-
plexity: one at the N terminus (residues 1–13) rich in
positive charges (6 of 13) and another between se-
quence positions 95–107 with many prolines and ser-
2These authors contributed equally to this work.
Figure 1. Eco1 Is a Member of the GCN5-
Related N-Acetyltransferase Superfamily
(A) Domain architecture of Eco1 from Sac-
charomyces cerevisiae. For details, see text.
(B) Eco1 is a member of the GCN5-related
N-acetyltransferase superfamily. The joint
multiple alignment of protein sequences of
members of the GNAT acetyltransferase su-
perfamily (structurally aligned, ) and se-
lected Eco1 sequences was obtained by
manual adjustments based on PF00583 do-
main fragment hits, secondary structure pre-
dictions, and physicochemical similarity of
amino acid types. The GenBank accession
numbers are AANAT_Oa, 6980559; Hat1_Sc,
5542194; AAC-3?_Sm, 3745773; AAC-6’_Em,
5542089; Eco1_Sc, NP_011218.1; Eco1_Sp,
NT_025025.5 (the latter two are conceptual
translations from genomic clones). The sec-
ondary structure elements of AANAT are
shown above (the numbering is derived from
). The framed residues correspond to the
regions designated as motifs D, A, and B by
. Residues in the S. cerevisiae Eco1 that
abolish enzymatic activity when mutated are
marked by an asterisk.
(C) Three-dimensional structure of GCN5 his-
tone N-acetyltransferase from Tetrahymena
thermophilum (1QSN) . The coloring and
labeling of the structural elements is ac-
cording to the sequence alignment in (A).
log is shown in yellow.
case of the latter, even the complete PFAM domain
nificant E value) for a segment comprising 98 residues.
We found that the N terminus of the S. pombe Eco1
homolog is extended by more than 600 amino acids and
the D. melanogaster homolog by 800 amino acids. In
the first case, the N-terminal extension is the known
fusion with polymerase ? [6, 17] and contains a UMUC
domain known to be involved in DNA replication and
UV protection ; in the fly protein case, no function
could be guessed for this extension, which could also
represent a genome assembly artifact.
Startingfrom thePF00583 modelhitsinEco1homologs
and relying on the highly similar secondary structure pre-
ing the Eco1 sequences and acetyltransferases with a
known 3D structure [12, 19, 20] was constructed and
plausibly extended both in N- and C-terminal directions.
The core alignment region comprises the sequence mo-
tifs D, A, and B; the ? strands 2 through 5; and the ?
helices 3 and 4 of the GNAT fold (Figures 1B and 1C)
[11, 12]. Apparently, Eco1 homologs lack the ? helix 2
of the histone acetyltransferase structure. Considering
the low level of sequence conservation with GNAT su-
To conclude, Eco1 protein sequence analysis results
with many positive charges (K/R) and a C2H2Zn finger-
like domain (residues 33–57, C2H2), possibly with a nu-
cleic acid or protein binding function; (2) a proline- and
serine-rich low complexity region (residues 95–107,
P/S), the probable interdomain linker; and (3) a C-ter-
minal domain with suggested acetyltransferase activity
(residues 111–266, ACT) (Figure 1A).
In order to test whether Eco1 has an enzymatic activity in
vitro, we expressed the yeast protein as a GST-fusion in
E. coli. When this recombinant protein was incubated
in the presence of
label to a protein was readily detectable (Figure 2A,
lanes 1–3). To confirm that this was not due to a very
tight binding of Eco1 to acetyl coenzyme A, Eco1 was
incubated with a cold substrate and then analyzed by
Western blot with a commercial anti-acetyl-lysine anti-
14C-acetyl coenzyme A, transfer of
prisingly, Eco1 purified from bacteria was found to be
likely due to autoacetylation occurring inside the cell in
the process of expression.
Transfer ofan acetyl groupfrom acetyl coenzymeA to
a primary amine in aqueous solution is an energetically
zymatic background acetylation . To rule out this
possibility, we decided to confirm that the observed in
vitro modification of Eco1 is dependent on its intact
structure. We mutated a number of amino acids con-
served among Eco1 homologs from different organisms
(indicated by asterisks in Figure 1B) and tested the mu-
tant forms of enzyme for activity in vitro. The mutation
leading to a temperature-sensitive allele of eco1 was
identified as glycine 211 to aspartic acid substitution
. In our model of Eco1 fold, this position corresponds
to the fourth residue in the ?4 helix in the motif A, the
longest and most highly conserved motif among the
members of GNAT superfamily of acetyltransferases.
Strand ?4 and the following helix ?3 form the core of
the acetyl coenzyme A binding site . Remarkably, a
mutation in Eco1 homolog of S. pombe, Eso1, that was
responsible for temperature sensitivity was an identical
glycine to aspartic acid substitution at position 799 cor-
respondingin thealignment toposition 211of Eco1.
Another mutation that we used in Eco1 was a glycine
to aspartic acid substitution at amino acid 225. This
position corresponds to glycine 691 in a Drosophila his-
tone acetyltransferase MOF. Substitution of this glycine
by a glutamic acid resulted in the absence of MOF activ-
ity in vivo . Similarly, mutation of corresponding gly-
ferase to aspartic acid resulted in loss of activity .
Two more mutations that were generated were arginine
222/lysine 223 to glycine/glycine and aspartic acid 232
to glycine. In the predicted secondary structure, all mu-
tated amino acids are located within motif A, either in
the helix ?3 or immediately preceding it. When the wild-
type and mutant forms of Eco1 were tested for self-
acetylation in an in vitro radioactive assay (Figure 2A),
only a wild-type enzyme was labeled efficiently (lanes
(lanes 4–15) that was weakly detected after the longer
exposure times. This result demonstrates that an Eco1
intact structure is required for the transfer of an acetyl
group and that this is a bona fide enzymatic reaction.
A majority of confirmed protein acetyltransferases are
tested whether Eco1 could use histones as a substrate.
Bovine histone H3 alone (Figure 2C, lanes 6–8) or a
mixture of all four core histones (lanes 3–5) were incu-
bated with14C-acetyl coenzyme A either alone or in the
presence of Eco1 or recombinant PCAF, a known his-
tone acetyltransferase, expressed in bacteria that was
used as a positive control. As expected, a measurable
level of apparently nonenzymatic acetylation of histone
H3 occurred in the absence of any added enzyme (lane
6). However, the addition of PCAF resulted in a greatly
increased acetylation of its primary targets, histones H3
and H4  (lanes 5 and 8), while the addition of Eco1
had no detectable effect (lanes 4 and 7).
Next, we assayed whether Eco1 can modify known
Figure 2. Eco1 Self-Acetylates but Does Not Acetylate Histones In
(A)Eco1 autoacetylatesinvitro. Wild-typeandmutant Eco1proteins
were expressed as GST fusions in bacteria. Increasing amounts of
Eco1 were incubated in the presence of
and separated on SDS-PAGE. Coomassie stain (upper panel) and
phosphorimager scan(lower panel)of thesame gelare shown.Eco1
band on Coomassie-stained gel is indicated with an arrow and a
nonspecific band with an asterisk. Although not clearly visible on
ration but is less abundant compared to the Eco1 protein.
(B) Autoacetylationof Eco1 isdetected by anacetyl-lysine antibody.
Wild-type Eco1 protein was incubated either alone (lane 1) or in the
presence of the increasing concentrations of cold acetyl coenzyme
A (lanes 2 through 5). Samples were separated on SDS-PAGE, and
Western blot analysis with an anti-acetyl-lysine antibody (Ac-K-103,
NEB)was performed.Thebackground acetylationfrom thebacterial
expression system (see text) is probably responsible for a weak
immunoreactivity with anti-acetyl-lysine antibody in the absence of
acetyl conenzyme A.
(C) Eco1 does not acetylate histones. Eco1 or PCAF were incubated
bovine core histones (lanes 3 through 5) or histone H3 alone (lanes 6
through 8). Phosphorimager scan of an SDS-PAGE gel is presented.
14C-acetyl coenzyme A
14C-acetyl coenzyme A either alone or in the presence of all
body (NEB) (Figure 2B). This antibody recognizes an
epitope comprised of an acetyl group covalently cou-
pled to the amino group of lysine side chain. Eco1 incu-
of the substrate, the intensity of the band detected by
the Western blot greatly increased (lanes 2–5). This indi-
cated that the reaction is accompanied by a covalent
transfer of the acetyl group to a lysine residue of the
protein. Eco1 was further analyzed by mass spectrome-
try to identify the acetylated amino acid residues. Sur-
Figure 3. Eco1 Acetylates Cohesin Subunits In Vitro
(A) Scc1 is acetylated by Eco1 in vitro. (Left and middle panels) In vitro acetylation assay was performed with cold acetyl coenzyme A and
Eco1, full-length Scc1 expressed in baculovirus, or both proteins together. Same Western blot was analyzed with anti-acetyl-lysine (left panel)
or anti-His5antibody (middle panel). (Right panel) Same assay was performed with radioactive acetyl coenzyme A. Phosphorimager scan of
SDS-PAGE gel is shown.
(B) Lysine 210 of Scc1 is modified by Eco1. The middle fragment of Scc1 was expressed as Gst fusion in bacteria. Either wild-type or proteins
with mutated lysines were used as substrates in an in vitro acetylation assay with Gst-Eco1. Upper and middle panels show Coomassie stain
and phosphorimager scan of the same gel. The lower panel is the Western blot probed with an anti-acetyl-lysine antibody.
(C) Scc3 and Pds5 are also acetylated by Eco1 in vitro. In vitro acetylation assays were performed with Gst-Eco1 and Scc3 and Pds5 proteins
expressed in baculovirus. Western blot with anti-acetyl-lysine antibody and a phosphorimager scan of a radioactive assay are shown.
components of the cohesin complex. First, yeast Scc1
protein was expressed in baculovirus system and incu-
bated with cold acetyl coenzyme A and Eco1 followed
by a Western blot analysis with an anti-acetyl-lysine
antibody. Incubation of Scc1 in the presence but not in
the absence of Eco1 resulted in a detectable degree of
acetylation (Figure 3A, left panel). The identity of the
band was confirmed by probing the same Western blot
with an anti-His5antibody recognizing a poly-histidine
tag on the recombinant Scc1 (Figure 3A, middle panel).
Scc1 was also labeled in a radioactive assay, although
weakly (Figure 3A, right panel). In addition, we observed
Eco1-dependent acetylation of the S. pombe Scc1 ho-
molog Rad 21 (data not shown).
S. cerevisiae Scc1 contains 31 lysines. Based on the
available prediction of its domain structure (our unpub-
lished data), Scc1 can be divided into three regions. The
N-terminal and the C-terminal regions are likely to fold
into defined domains with globular structure, and the
middle portion encompassing the separase cleavage
sites presumably constitutes an unstructured linker re-
gion. To map the acetylation site, we generated three
fragments of Scc1 corresponding respectively to amino
as GST fusions in E. coli. Only the middle portion of
Scc1 was acetylated in the in vitro assay. This region
contained only six lysines. They were subsequently mu-
tated to arginines, and the resulting mutants were as-
sayed for acetylation by Eco1 (Figure 3B). Mutating ly-
sine 210 to arginine resulted in loss of acetylation.
Mutating other lysines had no effect on the ability of the
middle portion of Scc1 to be modified by Eco1 either in
radioactive assay or by Western blot analysis with anti-
acetyl-lysine antibody. In addition, acetylation of lysine
210was confirmedbymassspectrometry analysis(data
plex can be acetylated by Eco1 in vitro. Scc3 and Pds5
proteins were expressed in baculovirus and tested as
substrates in the in vitro assay. All of them were acet-
ylated (Figure 3C), with Scc3 demonstrating the highest
degree of modification. The acetylation sites in Scc3
were further mapped by deletion analysis. Fragments
containing amino acids 1–174, 175–350, and 961–1150
were acetylated, while the middle portion of Scc3 was
not. Mass spectrometry analysis of the acetylated frag-
ments identified the modified amino acids as lysines 13,
36, 78, 106, 224, 1071, and 1086.
The analysis of the in vivo role of a new enzymatic
activity is clearly not a trivial task. For example, it took
3 years to demonstrate the galactosyltransferase activ-
ity of Fringe [25–27] after its prediction based on se-
quence considerations. To test whether any of the pro-
teins involved in cohesion are acetylated in vivo, we
performed immunoprecipitation of the myc-tagged can-
didate proteins from yeast under denaturing conditions
followed by Western blot analysis with commercially
ylated peptide corresponding to the region of Scc1
spanning lysine 210. In the case of Scc1 and Scc3, we
matched thequantity of theimmunoprecipitated protein
assay. Although acetylation of the recombinant protein
was readily detected, there was no acetylation of the
immunoprecipitated protein. A weak immunoreactivity
that was observed with some of the proteins after long
exposure times or using high antibody concentration
was likely nonspecific, since there was no difference
between protein precipitated from wild-type strain and
eco1 mutant strain grown at restrictive temperature. In
spite of our efforts, we have been unable to detect acet-
implicated in cohesion, PCNA and Ctf18, in vivo. To
further test whether acetylation of Scc1 has an in vivo
function, we constructed a yeast strain containing a
wild-type copy of Scc1 driven by a galactose-inducible
promoter and a mutant copy with lysine 210 substituted
for arginine under a normal Scc1 promoter. This strain
was viable when grown on medium with glucose as a
single carbon source and showed no obvious growth
or morphological defects (data not shown). Considering
that Eco1 function is essential in S. cerevisiae, this ob-
servation implies that if Scc1 acetylation is indeed oc-
curring in vivo it is unlikely to be the only essential Eco1
substrate. This leaves the most interesting question of
It still remains possible that one or more of cohesin
subunits are acetylated in the process of establishment
of cohesion, but this modification is transient, and the
chromatin. Alternatively, the specificity of the available
anti-acetyl-lysine antibodies may not allow detection of
acetylation of these particular proteins. Radioactive in
vivo labeling has been employed in the past for the
detection of acetylation of more transient nature, as is
the case with p53 . However, it also required overex-
pression of the protein in cells. Attractive candidates for
acetylation by Eco1 are DNA polymerases themselves,
such as Trf4 and DNA polymerase ?. The first was dem-
onstrated to be specifically required for the establish-
ment and perhaps also for maintenance of cohesion
. The second was shown to interact genetically with
Eso1 in S. pombe . Other possible targets might be
components of DNA replication machinery or chromatin
acetyltransferase is active in an in vitro assay toward a
variety of cohesion-related substrates. However, it has
no detectable activity toward histones. While this raises
the question of specificity of the modification observed
in vitro, it suggests that an in vivo target of the new
enzyme will most likely be a non-histone protein. The
fact that the G211D mutation that causes temperature
sensitivity in vivo (1) is located in the most conserved
region that is characteristic for known acetyltransfer-
ases and responsible for coenzyme A binding and (2)
triggers a dramatic reduction of enzymatic activity in
vitro strongly suggests that Eco1 functions as an acetyl-
transferase in vivo as well and the disruption of this
activity causes the mutant phenotype.
tid cohesion is regulated, directly or indirectly, by an
acetyltransferase. This observation opens an exciting
to lead to a discovery of a new enzymatic pathway that
is essential for cellular proliferation regulation.
This research was undertaken with the financial support of Boeh-
ringer Ingelheim International, the Austrian Industrial Research Pro-
motion Fund (FFF), and the Austrian Science Fund (FWF).
Received: November 20, 2001
Revised: December 18, 2001
Accepted: December 19, 2001
Published: February 19, 2002
1. Hirano, T. (2000). Chromosome cohesion, condensation, and
separation. Annu. Rev. Biochem. 69, 115–144.
2. Koshland, D.E., and Guacci, V. (2000). Sister chromatid cohe-
sion: the beginning of a long and beautiful relationship. Curr.
Opin. Cell Biol. 12, 297–301.
3. Nasmyth, K., Peters, J.M., and Uhlmann, F. (2001). Splitting
the chromosome: cutting the ties that bind sister chromatids.
Novartis Found. Symp. 237, 113–133.
4. Guacci, V., Koshland, D., and Strunnikov, A. (1997). A direct link
between sister chromatid cohesion and chromosome conden-
sation revealed through the analysis of MCD1 in S. cerevisiae.
Cell 91, 47–57.
mosomal proteins that prevent premature separation of sister
chromatids. Cell 91, 35–45.
6. Toth, A., Ciosk, R., Uhlmann, F., Galova, M., Schleiffer, A., and
Nasmyth, K. (1999). Yeast cohesin complex requires a con-
served protein, Eco1p(Ctf7), to establish cohesion between sis-
ter chromatids during DNA replication. Genes Dev. 13, 320–333.
Ctf7pisessential forsisterchromatidcohesion andlinksmitotic
Dev. 13, 307–319.
8. Jones, S., and Sgouros, J. (2001). The cohesin complex: se-
quence homologies, interaction networks and shared motifs.
Genome Biol. 2, 3.
9. Bateman, A., Birney, E., Durbin, R., Eddy, S.R., Howe, K.L., and
Sonnhammer, E.L. (2000). The Pfam protein families database.
Nucleic Acids Res. 28, 263–266.
10. Sander, C., and Schneider, R. (1991). Database of homology-
derived protein structures and the structural meaning of se-
quence alignment. Proteins 9, 56–68.
11. Neuwald, A.F., and Landsman, D. (1997). GCN5-related histone
N-acetyltransferases belong to a diverse superfamily that in-
cludes the yeast SPT10 protein. Trends Biochem. Sci. 22,
We report a novel acetyltransferase activity that was
in a protein with only a marginal degree of similarity
to the known GNAT superfamily acetyltransferases that
edge, this is the first successful case of nontrivial func-
tion assignment for an uncharacterized protein with a
hidden Markov model domain fragment search. Eco1
Current Biology Download full-text
12. Dyda, F., Klein, D.C., and Hickman, A.B. (2000). GCN5-related
N-acetyltransferases: a structural overview. Annu. Rev. Bio-
phys. Biomol. Struct. 29, 81–103.
13. Altschul, S.F., and Koonin, E.V. (1998). Iterated profile searches
with PSI-BLAST—a tool for discovery in protein databases.
Trends Biochem. Sci. 23, 444–447.
14. Altschul, S.F., Madden, T.L., Schaffer, A.A., Zhang, J., Zhang,
Z., Miller, W., and Lipman, D.J. (1997). Gapped BLAST and PSI-
BLAST: a new generation of protein database search programs.
Nucleic Acids Res. 25, 3389–3402.
16. Akhtar, A., and Becker, P.B. (2001). The histone H4 acetyltrans-
ferase MOF uses a C2HC zinc finger for substrate recognition.
EMBO Rep. 2, 113–118.
17. Tanaka, K., Yonekawa, T., Kawasaki, Y., Kai, M., Furuya, K.,
Iwasaki, M., Murakami, H., Yanagida, M., and Okayama, H.
(2000). Fission yeast Eso1p is required for establishing sister
chromatid cohesion during S phase. Mol. Cell. Biol. 20, 3459–
18. Sutton, M.D., Smith, B.T., Godoy, V.G., and Walker, G.C. (2000).
The SOS response: recent insights into umuDC-dependent mu-
tagenesis and DNA damage tolerance. Annu. Rev. Genet. 34,
19. Dutnall, R.N., Tafrov, S.T., Sternglanz, R., and Ramakrishnan,
V. (1998). Structure of the yeast histone acetyltransferase Hat1:
insights intosubstrate specificityand implications forthe Gcn5-
related N- acetyltransferase superfamily. Cold Spring Harb.
Symp. Quant. Biol. 63, 501–507.
20. Rojas, J.R., Trievel, R.C., Zhou, J., Mo, Y., Li, X., Berger, S.L.,
Allis, C.D., and Marmorstein, R. (1999). Structure of Tetrahy-
mena GCN5 bound to coenzyme A and a histone H3 peptide.
Nature 401, 93–98.
21. Tanner, K.G., Trievel, R.C., Kuo, M.H., Howard, R.M., Berger,
S.L., Allis, C.D., Marmorstein, R., and Denu, J.M. (1999). Cata-
lytic mechanism and function of invariant glutamic acid 173
from the histone acetyltransferase GCN5 transcriptional coacti-
vator. J. Biol. Chem. 274, 18157–18160.
22. Hilfiker, A., Hilfiker-Kleiner, D., Pannuti, A., and Lucchesi, J.C.
(1997). Mof, a putative acetyl transferase gene related to the
Tip60 and MOZ human genes and to the SAS genes of yeast,
is required for dosage compensation in Drosophila. EMBO J.
23. Lu, L., Berkey, K.A., and Casero, R.A., Jr. (1996). RGFGIGS is
and activity of human spermidine/spermine N1acetyltransfer-
ase. J. Biol. Chem. 271, 18920–18924.
24. Schiltz, R.L., Mizzen, C.A., Vassilev, A., Cook, R.G., Allis, C.D.,
and Nakatani, Y. (1999). Overlapping but distinct patterns of
histone acetylation by the human coactivators p300 and PCAF
within nucleosomal substrates. J. Biol. Chem. 274, 1189–1192.
25. Yuan, Y.P., Schultz, J., Mlodzik, M., and Bork, P. (1997). Se-
creted fringe-like signaling molecules may be glycosyltransfer-
ases. Cell 88, 9–11.
syltransferase activity of Fringe modulates Notch-Delta interac-
tions. Nature 406, 411–415.
27. Moloney, D.J., Panin, V.M., Johnston, S.H., Chen, J., Shao, L.,
Wilson, R., Wang, Y., Stanley, P., Irvine, K.D., Haltiwanger, R.S.,
et al. (2000).Fringe is a glycosyltransferasethat modifies Notch.
Nature 406, 369–375.
28. Gu, W., and Roeder, R.G. (1997). Activation of p53 sequence-
specific DNA binding by acetylation of the p53 C-terminal do-
main. Cell 90, 595–606.
29. Wang, Z., Castano, I.B., De Las Penas, A., Adams, C., and
Christman, M.F. (2000). Pol kappa: A DNA polymerase required
for sister chromatid cohesion. Science 289, 774–779.