Molecular Pathogenesis
The genes EMD, LMNA, and FHL1 – pathogenic variants in which cause Emery-Dreifuss muscular dystrophy (EDMD) – encode proteins critical for the organization of the nuclear envelope. Although not entirely elucidated, two main mechanisms (not necessarily mutually exclusive) are thought to be involved in EDMD pathogenesis [Broers et al 2006, Worman & Bonne 2007, Worman et al 2009]:
Structural strain caused by mechanical stress present in skeletal muscle and cardiac muscle
Modification of gene expression caused by abnormal chromatin organization associated with alteration of proliferation/differentiation and/or signaling pathways of muscle cells
Interactions of these nuclear envelope proteins with chromatin- and nuclear matrix-associated proteins are of particular interest. Both emerin and lamin A/C interact with nuclear actin, a component of the chromatin remodeling complex associated with the nuclear matrix, suggesting that either chromatin arrangement or gene transcription or both could be impaired in the disease [Maraldi et al 2002].
EMD
Gene structure. The gene has six exons. For a detailed summary of gene and protein information, see Table A, Gene.
Pathogenic variants. More than 130 different pathogenic variants have been reported to date (see the UMD-EMD Database). The majority of pathogenic variants (95%) are null variants: nonsense variants, deletions/insertions, and splice site variants. A few reported missense variants and in-frame deletions lead to decreased expression of emerin or to normal expression of a nonfunctional protein [Ellis et al 1998, Yates et al 1999, Yates & Wehnert 1999, Ellis et al 2000]. Most pathogenic variants are unique to a single family. On occasion, two or three families have the same pathogenic variant. No "hot spot" for pathogenic variants is observed in EMD; pathogenic variants are nearly randomly spread out along the gene. (For more information, see Table A.)
Normal gene product. Emerin is a 254-amino-acid serine-rich protein expressed in most tissues. It belongs to a family of type II integral membrane proteins, including lamina-associated protein 2 (LP2; β-thymopoietin) and lamin B receptor. The hydrophobic tail anchors the protein to the inner nuclear membrane and the hydrophilic remainder of the molecule projects into the nucleoplasm, where it interacts with the nuclear lamina [Manilal et al 1996, Yorifuji et al 1997].
Emerin binds directly to lamins A/C and to BAF (OMIM 603811), a DNA-bridging protein. This binding requires conserved residues in a central lamin A-binding domain and the N-terminal LEM domain of emerin, respectively [Clements et al 2000, Lee et al 2001]. BAF is required for the assembly of emerin and A-type lamins at the reforming nuclear envelope during telophase of mitosis and may mediate their stability in the subsequent interphase [Haraguchi et al 2001].
Abnormal gene product. Most pathogenic variants result in no emerin production. In the rare cases in which protein is expressed, either the gene product is lacking the transmembrane domain (in-frame distal deletions) resulting in mislocalization of the protein in the nucleoplasm or cytoplasm, or the abnormal protein is present at the nuclear rim (missense variants) but has weakened interactions with the lamina components [Ellis et al 1999, Fairley et al 1999, Ellis et al 2000].
FHL1
Gene structure. The gene has eight exons. For a detailed summary of gene and protein information, see Table A, Gene.
Pathogenic variants. Of the more than 40 disease-associated variants reported in FHL1, seven have been associated with EDMD [Gueneau et al 2009, Knoblauch et al 2010]. EDMD-associated variants are localized in the distal exons (5-8) of FHL1: two missense variants affecting highly conserved cysteines, one abolishing the termination codon, and four out-of-frame insertions or deletions [Gueneau et al 2009]. (For more information, see Table A.)
Normal gene product. Three FHL1 isoforms are produced by alternative splicing of FHL1.
FHL1 proteins belong to a protein family containing LIM domains (Lin-11, Isl-1, Mec3), which are highly conserved sequences comprising two zinc fingers in tandem, implicated in numerous interactions. Each of the two zinc fingers contains four highly conserved cysteines linking together one zinc ion [Kadrmas & Beckerle 2004].
The main isoform, FHL1A, is predominantly expressed in striated muscles [Lee et al 1998, Taniguchi et al 1998]. FHL1A can be localized to the sarcolemma, sarcomere, and nucleus of muscle cells [Brown et al 1999, Ng et al 2001]. It has been implicated in sarcomere assembly by interacting with myosin-binding protein C [McGrath et al 2006].
The two other (less abundant) isoforms, FHL1B and FHL1C, are expressed in striated muscles [Brown et al 1999, Ng et al 2001]. FHL1A, FHL1B, and FHL1C are, respectively, composed of 4.5, 3.5, and 2.5 LIM domains. FHL1B and FHL1C have different C-terminal domains, which correspond to nuclear import and export signals in FHL1B and to the RBP-J binding domain in FHL1B and FHL1C [Brown et al 1999, Ng et al 2001].
Abnormal gene product. Pathogenic variants in FHL1 affect FHL1 isoforms differently since they are located in alternatively spliced exons:
LMNA
Gene structure.
LMNA encodes four transcripts via alternative splicing – two major transcripts: the full-length lamin A (exon 1-12) and a shorter transcript lamin C (exon 1-10); and two minor transcripts: lamin A-delta-10, which lacks exon 10, and lamin C2, which has a different N-terminal start (alternative exon 1) from lamin C. For a detailed summary of gene and protein information, see Table A, Gene.
Pathogenic variants. More than 450 LMNA pathogenic variants are reported to date. See the UMD-LMNA Database [Bonne et al 2003] and Leiden Muscular Dystrophy pages©. The majority (85%) of pathogenic variants are missense variants. Nonsense variants, small deletions/insertions in-frame or with frameshift, and splice site variants also occur. Pathogenic variants are distributed along the length of the gene [Bonne et al 2000, Brown et al 2001]. A few recurrent pathogenic variants exist [Broers et al 2006]. (For more information, see Table A.)
Pathogenic variants associated with autosomal recessive (AR) disease generally occur at different residues from those responsible for autosomal dominant (AD) disease. As yet, variants cannot be predicted to cause AR or AD disease.
Table 7.
Selected LMNA Pathogenic Variants
View in own window
DNA Nucleotide Change | Predicted Protein Change | Reference Sequences |
---|
c.664C>T | p.His222Tyr 1 |
NM_005572.3
|
c.674G>A | p.Arg225Gln |
Variants listed in the table have been provided by the authors. GeneReviews staff have not independently verified the classification of variants.
GeneReviews follows the standard naming conventions of the Human Genome Variation Society (varnomen.hgvs.org). See Quick Reference for an explanation of nomenclature.
- 1.
Normal gene product. Four A-type lamins (A, AΔ10, C, and C2) are produced by LMNA alternative splicing. Lamin A and lamin C are the two main isoforms. They are initially expressed in muscle of the trunk, head, and appendages. Later, they are ubiquitously expressed. A few myeloid and lymphoid cell lines have no lamins.
The promoter 1C2 located in the first intron of LMNA allows transcription of lamin C2. The fourth lamin is lamin AΔ10 (missing exon 10) described in cancer cells [Alsheimer & Benavente 1996, Machiels et al 1996]. Lamins are type V intermediate filaments that form the nuclear lamina, a fibrous network underlying the inner face of the internal nuclear membrane.
Transcription factors such as c-fos, pRb, and Lco1 have been identified as binding partners of Lamin A/C, suggesting possible deregulation of signaling pathways and alteration of proliferation/differentiation of muscle cells [Broers et al 2006, Vlcek & Foisner 2007, Worman & Bonne 2007, Azibani et al 2014].
Abnormal gene product. Missense variants are reported in the majority of cases. Western blot analysis on fibroblasts of affected individuals demonstrates a normal level of protein expression, strongly suggesting that abnormal proteins are expressed [Muchir et al 2004]. Nonsense variants resulting in approximately 50% of normal protein levels have also been described [Bécane et al 2000, Muchir et al 2003].