Emery-Dreifuss muscular dystrophy (EDMD) is characterized
by a triad of symptoms: progressive muscle weakening,
contractures of the Achilles and other tendons, and potentially
life-threatening cardiac conduction defects (Emery, 1989).
EDMD is inherited through mutations in either of two genes,
STA (Bione et al., 1994) or LMNA (Bonne et al., 1999), which
encode nuclear lamina proteins named emerin and A-type
lamins, respectively (Cohen et al., 2001). Mutations in LMNA
can also give rise to other diseases (Bonne et al., 2000), including
dilated cardiomyopathy and lipodystrophy. The mechanisms of
these diseases, collectively termed laminopathies, are not
understood (Wilson et al., 2001; Morris, 2001).
Human emerin is a 254-residue integral protein of the
nuclear inner membrane (Manilal et al., 1996; Nagano et al.,
1996; Yorifuji et al., 1997). Emerin belongs to a family of
nuclear proteins defined by a ~40-residue motif termed the
LEM-domain (Lin et al., 2000). The LEM-domain family is
growing and includes MAN1 (Lin et al., 2000), lamina
associated polypeptide-2 (LAP2) (Foisner and Gerace, 1993),
otefin (Goldberg et al., 1998; Wolff et al., 2001) and Lem-3
(Lin et al., 2000; Lee et al., 2000). Emerin and the β-isoform
of LAP2 have a second region of high homology at their
transmembrane domains, and are similar throughout their
lengths. Both emerin and LAP2β interact with lamins. LAP2β
interacts specifically with lamin B1 (Foisner and Gerace,
1993), whereas emerin interacts with both A- and B-type
lamins (Fairley et al., 1999; Clements et al., 2000). In LMNA-
knockout mice, emerin becomes localized to both the nuclear
envelope and ER, suggesting that A-type lamins contribute to
(but are not essential for) the nuclear localization of emerin
(Sullivan et al., 1999). Localization at the inner nuclear
membrane appears to be important for emerin’s function, since
a mutation that prevents emerin from reaching the inner
membrane causes disease (Fairley et al., 1999).
The homology between LAP2β and emerin suggested to us
that these proteins might have related functions. In addition
to binding lamin B, LAP2β also interacts with chromatin in
vitro (Foisner and Gerace, 1993). A novel binding partner
for LAP2β on chromatin was identified in a yeast two-
hybrid screen (Furukawa, 1999); this partner, barrier-to-
autointegration factor (BAF), is an essential, highly conserved
DNA-bridging protein of unknown function (Lee and Craigie,
1998; Chen and Engelman, 1998; Zheng et al., 2000). The
LEM-domain is essential for LAP2β to bind BAF (Furukawa,
1999; Shumaker et al., 2001) and BAF-DNA complexes
(Shumaker et al., 2001). Because emerin has a LEM domain,
we tested the hypothesis that emerin binds BAF. Our results
for wildtype emerin and a collection of site-directed emerin
mutants strongly support this model, and define at least two
proposed functional domains within emerin.
Loss of emerin, a lamin-binding nuclear membrane
protein, causes Emery-Dreifuss muscular dystrophy. We
analyzed 13 site-directed mutations, and four disease-
causing mutations that do not disrupt emerin stability or
localization. We show that emerin binds directly to barrier-
to-autointegration factor (BAF), a DNA-bridging protein,
and that this binding to BAF requires conserved residues
in the LEM-motif of emerin. Emerin has two distinct
functional domains: the LEM-domain at the N-terminus,
which mediates binding to BAF, and a second functional
domain in the central region, which mediates binding to
lamin A. Disease mutation ∆95-99 mapped to the lamin-
binding domain and disrupted lamin A binding in vitro.
Two other disease-linked residues, Ser54 and Pro183,
mapped outside the BAF and lamin-binding domains,
suggesting that emerin may have additional functional
domains relevant to disease. The disease-linked emerin
proteins all remained active for binding to BAF, both in
vitro and in vivo, suggesting that disease can result from
the loss of specific molecular interactions between emerin
and either lamin A or putative novel partner(s). The
demonstration that emerin binds directly to BAF, coupled
to similar results for LAP2, provides proof in principle that
all LEM-domain nuclear proteins can interact with BAF,
with interesting implications for chromatin attachment to
the nuclear envelope.
Key words: Barrier to autointegration factor, Emery-Dreifuss
muscular dystrophy, lamin A, lamin-associated polypeptide 2, LEM-
domain, nuclear envelope, nuclear lamina, MAN1.
Distinct functional domains in emerin bind lamin A
and DNA-bridging protein BAF
Kenneth K. Lee1, Tokuko Haraguchi2, Richard S. Lee1, Takako Koujin2, Yasushi Hiraoka2
and Katherine L. Wilson1,*
1Department of Cell Biology, Johns Hopkins University School of Medicine, 725 N. Wolfe St., Baltimore, MD 21205, USA
2CREST Research Project of the Japan Science and Technology Corporation, Kansai Advanced Research Center, Communications Research
Laboratory, 588-2 Iwaoka, Iwaoka-cho, Nishi-ku, Kobe 651-2492 Japan
*Author for correspondence (e-mail: firstname.lastname@example.org)
Accepted 15 October 2001
Journal of Cell Science 114, 4567-4573 (2001) © The Company of Biologists Ltd
MATERIALS AND METHODS
Antiserum production and immunoblots
Polyclonal antibodies against recombinant human emerin were raised
in rabbit serum 2999, using untagged wildtype emerin residues 1-222
as antigen. Immunizations and serum production were done by
Covance Research Products (Denver PA). For immunoblots,
recombinant emerin proteins in bacterial lysates were resolved by
electrophoresis in 10% SDS-PAGE gels, transferred to nitrocellulose
(Schleicher and Schuell), blocked in PBS/0.1% Tween-20 (PBST)
containing 5% nonfat dry milk, and probed with serum 2999 (1:1000
dilution). Bound antibodies were detected using HRP-conjugated goat
anti-rabbit antibodies (1:50,000 dilution; Pierce) and enhanced
chemiluminescence (Amersham/Pharmacia Biotech).
An emerin cDNA was generated by PCR by E. Abrams and J.
Beneken from a human heart cDNA library obtained from R. Reed
(Johns Hopkins School of Medicine). The starting point for site-
directed mutagenesis was a cDNA encoding wildtype human emerin
residues 1-222, subcloned into the pET11c vector (Novagen). All
mutations were made using the QuickChange site-directed
mutagenesis kit (Stratagene), following the manufacturer’s
instructions, and verified by full-length double-stranded DNA
sequence analysis (data not shown). GFP-emerin constructs were
made as described (Haraguchi et al., 2001).
Emerin expression and blot overlay assays
Each emerin construct was transformed into E. coli strain BL21
(DE3). Transformed cells containing each plasmid were grown to an
OD600 of 0.6, and emerin expression was induced by 0.4 mM IPTG
for four hours. Cells were pelleted for 5 minutes at 14,000 g, and
resuspended in 2× SDS sample buffer. Proteins from unfractionated
bacterial lysates were separated on 10% SDS-PAGE gels, transferred
to nitrocellulose membranes (Schleicher and Schuell), and blocked for
1 hour in PBST containing 5% nonfat dry milk. Blots were then
washed twice in BRB (Blot Rinse Buffer; 10 mM Tris-HCl, pH 7.4,
150 mM NaCl, 1 mM EDTA, 0.1% Tween-20) for 5 minutes at 22-
24°C, and incubated with 20 µCi of 35S-cysteine/methionine labeled
probe protein (either BAF or lamin A; see below) diluted 1:200 into
BRB containing 0.1% fetal calf serum (final volume, 10 ml). The
lamin A construct in vector pET7a was a kind gift from Robert Moir
and Robert Goldman (Northwestern University, Chicago). Blots were
incubated overnight with 35S-labeled in vitro-transcribed/translated
probe protein at 4°C, washed twice in BRB, dried and exposed to
Hyperfilm MP (Amersham/Pharmacia Biotech). Emerin mutant
proteins m76 and m141 consistently migrated more slowly than other
recombinant emerins on SDS-PAGE.
Synthesis of 35S-Cys/Met labeled proteins and
We used the T7 promotors on expression vectors pET11c (for emerin
and emerin mutants), pET7a (for lamin A) and pET15b (for BAF) to
drive the expression of 35S-cysteine/methionine-labeled emerin, lamin
A and BAF proteins using the TNT Quick Coupled Transcription/
Translation System (Promega Corp., Madison WI), according to
the manufacturer’s protocol. Proteins were transcribed/translated
individually for 90 minutes at 30°C. For use as probes in blot overlay
experiments, each protein was diluted 1:200 into BRB/0.1% FCS and
used as described above. For immunoprecipitation experiments,
labeled proteins (10 µl each from a 50 µl TNT reaction) were
incubated (individually, or mixed as indicated) for 30 minutes at 22-
25°C to allow binding. We then added 300 µl of immunoprecipitation
(IP) buffer (20 mM Hepes pH 7.9, 150 mM NaCl, 10 mM EDTA, 2
mM EGTA, 0.1% NP-40, 10% glycerol, 1 mM DTT, 1 mM PMSF
and 20 ug/ml leupeptin) to each sample. To immunoprecipitate 35S-
labeled emerin, 4 µl of serum 2999 (immune or pre-immune) was
added to each reaction and incubated one hour on ice. BAF was
immunoprecipitated using rabbit serum 3000. We then added 50 µl of
washed protein A Sepharose beads (Amersham/Pharmacia Biotech),
incubated overnight at 4°C, centrifuged at 5000 g for 5 minutes to
pellet the beads, and washed the pellets five times with ice-cold IP
buffer. Bound proteins were removed from beads by boiling in 40 µl
2× SDS sample buffer, subjected to 17% SDS-PAGE, dried and
exposed to Hyperfilm (Amersham/Pharmacia Biotech).
GFP-emerin plasmid construction
GFP-emerin was a gift of Yuichi Tsuchiya and Kiichi Arahata. To
make a GFP fusion to emerin-m24, emerin-S54F and emerin-∆95-99
that included the transmembrane domain, the coding region of
pET11c-emerin-m24, pET11c-emerin-S54F and pET11c-emerin-
∆95-99 was first PCR-amplified using primers 5′-CGTCC-
TGGCGATCCTGGCCCAG-3′. Secondly, the PCR product was
digested with BspEI and BamHI, and inserted in the pEGFP-C1 vector
at the BspEI and BamHI sites. Finally, this construct was digested with
SacI and BamHI, and ligated with the SalI/BamHI fragment from full-
length GFP-emerin plasmids that include the transmembrane domain.
To make a GFP-fusion to emerin-P183T and emerin-P183H that
included the transmembrane domain, the coding region of GFP-
emerin was PCR-amplified using the following primers; 5′-
3′ for emerin-P183T, and 5′-CGGAGCTCCCTGGACCTGTCC-
TATTATCATACTTCCTCCTC-3′ and 5′-GGATCCGGTGGATCCC-
GGGCCCGCGGTACCGTAGAC-3′ for emerin-P183H. The PCR
product was digested with SacI and BamHI, and ligated with the
SacI/BamHI fragment from full-length GFP-emerin plasmids. The
DNA sequence of all fusion plasmids were confirmed using an
ABI377 DNA sequencer (Applied Biosystems, Norwalk, CT).
GFP-emerin expression and indirect immunofluorescence
staining in HeLa cells
HeLa cells were cultured in a 35 mm glass-bottom culture dish as
described previously (Haraguchi et al., 1997). Transfection of the
plasmid DNA encoding the wildtype and various mutations of GFP-
emerin was performed with LipofectaminePlus (Gibco BRL,
Rockville, MD) according to the manufacturer’s protocol except that
the incubation time of the cells with the reagent complexes was
reduced to 1.5 hours. Cells were cultured for 2 days under regular
culture conditions before being subjected to live microscopic
observation, as described previously (Haraguchi et al., 2000).
We hypothesized that residues conserved between emerin and
LAP2β might be important for emerin function, and therefore
targeted many of these conserved residues for mutation
(Fig. 1). We first generated a nearly full-length recombinant
human emerin protein consisting of residues 1-222, ending just
before the transmembrane domain. We then used site-directed
mutagenesis to construct 13 mutant emerin proteins, each
carrying a cluster of alanine substitutions in residues that are
identical between human emerin and human LAP2β. Mutant
clusters were numbered according to their most N-terminal
altered residue (Fig. 1).
Emerin binds directly to BAF in vitro
To test the hypothesis that emerin binds BAF, we first used in
vitro-transcribed/translated wildtype human BAF to probe
blots of immobilized emerin (‘blot overlay’ experiments).
JOURNAL OF CELL SCIENCE 114 (24)
4569Functional domains in emerin
Lysates from bacteria that expressed recombinant emerin
proteins (wildtype or mutant residues 1-222) were resolved by
SDS-PAGE, transferred to membranes and first probed with
antibodies against emerin (Fig. 2A), to demonstrate the
presence of each recombinant protein. Recombinant wildtype
emerin (residues 1-222) migrated with an apparent mass of 27
kDa in SDS-PAGE. Parallel blots were probed with 35S-labeled
human BAF, synthesized in coupled transcription/translation
reticulocyte lysates. Supporting our hypothesis, wildtype BAF
bound to wildtype recombinant emerin on the blots (Fig. 2B,
W). This interaction was specific since BAF did not bind to
several mutant emerins. All four LEM-domain mutations
(m11, m24, m30, m34) significantly reduced emerin binding
to BAF (Fig. 2B). Mutations in the central and C-terminal
regions of emerin did not disrupt the emerin-BAF interaction
(Fig. 2B; m70 to m214). These results suggested that emerin
interacts directly with BAF, and that residues within the LEM-
domain of emerin are required to bind BAF.
We used co-immunoprecipitation experiments to
independently confirm the emerin-BAF interaction. 35S-
labeled wildtype and mutant emerin proteins were
incubated with 35S-labeled human BAF, and precipitated
using polyclonal antibodies against human emerin (Fig. 3).
Supporting the blot overlay results, wildtype emerin
(residues 1-222) co-immunoprecipitated with wildtype
BAF (Fig. 3, WT). Co-immunoprecipitation results
confirmed that LEM-domain residues were essential for
emerin and BAF to interact in solution, since LEM-domain
mutants m11 to m34 all failed to co-immunoprecipitate
with BAF (Fig. 3). The other emerin mutants all co-
immunoprecipitated with BAF from solution (Fig. 3). Based
on these two independent lines of evidence we concluded that
emerin binds directly to BAF in vitro, and that this binding is
mediated by residues in the LEM-domain.
Central region of emerin binds lamin A
Emerin interacts with A-type lamins in vitro (Fairley et al.,
1999; Clements et al., 2000). To map the binding region for
lamin A, we tested our emerin mutants for binding to lamin A
in blot overlay experiments. Five mutations (m70, m76, m112,
m141 and m164) reduced emerin binding to lamin A; all five
mapped outside the LEM-domain, in the central region of
emerin (Fig. 2C). All other mutant emerins bound to lamin A
at least as well as wildtype emerin. Several emerin mutants,
notably m24 in the LEM-domain, reproducibly bound to lamin
A better than wildtype emerin (Fig. 2C), as estimated by
densitometry analysis (data not shown; see Discussion). When
Fig. 1. Mutagenesis of human emerin, targeting residues conserved with human LAP2β. Shown are the aligned amino acid sequences of human
emerin and human LAP2β, starting with the LEM-domain of each protein (residue 1 of emerin; residue 110 of LAP2β). Numbers on the top
line refer to the LAP2β sequence. The regions mutated in this study are indicated by lines; residues changed to alanine are indicated by A. The
number below each line refers to the amino acid sequence of emerin, and names each cluster of mutations according to its most N-terminal
altered residue. TMD, transmembrane domain.
Fig. 2. Blot overlay assays for emerin binding to BAF or lamin
A. Bacterial lysates containing wildtype (W) or mutant emerin
residues 1-222 were separated on gels, blotted and probed with:
(A) anti-emerin antibodies, (B) 35S-labeled BAF, or (C) 35S-
labeled lamin A. Mutants are numbered according to Fig. 1.
blots were probed with 35S-labeled lamin C, similar results
were seen but the signals were significantly weaker than for
lamin A (data not shown). Note that in competitive co-
immunoprecipitation assays, emerin prefers lamin C (Vaughan
et al., 2001). The first 566 residues of lamins A and C are
identical, but their C-termini differ (Lin and Worman, 1993).
Disease-associated emerin mutations
Most human emerin mutations yield cells that are null for
emerin protein. However, in four cases, comprising point
mutations S54F, P183H and P183T, and a small deletion
mutant protein is stable and localized at the nuclear envelope,
rather than being degraded like most other mutant emerins
(Fairley et al., 1999; Ellis et al., 1999; Haraguchi et al., 2001).
To determine if disease-causing mutations disrupted emerin
binding to BAF or lamin A, three of these ‘stable’ mutations
were introduced into recombinant emerin (residues 1-222). We
changed serine 54 to phenylalanine (S54F; referred to as
‘S54P’ in Fairley et al.) (Fairley et al., 1999), proline 183 to
histidine (P183H) (Ellis et al., 1999), and deleted five residues
to create the ∆YEESY mutation (referred to here as ∆95-99)
(Fairley et al., 1999). All three mutant proteins were tested for
direct binding to BAF and lamin A. Our controls were wildtype
emerin, mutant m24 (defective in binding BAF) (Fig. 2; Fig.
3) and mutant m141 (defective in binding lamin A; Fig. 2).
Mutants S54F and P183H both interacted with BAF in blot
overlay (Fig. 4A) and co-immunoprecipitation
assays (Fig. 4B), and also interacted with lamin
A in blot overlay assays (Fig. 4A). Thus, these
mutations did not disrupt binding to either BAF
or lamin A in vitro, consistent with their
positions within the proposed functional map
of emerin (Fig. 5). By contrast, mutation ∆95-
99 had no effect on emerin binding to BAF,
but significantly reduced its binding to lamin
A (Fig. 4A, lam A). This result strongly
supported the proposed lamin-binding domain
of emerin, where residues 95-99 map (Fig. 5).
These findings suggested that mutation ∆95-99
might cause disease by specifically disrupting
emerin attachment to lamins.
The above results showed that disease-
causing emerin mutants are active for
binding to BAF in biochemical assays. To
independently confirm these results, we tested
the disease-linked emerin mutants for binding
to BAF in living cells. Each disease mutation, plus the
alternative P183T allele (Ellis et al., 1999), was incorporated
into full-length emerin with Green Fluorescent Protein
attached to the N-terminus of emerin (GFP-emerin; see
Materials and Methods). Each mutant protein was transiently
expressed and localized in living HeLa cells. All four mutants
localized predominantly to the nuclear envelope during
interphase, with weak ER staining, and were indistinguishable
from wildtype emerin-GFP (Fig. 6A), as expected (Fairley et
al., 1999). These interphase results showed that our fusions to
GFP did not disrupt localization. We then followed the HeLa
cells as they progressed through mitosis, to determine if the
mutant emerins were able to interact with BAF in living cells,
based on a novel in vivo assay (Haraguchi et al., 2001). BAF
recruits emerin to co-localize at the ‘core’ region of telophase
chromosomes for about two minutes near the end of mitosis;
this ‘core’ localization appears to be critical for the assembly
of both emerin and A-type lamins (but not B-type lamins) into
re-forming nuclear envelopes (Haraguchi et al., 2001). ‘Core’
localization was present for wildtype emerin-GFP (Fig. 6B),
and absent in the negative control (Fig. 6B, mutant m24), as
expected. Notably, all four disease-linked mutations were
recruited to the ‘core’ region (Fig. 6B), demonstrating their
ability to bind BAF in vivo. These proteins subsequently
redistributed uniformly over the nuclear envelope, like
wildtype emerin, and continued to localize normally after
exiting mitosis (data not shown). The apparently normal
JOURNAL OF CELL SCIENCE 114 (24)
Fig. 3. Solution binding as assayed by co-
immunoprecipitation. Wildtype BAF, wildtype emerin (WT,
residues 1-222) and mutant emerin proteins (numbered as in
Fig. 1), were synthesized and 35S-labeled in vitro using
coupled transcription/translation reactions, and then
immunoprecipitated using immune (lanes Em, WT, 11-214)
or preimmune (pre) antiserum against emerin, or anti-BAF
antisera (BAF). In vitro translation of emerin yielded a 27
kDa long form (L), and often also yielded a prominent 23
kDa short form (S) (Östlund et al., 1999), assumed to arise
by translation initiation at an internal site, as well as several
Fig. 4. Effects of disease-associated mutations S54F, ∆95-99 and P183H on emerin
binding to BAF and lamin A. (A) Bacterial lysates containing wildtype (wt) emerin
protein, disease-linked emerins (S54F, ∆95-99, P183H) or alanine-substitution mutants
(m24 and m141) were separated on gels, blotted and probed with 35S-labeled BAF or
35S-labeled lamin A. (B) Wildtype and mutant emerin proteins were synthesized as 35S-
labeled proteins in vitro, mixed with 35S-labeled BAF, and immunoprecipitated with
immune (shown) or preimmune (not shown) antibodies against emerin (see Materials
4571Functional domains in emerin
recruitment of disease-causing emerin proteins to the ‘core’
region of assembling nuclear envelopes strongly supported our
in vitro findings that these mutant proteins are active for
binding to BAF. We propose that these mutations cause disease
at the molecular level, by specifically disrupting emerin
interactions with partners other than BAF during interphase.
Our discovery that emerin interacts with BAF in vitro brings a
potentially important new player into the picture for Emery-
Dreifuss muscular dystrophy. BAF is essential for life in C.
elegans(Zheng et al., 2000), where it is expressed in every cell
(M. Segura and K.L.W., unpublished). BAF is proposed to be
a DNA-bridging protein, based on the unique ability of BAF
dimers to assemble into discrete nucleoprotein complexes
consisting of six BAF dimers plus multiple dsDNAs (Zheng
et al., 2000). Cells that lack emerin also lack emerin-BAF
interactions, which might contribute to the molecular
mechanism of disease. In cells, emerin and BAF are strikingly
colocalized for about two minutes during telophase, at the
‘core’ region of telophase chromosomes (Haraguchi et al.,
2001). In cells that transiently express an exogenous mutant
BAF, emerin fails to localize at the core and is absent from the
subsequent assembled nuclei, suggesting a role for BAF in
recruiting and stabilizing emerin during nuclear assembly
(Haraguchi et al., 2001).
Proposed functional domains of emerin
Our strategy of mutagenizing small clusters of conserved
residues was highly effective. Every cluster of mutations from
residues 11 to 179 disrupted binding to either BAF or lamin A,
but not both, demonstrating that residues conserved between
emerin and LAP2β are indeed critical for emerin function. We
propose that the exposed (nucleoplasmic) region of emerin has
at least two independent domains, comprising an N-terminal
BAF-binding domain and a central lamin-binding domain, and
might also have additional domains relevant to disease (Fig. 5).
These domains are each discussed below.
The most N-terminal domain of emerin is the LEM motif
(residues 1-43), which is here demonstrated to bind BAF.
Consistent with this model, residues 1-65 (but not residues 1-
37) of emerin are sufficient to localize emerin to the ‘core’
region of telophase chromosomes in vivo (Haraguchi et al.,
2001). Our discovery that emerin binds BAF is also strongly
supported by the recently solved solution structure of the
constant region of LAP2 (Cai et al., 2001); this work showed
that the LEM-domain folds independently into a conserved
backbone structure (Cai et al., 2001; Laguri et al., 2001) with
surface features that complement a hydrophobic binding
pocket on the BAF dimer interface (Cai et al., 2001). The
ability of wildtype emerin and four disease-linked emerin
proteins to bind BAF, both in vitro and in living cells, strongly
suggests that (a) BAF interactions are central to emerin
function, and (b) for these particular mutant alleles, disease
may arise from disrupted binding to a partner other than BAF,
such as lamin A or a hypothetical novel partner.
Residues 70-178 comprise the proposed lamin A-binding
domain. This domain includes residues 117-170, which
function as a nuclear membrane retention signal for emerin
(Östlund et al., 1999), supporting our proposal that this region
interacts directly with lamins. Furthermore, EDMD-associated
mutation ∆95-99, which failed to bind lamin A in vitro, is more
susceptible to biochemical extraction from nuclei, consistent
with weakened binding to lamins (Ellis et al., 1998). Emerin
mutant ∆95-99 is localized at the nuclear envelope in EDMD
patients(Fairley et al., 1999) and when expressed in HeLa cells
(our results). This proper localization could be explained at
least two ways: this mutant might somehow remain competent
to bind lamin A in vivo, even though it fails to bind lamin A
in vitro. Alternatively, other partners (e.g. B-type lamins, BAF
or novel partners) might contribute to its localization in vivo.
Two findings support the idea that emerin localization in
humans depends on a partner other than lamin A, or multiple
partners. First, emerin localization at the nuclear envelope
is completely lost in C. elegans embryos that are depleted of
their only lamin (B-type; Gruenbaum et al., unpublished),
suggesting that lamins per se are essential for emerin
localization. Second, emerin and lamin A both fail to associate
with assembling nuclear envelopes in cells that express a
dominant mutant BAF (Haraguchi et al., 2001), implying that
BAF is key to localizing both emerin and lamin A. Together,
these findings indicate that emerin recruitment and retention at
the nuclear envelope is complicated, involving distinct
sequential interactions with BAF, A-type lamins and B-type
lamins. We suggest that emerin mutant ∆95-99 is recruited
appropriately by BAF, but its function is then compromised by
defective binding to lamin A. Thus in patients who express
emerin ∆95-99, emerin interactions with A-type lamins may be
Residues 179-222 define a potential third domain, which
was not required to bind either BAF or lamin A. Based on the
effectiveness of our mutagenesis strategy, and the fact that
mutations P183H and P183T cause disease, we propose that
this third region has a novel function. Interestingly, residues
176-222 are sufficient to localize the transmembrane domain
of emerin at the nuclear envelope (Haraguchi et al., 2001),
implying that the predicted ‘third’ domain of emerin might
interact with a partner found at or near the inner nuclear
Fig. 5. Functional domains of emerin defined in this study. Emerin
is depicted schematically, showing the LEM-domain (LEM),
transmembrane domain (TM), and position of each cluster of
mutations (inverted triangles; numbered as in Fig. 1). Mutations are
positioned to scale along the polypeptide sequence. Domains
defined in this study are the BAF binding domain (residues 1-50,
which include the LEM-domain), the lamin-binding domain
(residues 70-178) and a proposed third domain of unknown function (residues 179-222). Stars indicate the positions of human mutations that
cause Emery-Dreifuss muscular dystrophy. Shading at the right end of the proposed lamin-binding domain indicates less severely reduced
binding of lamin A to mutant m164.
membrane. Mutations at disease-linked residue P183
had no affect on emerin binding to BAF or lamin A,
either in vitro or in living HeLa cells. We therefore
propose that mutations at P183 (located within the
putative third domain) cause disease by disrupting
emerin binding to an unidentified new partner.
Our findings show both in vitro and in vivo that
the nucleoplasmic region of emerin has at least two
modular structural domains, which mediate its
binding to BAF and lamin A. Because two disease-
associated residues (S54 and P183) both lie outside
the BAF-binding and lamin-binding domains, we
speculate that these mutations might disrupt emerin
regulation, or define additional functional domains.
An important future question will be to determine
whether emerin interacts with its partners
simultaneously, or if binding to one partner can
displace or enhance binding to another partner.
Based on the enhanced lamin-binding activity of
some LEM-domain mutants, particularly m24, we
speculate that these domains might influence
each other intramolecularly. As precedent for
intramolecular regulation of LEM proteins, we note
that the binding affinity of LAP2 for BAF is reduced
three- to ninefold when the BAF-binding constant
region of LAP2 is linked to the ‘variable’ regions of
different LAP2 isoforms (Shumaker et al., 2001).
Implications for nuclear infrastructure
Our discovery that emerin binds BAF in a LEM-
domain-dependent manner, coupled to parallel
results for LAP2 (Furukawa, 1999; Shumaker et al.,
2001), strongly suggest that all LEM proteins can
bind BAF. Since BAF binds nonspecifically to
double-stranded DNA (Zheng et al., 2000), our
findings have important implications for chromatin
organization in the nucleus. LEM proteins, as a
family, are collectively positioned to play major roles
in chromatin attachment to the nuclear inner
membrane and lamin filaments. Emerin and other
LEM proteins are expressed in nearly all cells (Lin
et al., 2000), and some are abundant: the molar ratio
of LAP2 to lamin B in rat liver nuclei has been
estimated at 2-5%, enough to position one LAP2β
molecule every 25-50 nm along lamin filaments
(Foisner and Gerace, 1993). Furthermore, the
abundant α isoform of LAP2 co-localizes with lamin
A throughout the nuclear interior (Dechat et al.,
2000; Moir et al., 2000), meaning that the proposed
attachments between LEM proteins and BAF are not
restricted to the nuclear periphery, but could also
extend throughout the nuclear interior. Further study
JOURNAL OF CELL SCIENCE 114 (24)
Fig. 6. Localization of GFP-fused emerin mutants S54F,
∆95-99, P183H and P183T in living HeLa cells. HeLa
cells were transiently transfected to express the indicated
emerin mutant as a GFP-fusion protein. (A) GFP
fluorescence during interphase. (B) GFP fluorescence in
living cells 5-7 minutes after the metaphase-anaphase
transition, when wildtype emerin localizes to the ‘core’
region of telophase chromosomes. Bars, 10 µm.
4573Functional domains in emerin
of emerin-BAF interactions will be critical for understanding
chromatin organization in the nucleus, and the disease
mechanism of EDMD.
We thank R. Moir and R. Goldman for lamin constructs, R. Craigie
for human BAF cDNA, E. Abrams and J. Beneken for full-length
emerin cDNA, and A. Kowalski and J. Aniukwu for their help. We
thank M. Segura and R. Cole for thoughtful comments on the
manuscript. This work was supported by grants from the W.W. Smith
Charitable Trust and National Institutes of Health (R01-GM48646, to
K.L.W.), International Human Frontier Science Program (K.L.W. and
Y.H.), Japan Science and Technology Corporation (CREST; to Y.H.),
and Grant-in-Aid for Scientific Research B (T.H. and Y.H.).
Bione, S., Maestrini, E., Rivella, S., Mancini, M., Regis, S., Romeo, G. and
Toniolo, D. (1994). Identification of a novel X-linked gene responsible for
Emery-Dreifuss muscular dystrophy. Nat. Genet. 8, 323-327.
Bonne, G., Di Barletta, M. R., Varnous, S., Bécane, H.-M., Hammouda,
E.-H., Merlini, L., Muntoni, F., Greenberg, C. R., Gary, F. et al. (1999).
Mutations in the gene encoding lamin A/C cause autosomal dominant
Emery-Dreifuss muscular dystrophy. Nat. Genet. 21, 285-288.
Bonne, G., Mercuri, E., Muchir, A., Urtizberea, A., Bécane, H. M., Recan,
D., Merlini, L., Wehnert, M., Boor, R. et al. (2000). Clinical and molecular
genetic spectrum of autosomal dominant Emery-Dreifuss muscular dystrophy
due to mutations of the lamin A/C gene. Annals Neurol. 48, 170-180.
Cai, M., Huang, Y., Ghirlando, R., Wilson, K. L., Craigie, R. and Clore,
G. M. (2001). Solution structure of the constant region of nuclear envelope
protein LAP2 reveals two LEM-domain structures: one binds BAF and the
other binds DNA. EMBO J. 20, 4399-4407.
Chen, H. and Engelman, A. (1998). The barrier-to-autointegration protein is
a host factor for HIV type 1 integration. Proc. Natl. Acad. Sci. USA 95,
Clements, L., Manilal, S., Love, D. R. and Morris, G. E. (2000). Direct
interaction between emerin and lamin A. Biochem. Biophys. Res. Commun.
Cohen, M., Lee, K. K., Wilson, K. L. and Gruenbaum, Y. (2001).
Transcriptional repression, apoptosis, human disease and the functional
evolution of the nuclear lamina. Trends Biochem. Sci. 26, 41-47.
Dechat, T., Korbei, B., Vaughan, O. A., Vlcek, S., Hutchison, C. J. and
Foisner, R. (2000). Lamina-associated polypeptide 2α binds intranuclear A-
type lamins. J. Cell Sci. 113, 3473-3483.
Ellis, J. A., Craxton, M., Yates, J. R. W. and Kendrick-Jones, J. (1998)
Aberrant intracellular targeting and cell cycle-dependent phosphorylation of
emerin contribute to the Emery-Dreifuss muscular dystrophy phenotype. J.
Cell Sci. 111, 781-792.
Ellis, J. A., Yates, J. R. W., Kendrick-Jones, J. and Brown, C. A. (1999).
Changes at P183 of emerin weaken its protein-protein interactions resulting
in X-linked Emery-Dreifuss muscular dystrophy. Hum. Genet. 104, 262-
Emery, A. E. H. (1989) Emery-Dreifuss syndrome. J. Med. Genet. 26, 637-
Fairley, E. A., Kendrick-Jones, J. and Ellis, J. A. (1999). The Emery-
Dreifuss muscular dystrophy phenotype arises from aberrant targeting and
binding of emerin at the inner nuclear membrane. J. Cell Sci. 112, 2571-
Foisner, R. and Gerace, L. (1993). Integral membrane proteins of the nuclear
envelope interact with lamins and chromosomes, and binding is modulated
by mitotic phosphorylation. Cell 73, 1267-1279.
Furukawa, K. (1999). LAP2 binding protein 1 (L2BP1/BAF) is a candidate
mediator of LAP2-chromatin interaction. J. Cell Sci. 112, 2485-2492.
Goldberg, M., Lu, H. H., Stuurman, N., Ashery-Padan, R., Weiss, A. M.,
Yu, J., Bhattacharyya, D., Fisher, P. A., Gruenbaum, Y. and Wolfner,
M. F. (1998). Interactions among Drosophila nuclear envelope proteins
lamin, otefin, and YA. Mol. Cell Biol. 18, 4315-4323.
Haraguchi, T., Kaneda, T. and Hiraoka, Y. (1997). Dynamics of
chromosomes and microtubules visualized by multiple-wavelength
fluorescence imaging in living mammalian cells: effects of mitotic inhibitors
on cell cycle progression. Genes Cells 2, 369-380.
Haraguchi, T., Koujin, T., Hayakawa, T., Kaneda, T., Tsutsumi, C.,
Imamoto, N., Akazawa, C., Sukegawa, J., Yoneda, Y. and Hiraoka, Y.
(2000). Live fluorescence imaging reveals early recruitment of emerin, LBR,
RanBP2, and Nup153 to reforming functional nuclear envelopes. J. Cell Sci.
Haraguchi, T., Koujin, T., Segura, M., Lee, K. K., Matsuoka, Y., Yoneda,
Y., Wilson, K. L. and Hiraoka, Y. (2001). BAF is required for emerin
assembly into the reforming nuclear envelope. J. Cell Sci. 114, 4575-4585.
Laguri, C., Gilquin, B., Wolff, N., Romi-Lebrun, R., Courchay, K.,
Callebaut, I., Worman, H. J. and Zinn-Justin, S. (2001). Structural
characterization of the LEM motif common to three human inner nuclear
membrane proteins. Structure 9, 503-511.
Lee, K. K., Gruenbaum, Y., Spann, P., Liu, J. and Wilson, K. L. (2000).
C. elegans nuclear envelope proteins emerin, MAN1, lamin and
nucleoporins reveal unique timing of nuclear envelope breakdown during
mitosis. Mol. Biol. Cell 11, 3089-3099.
Lee, M. S. and Craigie, R. (1998). A previously unidentified host protein
protects retroviral DNA from autointegration. Proc. Natl. Acad. Sci. USA
Lin, F. and Worman, H. J. (1993). Structural organization of the human gene
encoding nuclear lamin A and nuclear lamin C. J. Biol. Chem. 268, 16321-
Lin, F., Blake, D. L., Callebaut, I., Skerjanc, I. S., McBurney, M. W.,
Pauline-Levasseur, M. and Worman, H. J. (2000). MAN1: an integral
protein of the inner nuclear membrane that shares the LEM domain with
lamina associated polypeptide2/thymopoietin, emerin and proteins of
Caenorhabditis elegans. J. Biol. Chem. 275, 4080-4087.
Manilal, S., Man, N. T., Sewry, C. A. and Morris, G. E. (1996). The Emery-
Dreifuss muscular dystrophy protein, emerin, is a nuclear membrane
protein. Hum. Mol. Genet. 5, 801-808.
Moir, R. D., Yoon, M., Khuon, S. and Goldman, R. D. (2000). Nuclear
lamins A and B1. Different pathways of assembly during nuclear envelope
formation in living cells. J. Cell Biol. 151, 1155-1168.
Morris, G. E. (2001). The role of the nuclear envelope in Emery-Dreifuss
muscular dystrophy. Trends Mol. Med. (in press).
Nagano, A., Koga, R., Ogawa, M., Kurano, Y., Kawada, J., Okada, R.,
Hayashi, Y. K., Tsukahara, T. and Arahata, K. (1996). Emerin deficiency
at the nuclear membrane in patients with Emery-Dreifuss muscular
dystrophy. Nat. Genet. 12, 254-259.
Östlund, C., Ellenberg, J., Hallberg, E., Lippincott-Schwartz, J. and
Worman, H. J. (1999). Intracellular trafficking of emerin, the Emery-
Dreifuss muscular dystrophy protein. J. Cell Sci. 112, 1709-1719.
Shumaker, D. K., Lee, K. K., Tanhehco, Y. C., Craigie, R. and Wilson, K.
L. (2001). LAP2 binds to BAF-DNA complexes: requirement for the LEM
domain and modulation by variable regions. EMBO J. 20, 1754-1764.
Sullivan, T., Escalente-Alcalde, D., Bhatt, H., Anver, M., Bhat, N.,
Nagashima, K., Stewart, C. L. and Burke, B. (1999). Loss of A-type
lamin expression compromises nuclear envelope integrity leading to
muscular dystrophy. J. Cell Biol. 147, 913-919.
Vaughan, O. A., Alvarez-Reyes, M., Bridger, J. M., Broers, J. L. V.,
Ramaekers, F. C. S., Wehnert, M., Morris, G. E., Whitfield, W. G. F.
and Hutchison, C. J. (2001). Both emerin and lamin C depend on lamin A
for localization at the nuclear envelope. J. Cell Sci. 114, 2577-2590.
Wilson, K. L., Zastrow, M. S. and Lee, K. K. (2001). Lamins and disease:
insights into nuclear infrastructure. Cell 104, 647-650.
Wolff, N., Gilquin, B., Courchay, K., Callebaut, I., Worman, H. J. and
Zinn-Justin, S. (2001). Structural analysis of emerin, an inner nuclear
membrane protein mutated in X-linked Emery-Dreifuss muscular dystrophy.
FEBS Lett. 501, 171-176.
Yorifuji, H., Tadano, Y., Tsuchiya, Y., Ogawa, M., Goto, K., Umetani, A.,
Asaka, Y. and Arahata, K. (1997). Emerin, deficiency of which causes
Emery-Dreifuss muscular dystrophy, is localized at the inner nuclear
membrane. Neurogenetics 1, 135-140.
Zheng, R., Ghirlando, R., Lee, M. S., Mizuuchi, K., Krause, M. and
Craigie, R. (2000). Barrier-to-autointegration factor (BAF) bridges DNA in
a discrete, higher-order nucleoprotein complex. Proc. Natl. Acad. Sci. USA