Portal de Periódicos da CAPES

Lamins and Disease

2001; Cell Press; Volume: 104; Issue: 5 Linguagem: Inglês

10.1016/s0092-8674(01)00261-6

1097-4172

Katherine L. Wilson, Michael S. Zastrow, KK Lee,

RNA regulation and disease

Two distinct human diseases (“laminopathies”) map to LMNA, the gene encoding A-type lamins (Bonne et al. 2000Bonne G. Mercuri E. Muchir A. Urtizberea A. Bécane H.M. Recan D. Merlini L. Wehnert M. Boor R. Reuner U. et al.Ann. Neurol. 2000; 48: 170-180Crossref PubMed Scopus (371) Google Scholar). These diseases took cell biologists by surprise, because the pathophysiological mechanisms are far from obvious. Each disease selectively strikes one or more specific tissues. Emery-Dreifuss muscular dystrophy (EDMD) begins in childhood and causes progressive muscle weakening, contractures of the Achilles, elbow, and neck tendons, and cardiac conduction defects that can cause sudden cardiac arrest (see Bonne et al. 2000Bonne G. Mercuri E. Muchir A. Urtizberea A. Bécane H.M. Recan D. Merlini L. Wehnert M. Boor R. Reuner U. et al.Ann. Neurol. 2000; 48: 170-180Crossref PubMed Scopus (371) Google Scholar). Mutations that cause EDMD are primarily missense mutations that map throughout the protein and act dominantly, although a few recessive mutations have also been identified (see Bonne et al. 2000Bonne G. Mercuri E. Muchir A. Urtizberea A. Bécane H.M. Recan D. Merlini L. Wehnert M. Boor R. Reuner U. et al.Ann. Neurol. 2000; 48: 170-180Crossref PubMed Scopus (371) Google Scholar). The severity of EDMD varies widely, even within the same family. Indeed, two other autosomal dominant diseases, limb girdle muscular dystrophy (type 1B) and dilated cardiomyopathy with conduction system disease (type 1A; see Bonne et al. 2000Bonne G. Mercuri E. Muchir A. Urtizberea A. Bécane H.M. Recan D. Merlini L. Wehnert M. Boor R. Reuner U. et al.Ann. Neurol. 2000; 48: 170-180Crossref PubMed Scopus (371) Google Scholar), also map throughout LMNA, and appear to be nonoverlapping clinical subsets of full-blown EDMD. The second distinct “laminopathy” is Dunnigan-type familial partial lipodystrophy, which begins at puberty and causes the selective loss of subcutaneous fat from the extremities, trunk, and gluteal region, and the accumulation of white fat tissue at the face, neck, back, and groin (Garg 2000Garg A. Am. J. Med. 2000; 108: 143-152Abstract Full Text Full Text PDF PubMed Scopus (331) Google Scholar). This lipodystrophy is dominant and rare, affecting less than one in 15 million people, with serious medical consequences including type II (insulin resistant) diabetes. Mutations that cause lipodystrophy are extremely specific, mapping to residues 465, 482, or 486 in all A-type lamins, or residues 582 or 584 in the lamin A tail. The key to these baffling diseases lies in understanding the structure and functions of the nuclear lamina, which is no small task. A human nucleus is typically 10 μm in diameter, and organizes ∼1 meter of chromosomal DNA. The lamina is a network of polymeric filaments inside the nucleus that consists of lamin proteins and associated lamin binding proteins (Figure 1). When stained by indirect immunofluorescence, the lamina stains prominently near the nuclear envelope, where the inner membrane is rich in lamin binding proteins (see below). However, this peripheral lamina is only part of the story. Over two-thirds of lamins are located in the nuclear interior (Moir et al. 2000aMoir R.D. Yoon M. Khuon S. Goldman R.D. J. Cell Biol. 2000; 151 (a): 1155-1168Crossref PubMed Scopus (302) Google Scholar and references therein; Liu et al. 2000Liu J. Ben-Shahar T.R. Riemer D. Treinin M. Spann P. Weber K. Fire A. Gruenbaum Y. Mol. Biol. Cell. 2000; 11: 3937-3947Crossref PubMed Scopus (323) Google Scholar). These internal lamins form stable structures, but the nature of these structures is completely unknown. Lamins have affinity for DNA, chromatin, and histones (see Stuurman et al. 1998Stuurman N. Heins S. Aebi U. J. Struct. Biol. 1998; 122: 42-66Crossref PubMed Scopus (568) Google Scholar), which might allow interior lamins to associate with and organize chromatin. For example, lamins are essential for DNA replication, which is initiated in a spatially organized manner beginning at a small number of lamin-containing foci near the nucleolus (Kennedy et al. 2000Kennedy B.K. Barbie D.A. Classon M. Dyson N. Harlow E. Genes Dev. 2000; 14: 2855-2868Crossref PubMed Scopus (243) Google Scholar, Moir et al. 2000bMoir R.D. Spann T. Herrmann H. Goldman R. J. Cell Biol. 2000; 149 (b): 1179-1192Crossref PubMed Scopus (170) Google Scholar). Furthermore, the nucleus and chromosomes are primed to respond to myriad signaling pathways during cell division, development, differentiation, and normal cell function. Some signals trigger dramatic, large-scale structural changes, such as chromosome condensation and nuclear breakdown during mitosis (Dechat et al. 2000Dechat T. Vlcek S. Foisner R. J. Struct. Biol. 2000; 129: 335-345Crossref PubMed Scopus (124) Google Scholar), the structural repression of selected chromatin in differentiating cells, and changes in nuclear shape (e.g., neutrophils and sperm). Other signals can precisely change the expression of a few specific genes. Thus, in addition to the well-established barrier function of the nuclear envelope, the animal nucleus assembles and maintains an infrastructure that accommodates the chromosomes and somehow facilitates all of the above-described activities. Lamins are key structural elements of this still-mysterious infrastructure. Below, we describe the lamins and their binding partners, highlighting findings that clarify the emerging “laminopathy” class of human disease. Lamins are nuclear-specific intermediate filament (IF) proteins. Like cytoplasmic IFs, lamins consist of a small N-terminal “head,” long coiled-coil “rod,” and globular C-terminal “tail” domains. Pairs of lamins interact in parallel via their coiled-coil regions to form lamin dimers. Lamin dimers interact head-to-tail to form polymers, which then associate with each other in an antiparallel manner (Stuurman et al. 1998Stuurman N. Heins S. Aebi U. J. Struct. Biol. 1998; 122: 42-66Crossref PubMed Scopus (568) Google Scholar). Beyond this stage, the structure of lamin filaments is not understood. Lamins are also the least understood IF proteins in terms of their mechanical properties (strength, flexibility). Vertebrate lamins fail to form stable 10 nm filaments in vitro, and are unique among IFs in their ability to assemble into orthogonal networks in vivo (Stuurman et al. 1998Stuurman N. Heins S. Aebi U. J. Struct. Biol. 1998; 122: 42-66Crossref PubMed Scopus (568) Google Scholar). Thus, one potential function for lamin binding proteins is to regulate lamin assembly. Consistent with this possibility, two alternatively spliced isoforms of Lamina Associated Polypeptide-2 (LAP2) are differentially expressed during development, and have distinct patterns of localization during nuclear envelope assembly and during interphase (Dechat et al. 2000Dechat T. Vlcek S. Foisner R. J. Struct. Biol. 2000; 129: 335-345Crossref PubMed Scopus (124) Google Scholar). The nuclear lamina is not a fixed structure. Although lamin filaments are stable and resist biochemical extraction during interphase, they depolymerize during mitosis due to phosphorylation at sites flanking the rod domain (Stuurman et al. 1998Stuurman N. Heins S. Aebi U. J. Struct. Biol. 1998; 122: 42-66Crossref PubMed Scopus (568) Google Scholar). Lamins also undergo posttranslational processing by proteolysis and isoprenylation (Stuurman et al. 1998Stuurman N. Heins S. Aebi U. J. Struct. Biol. 1998; 122: 42-66Crossref PubMed Scopus (568) Google Scholar), and cleavage during apoptosis (see Cohen et al. 2001Cohen, M., Lee, K.K., Wilson, K.L., and Gruenbaum, Y. (2001). Trends Biochem. Sci. 26, 41–47.Google Scholar). The prenylated precursor form of lamin A, which is proteolytically cleaved to the mature protein, interacts with a novel protein named Narf at sites within the nuclear interior (Barton and Worman 1999Barton R.M. Worman H.J. J. Biol. Chem. 1999; 274: 30008-30018Crossref PubMed Scopus (85) Google Scholar). During interphase, phosphorylation at serines, threonines, and tyrosines is proposed to control lamin interactions, for example to accommodate the assembly or rearrangement of lamin filaments (see references in Cohen et al. 2001Cohen, M., Lee, K.K., Wilson, K.L., and Gruenbaum, Y. (2001). Trends Biochem. Sci. 26, 41–47.Google Scholar). The assembly of B-type lamins after mitosis requires phosphatase PP1, which is recruited to the nuclear envelope by an integral membrane protein named AKAP149 (protein kinase A anchoring protein; Steen et al. 2000Steen R.L. Martins S.B. Taskén K. Collas P. J. Cell Biol. 2000; 150: 1251-1261Crossref PubMed Scopus (128) Google Scholar). Proteins such as AKAP149 and Narf suggest that lamina assembly and dynamics—and hence nuclear infrastructure—may be regulated locally for specific purposes within the nucleus. Humans have three lamin genes: LMNA and two genes that encode B-type lamins (LMNB1 and LMNB2). A- and B-type lamins are related, but have different sequences and biochemical properties (Stuurman et al. 1998Stuurman N. Heins S. Aebi U. J. Struct. Biol. 1998; 122: 42-66Crossref PubMed Scopus (568) Google Scholar). Because every mammalian cell expresses at least one B-type lamin, one or both B-type lamins are hypothesized to be essential. Through alternative splicing, LMNA encodes four isoforms, the most widely studied being lamins A and C. The LMNA gene appeared late in evolution, and A-type lamins are expressed primarily in differentiated cells (Cohen et al. 2001Cohen, M., Lee, K.K., Wilson, K.L., and Gruenbaum, Y. (2001). Trends Biochem. Sci. 26, 41–47.Google Scholar, and references therein). Thus, A-type lamins probably have specialized roles within the nucleus. An interesting question is to what extent A- and B-type lamins depend on each other for their assembly or function. Collectively, lamins are essential for life in multicellular animal eukaryotes. The null phenotype for lamins was recently reported in C. elegans, which has a single lamin (B-type; Liu et al. 2000Liu J. Ben-Shahar T.R. Riemer D. Treinin M. Spann P. Weber K. Fire A. Gruenbaum Y. Mol. Biol. Cell. 2000; 11: 3937-3947Crossref PubMed Scopus (323) Google Scholar). The loss-of-function phenotype is complicated, as predicted from partial loss-of-function studies of the Drosophila B-type lamin (flies also have an A-type lamin; see references in Cohen et al. 2001Cohen, M., Lee, K.K., Wilson, K.L., and Gruenbaum, Y. (2001). Trends Biochem. Sci. 26, 41–47.Google Scholar). In C. elegans, RNAi experiments show that lamin depletion is lethal during embryogenesis. As the pool of lamins becomes depleted, embryonic cells display many phenotypes including clustered (rather than evenly spaced) nuclear pore complexes, rapidly changing nuclear shape (on a timescale less than five minutes), and defects in chromosome organization and segregation (Liu et al. 2000Liu J. Ben-Shahar T.R. Riemer D. Treinin M. Spann P. Weber K. Fire A. Gruenbaum Y. Mol. Biol. Cell. 2000; 11: 3937-3947Crossref PubMed Scopus (323) Google Scholar). These phenotypes provide definitive new evidence that lamins are essential for the structural integrity of the nucleus, and are provocative in their implication that lamins are also essential for efficient chromosome segregation. Another phenotype appeared in a small number of “lamin-depleted” nematodes, which retained low levels of lamin expression and thereby escaped embryonic lethality. These animals had a high incidence of sterility due to reduced numbers of germ cells (Liu et al. 2000Liu J. Ben-Shahar T.R. Riemer D. Treinin M. Spann P. Weber K. Fire A. Gruenbaum Y. Mol. Biol. Cell. 2000; 11: 3937-3947Crossref PubMed Scopus (323) Google Scholar). Although germ cells are quite sensitive to disruption, these results suggest that the nuclear lamina is involved, directly or indirectly, in functions specific to particular cell types. New evidence from Drosophila supports this view. Bicaudal-D (BICD) is a Drosophila protein located primarily in the cytoplasm, which is required to establish oocyte cell fate and determine cell polarity. BICD has a coiled-coil region that interacts directly with the Drosophila B-type lamin in vitro (Stuurman et al. 1999Stuurman N. Häner M. Sasse B. Hubner W. Suter B. Aebi U. Eur. J. Cell Biol. 1999; 78: 278-287Crossref PubMed Scopus (32) Google Scholar). This interaction is specifically disrupted by a point mutation in BICD that dominantly disrupts BICD function in vivo, suggesting that its binding to lamin is physiologically relevant. If the lamina does play a role in cell polarity, it should be possible to identify lamin mutations that cause polarity defects during development. The mouse LMNA knockout strongly supports the idea that A-type lamins play cell-type-specific roles (Sullivan et al. 1999Sullivan T. Escalente-Alcalde D. Bhatt H. Anver M. Bhat N. Nagashima K. Stewart C.L. Burke B. J. Cell Biol. 1999; 147: 913-919Crossref PubMed Scopus (915) Google Scholar). LMNA null mice are born apparently normal, but develop severe forms of both muscular dystrophy and lipodystrophy starting around three weeks after birth, and die by eight weeks. A-type lamins are therefore essential only during adulthood, and the loss of LMNA is critical for only a few specific tissues, including muscle and adipose tissue. The big question for human laminopathy and nuclear infrastructure, is how to explain the specific pathophysiologies caused by defects in A-type lamins or lamin binding proteins such as emerin (below). Most proteins that are known (or suspected) to bind lamins are located at the nuclear inner membrane. Such proteins include the Lamin B Receptor (LBR), all isoforms of LAP1, most isoforms of LAP2, emerin, MAN1, otefin, and Young Arrest (Dechat et al. 2000Dechat T. Vlcek S. Foisner R. J. Struct. Biol. 2000; 129: 335-345Crossref PubMed Scopus (124) Google Scholar). The functions of these proteins are poorly understood; however their attachment to lamins, and in many cases to chromatin, suggest that they might link chromatin and lamins to the membrane. Nonmembrane proteins such as LAP2α could act as “free agents,” linking chromatin to lamins within the nuclear interior (Dechat et al. 2000Dechat T. Vlcek S. Foisner R. J. Struct. Biol. 2000; 129: 335-345Crossref PubMed Scopus (124) Google Scholar). LBR, a classic lamin binding protein, also binds to Hp1, a protein involved in the repression of gene expression (Ye et al. 1997Ye Q. Callebaut I. Pezhman A. Courvalin J.-C. Worman H.J. J. Biol. Chem. 1997; 272: 14983-14989Crossref PubMed Scopus (265) Google Scholar), suggesting that LBR might influence gene expression. There is a new hormonally regulated atypical P-type ATPase with nine transmembrane domains that localizes specifically to the nuclear inner membrane (Mansharamani et al. 2001Mansharamani M. Hewetson A. Chilton B.S. J. Biol. Chem. 2001; 276: 3641-3649Crossref PubMed Scopus (35) Google Scholar). This protein, named Ring Finger Binding Protein (RFBP), resembles a type IV phospholipid pump but lacks a domain required for pump activity. RFBP interacts directly with a RUSH protein, which is related to SWI/SNF transcription factors that remodel chromatin (see Mansharamani et al. 2001Mansharamani M. Hewetson A. Chilton B.S. J. Biol. Chem. 2001; 276: 3641-3649Crossref PubMed Scopus (35) Google Scholar). RFBP is the first nuclear membrane protein known to interact directly with a potential chromatin-remodeling partner. These findings suggest that RFBP and LBR (and other nuclear membrane proteins?) provide binding sites for proteins that regulate transcriptional access to chromatin. Alternatively, RFBP or LBR might play active roles in chromatin structure or transcription. The idea of a lamin-dependent infrastructure for gene regulation is supported by evidence that lamins interact with transcription factors, most notably Rb (retinoblastoma; see Cohen et al. 2001Cohen, M., Lee, K.K., Wilson, K.L., and Gruenbaum, Y. (2001). Trends Biochem. Sci. 26, 41–47.Google Scholar). A subset of nuclear membrane proteins, including LAP2, emerin and MAN1, belong to the newly defined “LEM domain” family (Lin et al. 2000Lin F. Blake D.L. Callebaut I. Skerjanc I.S. McBurney M.W. Pauline-Levasseur M. Worman H.J. J. Biol. Chem. 2000; 275: 4840-4847Crossref PubMed Scopus (264) Google Scholar). The LEM domain is a distinct ∼43-residue motif, which in LAP2 mediates binding to a ubiquitous, highly conserved and novel DNA-bridging protein named BAF (Furukawa 1999Furukawa K. J. Cell Sci. 1999; 112: 2485-2492Crossref PubMed Google Scholar). The function of BAF is unknown, but it is essential in C. elegans (Zheng et al. 2000Zheng R. Ghirlando R. Lee M.S. Mizuuchi K. Krause M. Craigie R. Proc. Natl. Acad. Sci. USA. 2000; 97: 8997-9002Crossref PubMed Scopus (180) Google Scholar). The loss of emerin, which binds A-type lamins (see Cohen et al. 2001Cohen, M., Lee, K.K., Wilson, K.L., and Gruenbaum, Y. (2001). Trends Biochem. Sci. 26, 41–47.Google Scholar), causes the X-linked recessive form of EDMD (Bione et al. 1994Bione S. Maestrini E. Rivella S. Mancini M. Regis S. Romeo G. Toniolo D. Nat. Genet. 1994; 8: 323-327Crossref PubMed Scopus (730) Google Scholar). Thus, “laminopathies” can arise from mutations in proteins that bind to lamins, as well as lamins themselves. Thus, disease models for EDMD must account for the functions of both emerin and A-type lamins. If LEM proteins (e.g., emerin) also interact with transcription factors, then disease may spring from defective expression of specific genes. Lamins are clearly polyfunctional, having both essential (housekeeping) and nonessential (cell-type-specific) roles. Essential roles are those required for cell survival, such as structural roles in DNA replication, chromosome segregation, transcription, and nuclear integrity. Mutations that knock out essential lamin functions would be lethal during embryogenesis, and would not cause disease. Dramatically, LMNA knockout mice demonstrate that all A-type lamins are nonessential, in the strictest sense of the word, since these mice are born normal and live at least three weeks after birth (Sullivan et al. 1999Sullivan T. Escalente-Alcalde D. Bhatt H. Anver M. Bhat N. Nagashima K. Stewart C.L. Burke B. J. Cell Biol. 1999; 147: 913-919Crossref PubMed Scopus (915) Google Scholar). It will be interesting to learn the knockout and mutant phenotypes for B-type lamins. Even if B-type lamins are essential, mild mutations might selectively disrupt cell-type-specific roles that will help us understand their full range of functions. The lipodystrophy mutations strongly support the idea that lamins can have cell-type-specific functions. Mutations that cause lipodystrophy are restricted to two small regions of LMNA, and the most frequent disease alleles are missense mutations at Arg482. When one such mutation was introduced into lamin A and expressed in cells, the mutant lamin protein behaved normally; it localized at the nuclear envelope, bound to emerin, and did not disrupt the endogenous lamina (Holt et al. 2001Holt, I.H., Clements, L., Manilal, S., Brown, S.C., and Morris, G.E. (2001). Eur. J. Hum. Genet., in press.Google Scholar). This normal behavior, coupled with its dominant tissue-specific disease phenotype, is most simply explained by models in which the mutant lamin interferes with a specific signaling or gene expression event. The nuclear lamina mediates nuclear structure, chromatin organization, and chromosome segregation, and may have structural roles in transcription and the elongation phase of DNA replication. The lamina also retains nuclear membrane proteins, determines nuclear shape, and controls the spatial distribution of nuclear pore complexes. This astonishing array of functions is comparable in principle to the cytoskeleton. Regulated, dynamic changes in the oligomeric or polymeric state of lamins may be crucial for particular interactions (e.g., with distinct components of the DNA replication machinery). In other words, lamins might assemble into different structures for different jobs. Based on current knowledge, we can make at least two predictions about disease mechanisms. First, each disease may have a unique mechanism, depending on which aspect(s) of lamina function is disrupted (e.g., mechanical stability versus gene expression). Second, the pathophysiology may not reflect the primary molecular defect per se (e.g., lamin disorganization). Instead, disease may arise from the downstream effects on chromatin structure or gene expression that are caused by lamin disorganization, failure to provide attachment sites for transcriptional regulators, or reduced binding affinity for other essential partners. These “downstream effects” may be responsible for the majority of clinical symptoms and tissue-specific pathology in human laminopathies. By dissecting the molecular mechanism of each laminopathy, we will gain unprecedented insight into the inner workings of the nucleus, and perhaps understand how to alleviate disease symptoms.