A laminopathic mutation disrupting lamin filament assembly causes disease-like phenotypes in C. elegans

Mutations in the human LMNA gene underlie many laminopathic diseases, including EmeryDreifuss muscular dystrophy (EDMD); however, a mechanistic link between the effect of mutations on lamin filament assembly and disease phenotypes has not been established. We studied the ∆K46 C. elegans lamin mutant, corresponding to EDMD-linked ∆K32 in human lamins A/C. Cryo-electron tomography of lamin ∆K46 filaments in vitro revealed alterations in the lateral assembly of dimeric head-to-tail polymers, which causes abnormal organization of tetrameric protofilaments. GFP:∆K46 lamin expressed in C. elegans was found in nuclear aggregates in post-embryonic stages along with LEM-2. GFP:∆K46 also caused mislocalization of emerin away from the nuclear periphery, consistent with a decreased ability of purified emerin to associate with lamin ∆K46 filaments in vitro. GFP:∆K46 animals had motility defects and muscle structure abnormalities. These results show that changes in lamin filament structure can translate into disease-like phenotypes via altering the localization of nuclear lamina proteins and suggests a model for how the ∆K32 lamin mutation may cause EDMD in human. http://www.molbiolcell.org/content/suppl/2011/06/04/mbc.E11-01-0064.DC1 Supplemental Material can be found at: 2 Introduction The nuclear lamina is a meshwork of proteins adjacent to the nucleoplasmic face of the inner nuclear membrane (INM) and serves as both a main nuclear structural element as well as an interface between the nucleoplasm and the nuclear membrane (Fawcett, 1966; Aebi et al., 1986). The evolutionarily conserved protein lamin is the major component of the nuclear lamina and is found as two types (A-type and B-type) in humans (Gruenbaum et al., 2003). Mutations in the human A-type lamin gene (LMNA) cause at least 14 diseases, collectively termed laminopathies (Broers et al., 2006; Mattout et al., 2006; Worman and Bonne, 2007). Among the autosomaldominant laminopathies are the muscle diseases Emery-Dreifuss muscular dystrophy (EDMD), LMNA-related congenital muscular dystrophy (L-CMD), limb-girdle muscular dystrophy, and dilated cardiomyopathy; the aging and skin disorder restrictive dermopathy; the neuropathic Charcot-Marie-Tooth disorder; the lipodystrophies mandibuloacral dysplasia and Dunnigan-type familial partial lipodystrophy; and the systemic premature aging syndromes Hutchison-Gilford progeria syndrome and atypical Werner syndrome (Vlcek and Foisner, 2007; Quijano-Roy et al., 2008). Intriguingly, the wide range of tissue types affected and severity of the disease phenotypes cannot be directly correlated with the position of the mutation in the LMNA gene (Worman and Bonne, 2007). There are three major models to explain how lamin mutations may lead to tissue-specific phenotypes. The first model gives lamin a role as the main structural element of the nucleus; disrupting this foundation may cause tissues under higher stress (such as the mechanical load on muscle cells) to be more prone to lesions leading to specific tissue degeneration (Ostlund et al., 2001). The second model suggests that lamin is a key regulator of signaling pathways affecting DNA replication, transcription, and chromatin organization, mediated through interactions between lamin, lamin-binding proteins, and chromatin (Cohen et al., 2001; Gruenbaum et al., 2005; Vlcek and Foisner, 2007; Dechat et al., 2008). The third model suggests that lamins A and C are involved in regulating cell type-specific gene expression during adult stem cell differentiation (Gotzmann and Foisner, 2006). These three models are not mutually exclusive; therefore, studying mutant lamin structure in isolation from the physiological effects of the mutation may not fully explain how a mutation can cause a specific laminopathy. Based on its sequence and predicted structure, lamin is classified as a type-V intermediate filament (IF) protein (Steinert and Roop, 1988). It has a short, unstructured amino-terminal head domain, an alpha-helical coiled-coil rod domain made of four segments of heptad repeats, and a carboxy-terminal tail domain containing a nuclear localization signal and an immunoglobulin (Ig)-fold (Strelkov et al., 2004; Ben-Harush et al., 2009). Lamins assemble into a network of 10nm filaments beneath the INM, as first described in the Xenopus germinal vesicle (Aebi et al., 1986), but their structure in somatic cells is still unknown. While the general morphology of the lamin network has been studied in vivo, further elucidation of the higher order assembly pattern and dynamics of filament formation has required analysis of lamin structure in vitro (Moir et al., 1991; Geisler et al., 1998; Stuurman et al., 1998; Karabinos et al., 2003). Intriguingly, under the tested experimental conditions, lamins from most species, including human lamins, form paracrystalline arrays instead of 10-nm filaments in vitro (Moir et al., 1991; Stuurman et al., 1998, Melcer 2007, Herrmann and Foisner 2003; Taimen et al., 2009). While these structures have been recapitulated by a vast overexpression of Drosophila and Xenopus lamins in insect cells, their physiological relevance is not clear (Klapper et al., 1997). Importantly, depending on the specific assembly conditions, lamin from Caenorhabditis elegans can form either paracrystalline arrays or 10 nm IF-like filaments (Karabinos et al., 2003; Foeger et al., 2006);