| Human Molecular Genetics | Pages |
©1999 Oxford University Press |
Heart to heart: from nuclear proteins to Emery-Dreifuss muscular dystrophy
Introduction
Clinical Features
Skeletal features
Cardiac features
Emerin, Lamins And The Nuclear Envelope
Mutations In Emerin And Lamin A/C
Acknowledgements
References
Heart to heart: from nuclear proteins to Emery-Dreifuss muscular dystrophy
Received April 20 1999; Revised and Accepted June 3, 1999
Emery-Dreifuss muscular dystrophy has some remarkably specific features, with only cardiac and skeletal tissues being affected. Equally remarkably, the disease is caused by mutations in widely expressed genes for the nuclear membrane/lamina proteins, emerin and lamin A/C. How do mutations in proteins at the heart of the cell lead to stiff joints and sudden heart failure? This and related questions are the subject of this review.
INTRODUCTION
When the gene for X-linked Emery-Dreifuss muscular dystrophy (EDMD; OMIM 310300) was first identified in 1994, it was clear from the sequence that the protein product, emerin, was a likely transmembrane protein, since it has a hydrophobic helix of ~20 amino acids near its C-terminus (1). It was widely anticipated that emerin would be found at the plasma membrane, since proteins associated with the sarcolemma are encoded by the genes responsible for Duchenne/Becker and most cases of limb-girdle and congenital muscular dystrophies (2). The discovery that emerin is a nuclear membrane protein (3,4) came as a complete surprise and left one wondering how the absence of a nuclear protein can cause the specific clinical features associated with EDMD [cardiac conduction defects, early contractures at the neck, ankles and elbows and slowly progressive wasting of certain specific muscles (5)]. The possibility that emerin is also found at intercalated discs in the heart (6) and perhaps at myotendinous junctions in skeletal muscle seemed to offer an attractive explanation, but the localization was not supported by a subsequent study using a wider range of anti-emerin antibodies (7). Further evidence for a problem at the nuclear periphery in EDMD came with the recent identification of lamin A/C as the gene on chromosome 1q21.3 responsible for an autosomal dominant form of EDMD (OMIM 181350) (8).
CLINICAL FEATURES
The peculiar characteristics of EDMD have been reviewed by Emery (5), clinical criteria have been defined by Yates (9) and Merlini has published a comprehensive reference list dating back to 1902 (10). There are two problems in understanding the pathogenesis of EDMD. First, emerin and lamin A/C are expressed in many tissues (1,4), so why are skeletal and cardiac muscles specifically affected? Second, how do molecular defects at the nuclear rim produce specific features, like contractures and cardiac conduction defects? Might specific cell death be responsible for the phenotype or must we look for more subtle changes? Muscle biopsies from EDMD patients, though myopathic, usually show little, if any, necrosis (5) although routine biopsies are often taken from less severely affected muscles. Conduction defects in myotonic dystrophy (DM) are often associated with visible damage to the conduction system, even though they are less advanced than in EDMD, complete heart block being rare in DM (11). For EDMD itself, few post-mortem studies have been published and, although replacement of myocardium by fibrous and adipose tissue has been observed, specific degeneration of the conduction system has not been reported (12). An emerin knockout mouse with cardiac symptoms would be a valuable research tool in view of the difficulty of obtaining EDMD hearts.
The EDMD phenotype displays very variable expressivity, both between families with different mutations and within families with the same mutation. Together with the slow development of many characteristic features, this can make diagnosis difficult, in particular in sporadic cases and in children and young adults (5,9). Confusion with both limb-girdle MDs (13) and rigid spine syndrome (14) have occurred. One family with an emerin null mutation was first thought to have limb-girdle MD and was tested for emerin deficiency only after a maternal cousin was found with typical EDMD features (13). Out of 17 family members with a lamin A/C mutation, 12 had cardiac symptoms only, while five showed the full phenotype (15). The phenotypes caused by emerin and lamin A/C mutations are clinically very similar, if not identical (15,16), suggesting a close functional relationship between the two proteins.
Skeletal features
Contractures are characterized by permanent muscle shortening and it can hardly be a coincidence that the most affected muscles in EDMD are in the lower leg and the upper arm which are associated with contractures at the ankle (Achilles tendon) and elbow, respectively. For example, an atypical autosomal EDMD patient displaying wasting of pelvic girdle and thigh, as well as peroneal, muscles (but not the upper arm) had contractures at the knee and hip (but not the elbow) (15,17). It is not clear, however, whether the primary changes leading to contracture and wasting occur in the muscle fibres, the tendons or the myotendinous junctions. This knowledge would help considerably in tracing the molecular pathogenesis. Contractures also occur in other muscular dystrophies, possibly as a result of muscle inactivity (16), but their early occurrence is characteristic of EDMD, with surgical release at the Achilles tendon commonly required in childhood.
Cardiac features
Cardiac arrythmias develop in virtually all EDMD patients by the third decade and progress towards complete heart block. Consequently, sudden heart failure in middle age is by far the most common cause of death. In normal heart, the sinoatrial (SA) node transmits ~90 impulses/min to the ventricles via the atrioventricular (AV) node and the bundle of His which branches into the subendocardial network of Purkinje fibres. Purkinje fibres are made up of specialized cardiomyocytes with few contractile myofibrils. Without the AV node, the ventricle will continue a slow rate of <40 beats/min typical of complete heart block, but this is unstable and subject to tachycardias or complete heart failure (asystole), so an external pacemaker must be fitted. Complete heart block in EDMD patients could be caused by specific loss of conduction system function, but the possibility remains that the conduction system is only particularly sensitive to pathological changes which occur in all cardiomyocytes, since ventricular dysfunction and a generalized loss of atrial myocardium have also been observed in some EDMD cases (12). This distinction has practical implications: should we look for molecular changes in the conducting system specifically or cardiomyocytes generally?
EMERIN, LAMINS AND THE NUCLEAR ENVELOPE
Emerin is a 254 amino acid, serine-rich protein which migrates on SDS-PAGE as a 34 kDa band (slower than the predicted 29 kDa) (3,4). It is a type II integral membrane protein, anchored to the inner nuclear membrane via its hydrophobic C-terminal tail, with the rest of the molecule projecting into the nucleoplasm (4). The inner membrane localization has now been confirmed by both electron microscopy (6,18) and antibody accessibility studies (19). The relationship of emerin to other inner nuclear membrane proteins, including the lamina-associated protein LAP2, which exists in alternatively spliced forms, is illustrated in Figure 1 (see legend for molecular details). Rat (20) and mouse (21) emerin sequences show >70% identity with the human sequence, but there are no shared sequences in the Drosophila melanogaster or Caenorhabditis elegans databases, except for the LEM domain, which is also present in mammalian LAP2 and MAN1 proteins (Fig. 1). Putative glycosylation and fibronectin-binding sites in human emerin (1) are not conserved and are obviously spurious. There are two major types of lamin: A-type (lamins A and C) and B-type (B1 and B2). Lamins A and C are two alternatively transcribed products of a gene assigned to human chromosome 1q21.3 (22); lamin C is essentially a short form of lamin A. The lamin B1 and lamin B receptor (LBR) genes localize to chromosomes 5q23.3-31.1 and 1q42.1, respectively (22), while lamin B2 is on chromosome 19p13.3 (23). LAP2 [chromosome 12q22 (24)] specifically interacts with lamin B1 in a phosphorylation-dependent manner whereas LAP1 interacts with lamins A, C and B1 (25). LBR interacts with B-type lamins (26), although possibly indirectly (27). Direct interaction of emerin with lamins has not been demonstrated although they do partially co-localize in interphase cells and at certain stages of mitosis (28). Thus, both emerin and lamins are found in discrete intranuclear spots in cultured cells (28,29), in nuclear `channels' in cardiomyocytes (7) and in the nuclear matrix remaining after chromatin extraction (30), as well as at the nuclear periphery. Current ideas about internal nuclear structure suggest that chromosomes occupy discrete regions in interphase nuclei with RNA processing and possibly other activities occurring in the spaces between them (31). Parts of chromosomes at the nuclear periphery are often inactive heterochromatin, suggesting that lamins and associated proteins may have a role in gene regulation. LBR, for example, interacts with the heterochromatin protein HP1 (32). The presence of emerin and lamins throughout the nucleus clearly extends their possible role in gene regulation, although the possibility that they represent a nuclear pool in transit to the nuclear periphery has not been ruled out. Tissue-specific differences in expression are now well documented for both emerin (3,7) and lamins (33,34) and could be related to differences in gene expression and differentiation.
Figure 1. Schematic diagram of interactions at the nuclear membrane/lamina. The assembly of the nuclear lamina and membrane has been reviewed (51,52). Five known transmembrane proteins of the inner nuclear membrane are shown. LAP2 is the best characterized; the [beta] form is shown, but additional splicing variants occur, including LAP2[gamma], which is shorter, and LAP2[alpha], which lacks the transmembrane sequence (24). LAP2[beta] interacts with chromatin through its N-terminal sequence (amino acids 1-85) and with B-type lamins through amino acids 298-373 (53,54). Between these two regions is a 38 amino acid sequence (114-152) which shows 40% homology with a sequence at the N-terminus of emerin (6-44) (1). This homology, termed the `LEM domain', is also shared by the recently identified MAN1 (55), which has two transmembrane sequences (H.J. Worman, personal communication). The LAP2-binding region of lamin B1 (78-258) has also been identified (56). The hydrophobic transmembrane sequences of emerin and LAP2 also show sequence homology, but the lamin B and chromatin-binding sequences of LAP2 are unrelated to emerin. LAP1C is unrelated in sequence to LAP2 (or emerin) but it binds to both A- and B-type lamins (57). LAP1 also occurs as three splicing variants, LAP1A, B and C. The lamin B receptor has eight transmembrane sequences related to sterol reductases (58); its chromatin-binding sequence (amino acids 71-100) has been identified (26) but the site of interaction with B-type lamins is not known and direct interaction has been questioned (27). An additional integral nuclear membrane protein, otefin, has been identified in Drosophila (59). Lamins B and A (but not lamin C) also attach directly to the inner nuclear membrane by isoprenylation, although this attachment is removed from lamin A by subsequent proteolysis (60).
The current hypothesis for targeting of inner nuclear membrane proteins is that the proteins are first inserted via their transmembrane sequences into the endoplasmic reticulum system (which includes the nuclear membrane). They then diffuse freely until they are trapped in the nuclear membrane by interactions of nucleoplasmic domains with, for example, the nuclear lamina and/or chromatin. Transfection studies with emerin fragments at first seemed inconsistent with this view, since the transmembrane domain alone seemed capable of directing emerin to the nuclear membrane (6), but more recent studies suggest that the isolated transmembrane domain locates throughout the ER system, as expected (19,35). Both later studies identified a region of emerin essential for nuclear envelope targeting [117-170 (19) or 107-175 (35) (Fig. 2)]. However, Östlund et al. (19) showed that residues 117-170, when attached to an artificial transmembrane sequence, remain in the ER unless additional emerin residues (3-117) are present, suggesting that two separate emerin sequences are involved in interactions with nuclear proteins or structures. In photobleaching experiments, emerin was more mobile in the nuclear membrane than LBR (19), suggesting that emerin is less firmly attached to immobile structures, such as chromatin. Although emerin is small enough to diffuse through the lateral channels of the nuclear pore complexes to the inner nuclear membrane, formation of even relatively small protein complexes inside the nucleus would prevent it from diffusing out again. No emerin-binding proteins have yet been identified, although both lamin A (36) and emerin (15) may bind actin, the emerin sequence involved being different from the nuclear targeting sequences. Studies of missense mutations in EDMD patients tend to support multiple nuclear targeting sequences. Five known missense mutations or small in-frame deletions are shown in Figure 2. The first, S54F, has not been studied at the protein level and the last, a deletion of six amino acids from the transmembrane domain, results in greatly reduced emerin levels and absence from the nuclear membrane in muscle biopsy (37). Q133H (38) and P183H or T (39) both have near normal emerin levels on western blots, but the former caused reduced nuclear membrane staining in muscle biopsy and the latter caused partial fractionation of emerin into non-nuclear fractions of lymphoblastoid cell lines (LCLs) with easier extractability of the nuclear emerin fraction. The five amino acid deletion of YEESY (Fig. 2) caused reduced emerin levels in LCLs with similar changes in emerin distribution and extractability (20). Q133 lies within and P183 just outside the Tsuchiya-Östlund sequence (green in Fig. 2), while YEESY is within the second region identified by Östlund et al. (19). Cartegni et al. (6) described an interesting two base frameshift deletion which by chance added a novel hydrophobic sequence to the first 169 amino acids of emerin. This chimeric emerin, having both a Tsuchiya-Östlund sequence and a hydrophobic C-terminal domain, locates correctly to the nuclear membrane (although in reduced amounts) (6,19). Although there are several potential serine protein kinase sites in emerin, S54 is not one of these; the first tyrosine in YEESY, however, is a potential phosphorylation site. Six different monoclonal antibody-binding sites on native emerin have been identified (Fig. 2) and there is some evidence that the two within the Tsuchiya-Östlund sequence can form a conformational epitope, implying protein folding in this region (7).
Figure 2. Features of the human emerin sequence. Emerin is represented as anchored in the inner nuclear membrane by its C-terminal `thymopoietin homology' domain (black; 1). Positions of exon boundaries, monoclonal antibody epitopes (blue; 7) and point mutations/in-frame deletions (orange; 41) are shown in relation to the amino acid sequence (approximate positions only). Potential phosphorylation sites (P) are shown in black (casein II kinase), red (protein kinase C), green (cAMP-dependent) and blue (tyrosine kinase); phosphorylation has not yet been shown at any specific site. The N-terminal `thymopoietin homology' domain, or LEM domain (LAP2-emerin-MAN1; H.J. Worman, personal communication), is shown in blue and the Tsuchiya-Östlund nuclear targeting sequence (19,35) in green. A sequence with homology to catenin-binding sequences in the APC tumor suppressor gene (6) is shown as `APC'.
MUTATIONS IN EMERIN AND LAMIN A/C
Nearly all emerin mutations result in complete loss of emerin or very great reduction in emerin levels (37). This enables rapid diagnosis of most X-linked cases by immunohistochemistry on skin biopsies or buccal smears or by western blotting of white blood cells (37,38,40). Missense mutations which are exceptions to this general rule have been discussed above. Most mutations are either nonsense point mutations or small frameshifting deletions/insertions, all of which result in early stop codons (41), although deletions of the whole gene also occur (41,42). There are no obvious mutation hotspots (41). Skewed X inactivation can occur, resulting in symptomatic females and a pseudo-autosomal pattern of inheritance when emerin levels are reduced by >95% (37). Since emerin mRNA levels are usually normal, absence of emerin suggests that truncated mutant emerins which lack the C-terminal transmembrane sequence are unstable under normal conditions (3,37).
In autosomal dominant EDMD, however, normal lamin A/C continues to be produced from the unaffected allele (8) so one might expect, naively perhaps, 50% of normal levels. Three mechanisms are commonly proposed for dominant negative effects: haploinsufficiency, gain of deleterious function or the existence of a functional multimer which cannot tolerate one defective subunit. A-type lamins certainly form multimeric filaments by side-to-side and end-to-end interactions (43) and missense mutations in lamin A/C are found in helical domains involved in polymerization (8). However, the early stop codon after only six amino acids in one family (8) suggests that EDMD can be caused by the mere absence of the product of one allele. A lamin A/C knockout mouse would be informative if it displayed EDMD, since gain of function would be ruled out and the heterozygote with 50% lamin A/C could be compared with the null homozygote. All mutations found so far would affect both lamin A and lamin C (8) and emerin levels and distribution are unaffected in autosomal EDMD (8,37). Although the tissue distribution of emerin resembles lamin A (7) and mutations in either protein give the same phenotype (15), there is no direct evidence for an emerin-lamin A interaction and emerin has no sequence similarity to LAP1, which does bind lamin A.
There may be other genes for EDMD yet to be discovered. One EDMD family with autosomal inheritance has no mutation in emerin or lamin A/C (44). Sporadic cases account for a high proportion of EDMD cases and rarely have emerin mutations (37). It is unlikely that all of these have lamin A/C mutations. Because of the similarities between EDMD and limb-girdle MD, there is speculation that autosomal dominant LGMD1B, mapped to chromosome 1q (45), may have a lamin A/C mutation and a more remote possibility that AD-LGMD1A on chromosome 5q (46) may have a lamin B1 mutation. Autosomal recessive forms of EDMD have also been reported (47).
Several authors have speculated on the mechanism by which changes in nuclear envelope proteins might cause the specific EDMD phenotype. Östlund et al. (19) suggested that emerin may regulate gene expression in heart and muscle by interacting with specific transcription factors or DNA sequences. Tsuchiya et al. (35) suggested that heart and muscle cells are particularly sensitive to mechanical stress and that emerin may be part of a nucleo-cytoskeletal network which protects cells from stress (48). Electron microscopy has revealed physical damage to skeletal muscle nuclei in EDMD, including detachment of membrane and lamina (49,50), although it is not yet clear whether this is a cause of the pathology or a secondary effect. Manilal et al. (7) suggested that, since cardiac and skeletal muscle nuclei appear to lack lamin B1, they may be particularly sensitive to loss of either emerin or lamin A/C. This hypothesis assumes that emerin-lamin A interactions provide back-up for the LAP2/LBR-lamin B1 system, so that problems occur when both systems fail. Finally, Cartegni et al. (6) proposed that absence of emerin from intercalated discs in the heart is responsible for conduction defects, but this was not supported by the separation of disc-staining antibodies from nucleus-staining antibodies in antisera or by the existence of monoclonal antibodies which stained only the nuclear membrane (7). However, Östlund et al. (19) have recently shown that overexpressed emerin in COS cells can find its way to the plasma membrane, where its diffusion is restricted, thus resurrecting the issue of non-nuclear emerin.
In conclusion, molecular studies of emerin function are well under way and nuclear lamins will certainly have to be looked at in a new light. Whether an emerin-lamin A/C system, analogous to the LAP2-lamin B system, will emerge remains to be seen. Hopefully, studies of a rare genetic disease may help us to understand the much commoner medical problem of cardiac conduction defects in general.
ACKNOWLEDGEMENTS
We thank Howard Worman and Juliet Ellis for providing preprints of their publications and Alan Emery, Caroline Sewry and Kathy Wilson for valuable comments on the manuscript. This work was supported by a British Heart Foundation grant (PG97142).
REFERENCES
+To whom correspondence should be addressed. Tel: +44 1978 293330; Fax: +44 1978 290008; Email: morrisge{at}newi.ac.uk
This page is run by Oxford University Press, Great Clarendon Street, Oxford OX2 6DP, as part of the OUP Journals
Comments and feedback: jnl.info{at}oup.co.uk
Last modification:
Copyright© Oxford University Press, 1999.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  J. M. Holaska Emerin and the Nuclear Lamina in Muscle and Cardiac Disease Circ. Res., July 3, 2008; 103(1): 16 - 23. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Heydemann, A. Demonbreun, M. Hadhazy, J. U. Earley, and E. M. McNally Nuclear sequestration of {delta}-sarcoglycan disrupts the nuclear localization of lamin A/C and emerin in cardiomyocytes Hum. Mol. Genet., February 15, 2007; 16(4): 355 - 363. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Halaschek-Wiener and A. Brooks-Wilson Progeria of Stem Cells: Stem Cell Exhaustion in Hutchinson-Gilford Progeria Syndrome J. Gerontol. A Biol. Sci. Med. Sci., January 1, 2007; 62(1): 3 - 8. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. Ozawa, Y. K. Hayashi, M. Ogawa, R. Kurokawa, H. Matsumoto, S. Noguchi, I. Nonaka, and I. Nishino Emerin-Lacking Mice Show Minimal Motor and Cardiac Dysfunctions with Nuclear-Associated Vacuoles Am. J. Pathol., March 1, 2006; 168(3): 907 - 917. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. M. Hodzic, D. B. Yeater, L. Bengtsson, H. Otto, and P. D. Stahl Sun2 Is a Novel Mammalian Inner Nuclear Membrane Protein J. Biol. Chem., June 11, 2004; 279(24): 25805 - 25812. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 O. Cohen-Barak, Z. Yi, N. Hagiwara, K. Monzen, I. Komuro, and M. H. Brilliant Sox6 regulation of cardiac myocyte development Nucleic Acids Res., October 15, 2003; 31(20): 5941 - 5948. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Liu, K. K. Lee, M. Segura-Totten, E. Neufeld, K. L. Wilson, and Y. Gruenbaum MAN1 and emerin have overlapping function(s) essential for chromosome segregation and cell division in Caenorhabditis elegans PNAS, April 15, 2003; 100(8): 4598 - 4603. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. Markiewicz, T. Dechat, R. Foisner, Roy. A Quinlan, and C. J. Hutchison Lamin A/C Binding Protein LAP2alpha Is Required for Nuclear Anchorage of Retinoblastoma Protein Mol. Biol. Cell, December 1, 2002; 13(12): 4401 - 4413. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Vlcek, B. Korbei, and R. Foisner Distinct Functions of the Unique C Terminus of LAP2alpha in Cell Proliferation and Nuclear Assembly J. Biol. Chem., May 17, 2002; 277(21): 18898 - 18907. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Y. Gruenbaum, K. K. Lee, J. Liu, M. Cohen, and K. L. Wilson The expression, lamin-dependent localization and RNAi depletion phenotype for emerin in C. elegans J. Cell Sci., January 3, 2002; 115(5): 923 - 929. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Izumi, O. A. Vaughan, C. J. Hutchison, and D. M. Gilbert Head and/or CaaX Domain Deletions of Lamin Proteins Disrupt Preformed Lamin A and C But Not Lamin B Structure in Mammalian Cells Mol. Biol. Cell, December 1, 2000; 11(12): 4323 - 4337. [Abstract] [Full Text]
|
|
|
|
|
|
K. K. Lee, Y. Gruenbaum, P. Spann, J. Liu, and K. L. Wilson C. elegans Nuclear Envelope Proteins Emerin, MAN1, Lamin, and Nucleoporins Reveal Unique Timing of Nuclear Envelope Breakdown during Mitosis Mol. Biol. Cell, September 1, 2000; 11(9): 3089 - 3099. [Abstract] [Full Text]
|
|
|
|
 |
|
 R. A. Hegele, C. M. Anderson, J. Wang, D. C. Jones, and H. Cao Association Between Nuclear Lamin A/C R482Q Mutation and Partial Lipodystrophy with Hyperinsulinemia, Dyslipidemia, Hypertension, and Diabetes Genome Res., May 1, 2000; 10(5): 652 - 658. [Abstract] [Full Text]
|
|
|
|
|
|
N. Hagiwara, S. E. Klewer, R. A. Samson, D. T. Erickson, M. F. Lyon, and M. H. Brilliant Sox6 is a candidate gene for p100H myopathy, heart block, and sudden neonatal death PNAS, April 11, 2000; 97(8): 4180 - 4185. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J Gotzmann, S Vlcek, and R Foisner Caspase-mediated cleavage of the chromosome-binding domain of lamina-associated polypeptide 2 alpha J. Cell Sci., January 11, 2000; 113(21): 3769 - 3780. [Abstract] [PDF]
|
|
|
|
|
|
H. Cao and R. A. Hegele Nuclear lamin A/C R482Q mutation in Canadian kindreds with Dunnigan-type familial partial lipodystrophy Hum. Mol. Genet., January 1, 2000; 9(1): 109 - 112. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. M. Graham and W. A. Owens Pathogenesis of Inherited Forms of Dilated Cardiomyopathy N. Engl. J. Med., December 2, 1999; 341(23): 1759 - 1762. [Full Text]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||