Nuclear lamina

The nucleus is the innermost sanctuary of every eukaryotic cell. It is encapsulated by the nuclear envelope, a double membrane structure consisting in the outer and inner nuclear membranes. Transport between the cytoplasm and the nucleus is facilitated by nuclear pore complexes (NPC), multimeric structures that span both layers of the nuclear envelope. The architectural integrity of the nucleus is maintained by the nuclear lamina, a thick (20–50 nm) proteinaceous meshwork that consists in type V intermediate filament proteins: the A- and B-type lamins. Expression of A-type lamins is limited to most somatic lineages, whereas B-type lamins are expressed in pluripotent stem cells and in their differentiated progeny.

A- and B-type lamins have been identified only in metazoans, a taxon including all multicellular animals arising around 700–800 million years ago (Peter et al. 2012; Melcer et al. 2007). Although plants harbour a nuclear lamina that is similar in its organisation and structure to the lamina found in multicellular animals, it is debatable whether its components are lamin analogues (Ciska and Moreno Diaz de la Espina 2013). Nevertheless, even in unicellular eukaryotes such as Dictyostelium or Trypanosomes, a lamina-type structure lines the nuclear periphery, highlighting the importance of this structural component in eukaryotic nuclei (Krüger et al. 2012; Batsios et al. 2012; DuBois et al. 2012). In addition to supporting the structural integrity of the nucleus, the components of the nuclear lamina play fundamental roles in sustaining cellular physiology.

A-type lamins

Lamin A, lamin AΔ10, lamin C and lamin C2 are all generated by alternative splicing of a single transcript from the LMNA gene. A-type lamins are absent in undifferentiated pluripotent stem cells (Constantinescu et al. 2006) but are ubiquitously expressed in most somatic lineages, in particular those of mesenchymal origin (Röber et al. 1989; Broers et al. 1997). Exceptions to this rule are the various neuronal and hematopoietic lineages in which A-type lamins can be expressed at low, or undetectable, levels (Röber et al. 1990; Jung et al. 2012). Lamin AΔ10 has been detected at low levels in cancer cells (Machiels et al. 1996), whereas the expression of lamin C2 is restricted to germ cells (Furukawa et al. 1994). In human fibroblasts, both lamin A and C are expressed at roughly equal levels but the relative amount of each isoform can vary greatly between tissues. For instance, neuronal lineages express little to no lamin A, whereas lamin C can readily be detected (Jung et al. 2012).

Lamin A (but not lamin C) undergoes extensive post-translational processing: the C-terminal CaaX motif of lamin A is farnesylated, cleaved, methylated and finally cleaved again to form the mature form of lamin A. Deletion of the protease involved in lamin A processing or mutations impairing the cleavage site can lead to the accumulation of permanently farnesylated, uncleaved pre-lamin A, and have been associated with restrictive dermopathy (RD) and Hutchinson-Gilford progeria syndrome (HGPS), respectively (see below).

B-type lamins

B-type lamins are subjected to similar post-translational processing. Both lamin B1 and B2 are farnesylated and cleaved at the –aaX motif but, in contrast to lamin A, B-type lamins undergo no further cleavage. Furthermore, lamin B1 and B2 are expressed in all cell types including embryonic stem cells (ESC), whereas lamin A/C expression is restricted to somatic lineages. Lamin B3 is a minor splice variant of lamin B2 and is exclusively expressed in male germ cells (Furukawa and Hotta 1993). B-type lamins are involved in various cellular processes, including DNA replication (Moir et al. 1994), cell cycle progression, chromatin remodelling and chromosome organisation (Moir et al. 1994; Guelen et al. 2008; Solovei et al. 2013), mitotic spindle assembly (Tsai et al. 2006; Kim et al. 2011) and, as most recently found, senescence and ageing (Dreesen et al. 2013a, 2013b; Shimi et al. 2011; Freund et al. 2012; Shah et al. 2013; Sadaie et al. 2013).

Diseases associated with mutations in nuclear lamina proteins

The number of known mutations in components of the nuclear lamina and lamina-associated factors is currently well over 400, resulting in a multitude of different diseases collectively termed laminopathies. Although partial overlaps occur in the clinical symptoms of the various laminopathies, they are extremely heterogeneous. This is particularly striking as many of them are caused by single point mutations in the LMNA gene. Mutations in LMNA can cause various muscular dystrophies (Emery-Dreifuss muscular dystrophy [EDMD], limb girdle muscular dystrophy [LGMD]), as well as lipodystrophy (familial partial lipodystrophy [FPLD]), dilated cardiomyopathy (DCM), neuropathy (Charcot-Marie-Tooth [CMT], autosomal dominant leukodystrophy [ADLD]), skin pathology (restrictive dermopathy [RD]), bone disease (mandibuloacral dysplasia [MAD]) and accelerated ageing/progeroid syndromes (Hutchinson Gilford progeria syndrome [HGPS] and atypical Werner syndrome) (Worman and Bonne 2007; Burke and Stewart 2014).

Mutations in LMNA: HGPS

Arguably, one of the best-studied laminopathy is the early-onset accelerated ageing syndrome HGPS, initially described by Jonathan Hutchinson and Hastings Gilford in 1886-87. Children with progeria appear normal at birth but start to develop symptoms including thinning of the skin and alopecia after 2–3 years of age, and die in their mid-teens because of cardiovascular failure (De Sandre-Giovannoli et al. 2003; Eriksson et al. 2003). HGPS is caused by a de novo autosomal dominant mutation in LMNA (c.1824C→T) giving rise to an aberrantly spliced, truncated form of lamin A, called progerin. At the cellular level, the most obvious phenotype of HGPS is the presence of morphologically abnormal nuclei (Fig. 1; Goldman et al. 2004; Eriksson et al. 2003). Nuclear architecture defects can be corrected by introducing modified oligonucleotides that target the activated cryptic splice site, prevent aberrant splicing and thereby suppress expression of the mutant protein (Scaffidi and Misteli 2005). Nuclear shape can also be restored by treating HGPS-derived cells with farnesyltransferase inhibitors (FTIs), drugs that inhibit lamin A and progerin farnesylation (Capell et al. 2005; Toth et al. 2005). FTIs were used in the first clinical trial for progeria patients and some beneficial effects have been reported (Couzin-Frankel 2012; Gordon et al. 2014).

Fig. 1
figure 1

Three-dimensional rendering of the nuclear lamina in (a) wild-type and (b) HGPS fibroblast nuclei by three-dimensional structured-illumination super-resolution microscopy. Staining: lamin A/C antibody. Bar 6 μm

Full size image

Recent studies have demonstrated that the inhibition of the N-acetyltransferase activity of NAT10 towards cytoplasmic microtubules, by a small molecule called “remodelin”, is also able to restore nuclear shape in HGPS or lamin-A-depleted cells. In addition, remodelin appears to improve the proliferation of HGPS cells (Larrieu et al. 2014). Interestingly, cells treated with both remodelin and FTI do not exhibit additional improvements in comparison with cells treated with either remodelin or FTI alone. This suggests that remodelin and FTIs do not act synergistically, but rather target a common pathway. In wild-type cells, FTIs have also been shown to affect microtubule dynamics and to cause nuclear abnormalities that can be prevented in the presence of remodelin (Suzuki et al. 1998; Verstraeten et al. 2011; Larrieu et al. 2014). Thus, both FTI and NAT10-modifying compounds might share parts of their mechanism of action with respect to nuclear shape.

Nuclear abnormalities are certainly a major phenotype of HGPS but there is more to progerin than meets the eye. Shortly after the HGPS mutation was identified, fibroblasts from HGPS patients were shown to exhibit increased DNA damage (Liu et al. 2005, 2006; Musich et al. 2009). Although treatment with FTIs or remodelin improves the aberrant nuclear shape of HGPS fibroblasts, only remodelin reduces DNA damage. These results suggest that nuclear morphology defects and DNA damage are independent phenotypes arising from progerin expression (Liu et al. 2005, 2006; Musich et al. 2009). Indeed, consistent with the persistent activation of DNA damage checkpoints, progeric fibroblasts exhibit a limited proliferative capacity and have significantly shorter telomeres than age-matched controls (Decker et al. 2009; Allsopp et al. 1992). In addition, the ectopic expression of progerin in wild-type fibroblasts inhibits their proliferation and triggers premature senescence (Benson et al. 2010). However, progerin-induced proliferative inhibition is alleviated by telomerase activation and, to some extent, by the inactivation of the p53 pathway (Kudlow et al. 2008). These results have provided the first evidence that progerin directly or indirectly damages telomeres, thereby activating p53, in a manner that can be alleviated by telomerase. More recently, Benson and colleagues extended these results by showing that the DNA damage foci in progerin-expressing cells co-localise with human TRF1, a component of the telomere-associated shelterin complex (Benson et al. 2010). Telomere-specific DNA damage is particularly detrimental for cells as it cannot be repaired by conventional DNA repair pathways and thus, triggers permanent growth arrest.

What is the physiological relevance of these findings and how does the accelerated ageing in HGPS patients relate to normal ageing? Increased DNA damage is a hallmark of cells undergoing senescence because of shortened or deprotected telomeres and shortened telomeres have been associated with human ageing (d’Adda di Fagagna et al. 2003; Takai et al. 2003; Canela et al. 2007). Nevertheless, one difference is that during normal ageing, telomere dysfunction is mainly a consequence of the end replication problem, which might also be enhanced by DNA damaging agents such as oxidative stress (Wang et al. 2009); whereas in HGPS, telomeres might be damaged more directly by progerin (Benson et al. 2010). However, the precise mechanism by which telomeres are damaged and whether only a subset of telomeres, such as those located at the nuclear periphery, are affected still needs to be established.

Mutations in lamin A versus lamin C

Although the HGPS mutation (c.1824C→T) is found exclusively in the lamin A splice isoform of LMNA, most other progeroid mutations (c.412G→ A, c.428C→T, c.433G→A, c.1583C→T, c.1619 T→C, c.1626G→C) and mutations involved in muscular dystrophies (EDMD, LGMD), lipodystrophy (FPLD), DCM, neuropathy (CMT), restrictive dermopathy (RD) and bone disease (MAD) simultaneously affect both lamin A and lamin C splice variants (Fig. 2). The subcellular localisation and similar domain structure of lamins A and C suggest that mutations in the LMNA locus affect cell physiology by equally altering both lamin A and C. Various mouse models have indeed shown that A-type lamin isoforms can behave independently of each other and fulfil similar functions (Fong et al. 2006). The finding that mice expressing exclusively lamin A or lamin C are disease-free as compared with Lmna knockout mice indicates a high level of functional redundancy (Sullivan et al. 1999; Kim and Zheng 2013; Kubben et al. 2011). However, LMNA mutations causing EDMD, DCM or lipodystrophy have been demonstrated to give rise to different effects when expressed in each isoform, including differential binding to other lamina components (Motsch et al. 2005; Sylvius et al. 2008). In particular, three EDMD-causing mutations (c.448A→C, c.1580G→C, c.1589T→C) disrupt lamin A binding to emerin, without affecting the association of emerin with lamin C (Motsch et al. 2005). Similarly, introducing the FPLD c.1444C→T mutation into lamin C impairs its localisation to the nuclear lamina, whereas the introduction of the same mutation into lamin A causes a much milder phenotype (Broers et al. 2005). Lastly, a decrease of lamin A protein abundance without affecting lamin C is also seen in the central nervous system in which only lamin A is specifically downregulated by miR-9 microRNA (Jung et al. 2012). Thus, despite an obvious functional redundancy between the two A-type lamins, mutations that affect both protein isoforms can impair cell function in a differential and/or synergistic manner.

Fig. 2
figure 2

Spatial distribution of LMNA mutations along A-type lamin cDNA. Representation of reported LMNA mutations (source: Universal Mutation Database LMNA). Mutations are represented by their position on the LMNA cDNA and categorised according to their phenotype(s): DCM (dilated cardiomyopathy), EDMD (Emery-Dreifuss muscular dystrophy), CMT (Charcot-Marie-Tooth), FPLD (familial partial lipodystrophy), LGMD (limb girdle muscular dystrophy), progeroid (HGPS and atypical Werner syndrome), MAD (mandibuloacral dysplasia), RD (restrictive dermopathy). A-type lamin protein domains are indicated: coil 1A (green), coil 1B (light blue), coil 2 (dark blue), nuclear localization signal (pink), lamin A-specific N-terminus (grey)

Full size image

Diseases associated with perturbations of B-type lamins

In contrast to lamin A, no point mutations in lamin B1 have been linked to any diseases. However, elevated levels of lamin B1 have been observed in cells from patients with ataxia telangiectasia (AT) and adult-onset autosomal dominant leukodystrophy (ADLD), diseases whose clinical symptoms include neurological defects. Whilst the underlying cause of the elevated lamin B1 levels in AT remains unclear, ADLD is caused by a duplication of the lamin B1 locus and leads to symmetrical demyelination of the central nervous system. Similar to patients with multiple sclerosis (MS), ADLD patients progressively lose fine motor skills and suffer from autonomic symptoms including bowel/bladder dysfunction, orthostatic hypotension and male impotence. However, contrasting the neurodegenerative phenotype of MS, demyelination in ADLD is symmetrical and not associated with brain inflammation and loss of oligodendrocytes.

In the past two years, several in vitro and in vivo models have been used to investigate the way that elevated levels of lamin B1 affect cellular physiology. Heng et al. (2013) generated a bacterial artificial chromosome (BAC)-based transgenic ADLD mouse model by expressing lamin B1 under the control of its endogenous promoter. These lamin B1BAC mice exhibit several pathophysiological features of ADLD, including impaired cognitive function and age-dependent motor deficits (Heng et al. 2013). Ultrastructural analysis of 24-month-old ADLD mice revealed aberrant myelin formation, demyelination and axonal degeneration. Consistent with the non-inflammatory phenotype of ADLD patients, lamin B1BAC mice show no evidence of microglia activation or reactive astrocytes. Nevertheless, how do elevated levels of lamin B1 result in myelin loss? Analysis of the protein composition between lamin B1BAC and wild-type mice has revealed that lamin B1BAC mice exhibit a significant downregulation of proteolipid protein (PLP). PLP is a major component of the myelin sheet and has previously been implicated in other myelin-related diseases. In addition, elevated levels of lamin B1 in oligodendrocytes result in the transcriptional repression of the genes involved in myelin biosynthesis (myelin-basic protein, proteolipid protein and myelin oligodendrocyte glycoprotein; Lin and Fu 2009).

Although these results provide a link between lamin B1 overexpression and myelin abnormalities, the precise mechanism by which lamin B1 overexpression leads to reduced myelin remains unclear. Lamins have been shown to interact with DNA (Kind et al. 2013; Guelen et al. 2008) and, to some extent, might regulate gene expression (Finlan et al. 2008; Reddy et al. 2008; Shevelyov et al. 2009). Thus, elevated levels of lamin B1 might directly affect the transcriptional regulation of PLP.

Another possibility is that lamin B1 overexpression triggers a cellular program that in turn leads to myelin destabilisation. We and others have developed cell-based in vitro models to study the consequences of lamin B1 overexpression on cell proliferation and function (Barascu et al. 2012; Dreesen et al. 2013a, 2013b). However, in human fibroblasts, a 2– to 3-fold overexpression of lamin B1 only results in a moderate proliferation defect. These results have prompted us to investigate the reasons why human fibroblasts are relatively resistant to lamin B1 overexpression and why lamin B1 duplication in ADLD preferentially affects the central nervous system. We hypothesised that differences in the composition of the nuclear lamina in fibroblasts versus neuronal lineages render the latter more susceptible to aberrant lamin B1 levels. One such difference is that neuronal lineages express dramatically lower levels of lamin A than fibroblasts (Lehner et al. 1987; Röber et al. 1989; Zhang et al. 2011). We therefore increased lamin B1 in fibroblasts with reduced (50 % reduction) levels of lamin A/C. Strikingly, lamin B1↑lamin A/C↓ cells exhibited a pronounced proliferation defect, arrested at the G0/G1 stage of the cell cycle and stained positive for senescence-associated-β-gal activity. In addition, lamin B1↑lamin A/C↓ cells exhibited 53BP-1 DNA damage foci that were associated with telomeres (Dreesen et al. 2013a, 2013b). Reminiscent of the situation in progerin-expressing fibroblasts, the introduction of telomerase prevented the accumulation of telomeric DNA damage foci and restored the proliferation of lamin B1↑lamin A/C↓ cells (Dreesen et al. 2013a, b). Taken together, these results suggest that a reduction in lamin A/C levels dramatically potentiates the phenotypic consequences of lamin B1 overexpression. The question remains as to how perturbations in the nuclear lamina cause telomeric DNA damage.

Nuclear lamina remodelling

The nuclear lamina was at first thought to be a static meshwork but is now considered to be a highly plastic and dynamic structure. The composition of the nuclear lamina and the stoichiometry of lamina components varies between tissues and changes during development, the cell cycle and in different cell fates (Broers et al. 1997; Swift et al. 2013).

Remodelling during embryonic development and differentiation

Pioneering studies in the 1980s revealed that the composition of the nuclear lamina undergoes profound changes during embryonic development in chicken and mouse. Nigg and colleagues demonstrated that early chicken embryos contained substantial amounts of B-type lamins, whereas A-type lamins were absent and accumulated only during later stages of development (Lehner et al. 1987). A-type lamins appeared after ~8 days in the ectoplacental cone, the tissue that eventually forms the placenta and after 9–11 days in the embryo proper.

The differential expression of lamins during embryonic development can also be recapitulated at the cellular level in vitro: pluripotent ESC and undifferentiated teratocarcinoma stem cells exclusively express B-type lamins, whereas the expression of A-type lamins is confined to their differentiated progeny. Conversely, the reprogramming of somatic cells into induced pluripotent stem cells (iPSC) or even the exposure of somatic cells to ESC extracts results in the downregulation of lamin A/C (Zhang et al. 2011; Liu et al. 2011; Bru et al. 2008).

The differential expression of lamins during embryonic development in vivo and during differentiation in vitro raises the question as to whether the lamina is actively involved in regulating gene expression during development. At least in the mouse, neither A- nor B-type lamins appear to be necessary for early embryonic development: Lmna-/- mice develop to term and only exhibit impaired growth and muscular dystrophy after birth (Sullivan et al. 1999). Similarly, mice lacking both lmnb1 and lmnb2 also develop to term but die shortly thereafter. Perhaps most surprising is the fact that E12.5 and E18 embryos from Lmnb1+/- Lmnb2+/- intercrosses reveal a normal Mendelian distribution and that all internal organs form properly. In agreement with these findings, mESC derived from Lmnb1/2-/- mice retain their pluriopotency markers and differentiate into trophectoderm with a similar efficiency as wild-type mESC (Kim et al. 2011). Taken together, these results suggest that B-type lamins are not actively involved in regulating transcriptional programs during ES cell differentiation.

Haematopoietic system

Most cells of the haematopoietic system either do not express A-type lamins or express it at extremely low levels (Röber et al. 1990; Guilly et al. 1990). However, a recent study demonstrated that T-lymphocytes, which are generally devoid of A-type lamins, show a transient increase in lamin A/C expression upon T-cell activation (Gonzales-Granado et al. 2014). These results might be physiologically relevant as elevated levels of lamin A enhance T-cell activation both in vitro and in vivo, whereas lamin A-deficiency reduces T-cell activation. Do these results have implications for laminopathies? Children with progeria appear to have a normal haematopoietic system and respond normally to various infections. A detailed analysis of telomere length in haematopoietic lineages including T-cells, B-cells, natural killer cells and granulocytes, revealed that the median telomere length in three out of four HGPS patients is comparable with that of age-matched controls (Decker et al. 2009). In contrast, fibroblasts from the same patients exhibit dramatically shortened telomeres. Is it possible that the reactivation of lamin A and progerin in activated T-cells from HGPS patients is too transient or at too low a level to damage telomeres? The reactivation of lamin A in activated T-cells and the apparent resistance of the immune system to progerin is intriguing and would benefit from further investigation.

Central nervous system

In most adult tissues, lamins A and C are found in roughly equal amounts. An exception to this rule is the central nervous system: whilst lamin C is expressed in most cell types of the adult mouse brain, lamin A levels remain low or undetectable (Jung et al. 2012). As lamin A and lamin C are alternatively spliced isoforms of the LMNA gene, what could account for their dramatically different expression levels? Jung and colleagues found that the 3’untranslated region of pre-lamin A mRNA but not lamin C mRNA, contains a putative binding site for miR-9, a microRNA that is specifically expressed in the brain and is a key player in neural development (Leucht et al. 2008). Subsequent experiments by various groups have shown that miR-9 expression in human fibroblasts, HeLa cells and iPSC-derived mesenchymal stem cells leads to reduced levels of lamin A but not lamin C. Importantly, the miR-9-dependent removal of lamin A from neuronal tissues is of critical clinical importance to patients with HGPS: miR-9 prevents the accumulation of mutant lamin A (progerin) thereby protecting them from complications in the central nervous system (Nissan et al. 2012; Jung et al. 2012, 2014).

The lamina composition in the brain also appears to change over time. In mice, the temporal expression of lamin B1 is developmentally regulated and changes during brain development: lamin B1 protein and mRNA levels are highest at birth and subsequently decline throughout adult life (Lin and Fu 2009). Similarly, reduced lamin B1 levels can be observed in rats during oligodendrocyte maturation in vitro (Dugas et al. 2006). Interestingly, lamin B1 levels are inversely correlated with myelin-specific proteins, including myelin basic protein and myelin-associated glycoprotein. Therefore, the orchestration of the onset and levels of lamin B1 expression is important as excessive production of lamin B1 (because of genomic duplication of the LMNB1 locus) is associated with ADLD and demyelination in the central nervous system (Lin and Fu 2009).

Tissue rigidity

The central nervous system not only has a highly characteristic ratio of lamin A abundance compared with that of lamin C, it also has one of the lowest ratios of A-type to B-type lamins. Swift et al. (2013) showed that A:B lamin stoichiometry changes across tissues and is correlated with their elasticity. Thus, tissues such as cartilage, bone, muscle or heart have a high A:B ratio (high A-type lamin expression compared with B-type lamins), whereas softer tissues such as liver, kidney and brain are characterised by a low A:B ratio (low A-type lamin expression compared with B-type lamins). This variation of A:B-type lamin abundance is mainly dependent on the modulation of A-type lamins, as the levels of B-type lamins remain relatively constant across tissues. What could explain the diverse A:B-type ratio in various tissues and how is it regulated? In a series of papers, the Discher group demonstrated that the lamina composition dynamically changes in order to adapt to various tissue environments. This is facilitated by the phosphorylation of lamin A, which in turn changes the physical properties of the lamina. Lamin A phosphorylation occurs when cells are grown on soft matrices, whereas lamin A/C levels are elevated with increasing stiffness of the matrix. Thus, both lamin A/C levels and its phosphorylation status modify the mechanical properties of the nucleus and facilitate cell growth in various tissue environments (Swift et al. 2013; Buxboim et al. 2014).

Nuclear lamina dissolution during mitosis

The most drastic remodelling of the nuclear lamina occurs during cell division, when the nuclear envelope and the lamina are broken down (Güttinger et al. 2009). During cell division, A- and B-type lamin filaments are depolymerised following their phosphorylation by kinases including protein kinase C (PKC), Aurora A and polo-like kinase 1 (Gerace and Blobel 1980; Güttinger et al. 2009; Mall et al. 2012). Phosphorylation of A-type lamins induces their solubilisation and release, first into the nucleoplasm and, after nuclear envelope breakdown, into the cytosol. In contrast, phosphorylated B-type lamins remain associated with mitotic endoplasmic reticulum (ER) membranes (Georgatos et al. 1997). Reassembly of the nuclear lamina after mitosis starts by the binding of the ER tubules to chromatin (Anderson and Hetzer 2008), assembly of NPC prepores (Sheehan et al. 1988), remodelling of the ER tubules into flattened nuclear envelope patches and the binding of inner nuclear membrane proteins to chromatin (Anderson and Hetzer 2007). Dephosphorylation of lamins by protein phosphatases and inactivation of cyclin-dependent kinase 1 is required to reintegrate the nuclear lamina proteins into a new nuclear lamina; this will occur only once nuclear import has been restored (Newport et al. 1990).

Viruses take advantage of this mechanism to modulate the nuclear lamina structurally and to egress from the nucleus into the cytoplasm. Herpes simplex virus type 1 and the murine cytomegalovirus achieve this by recruiting the cellular PKC to the nuclear periphery (Park and Baines 2006). This in turn triggers phosphorylation of nuclear envelop proteins, leads to irregular lamina distribution and dissolution, and facilitates the exit of virions from the nucleus (Muranyi et al. 2002; Camozzi et al. 2008).

Lamina remodelling in cancer cells

The Papanicolaou smear test is routinely used to identify squamous cell carcinomas, one of the most common and malignant neoplasms among women. The test typically includes the search for nuclear abnormalities, such as enlarged and irregularly shaped nuclei (Bengtsson and Malm 2014). Such alterations of the nuclear lamina and the variable levels of A-type lamins have been recurrently observed in cancer cells, although their exact significance and relevance remain unclear (de Las Heras et al. 2013). Down-regulation of A-type lamins has been reported in gastrointestinal neoplasms (Moss et al. 1999), gastric carcinoma (Wu et al. 2009), colon (Belt et al. 2011), breast (Capo-chichi et al. 2011a, 2011b) and ovarian cancers (Capo-chichi et al. 2011a, 2011b), whereas others have shown the up-regulation of lamins in colorectal (Willis et al. 2009) and prostate cancer (Kong et al. 2012). Thus, although different lamin A expression levels can be correlated with the prognosis of various cancers, it remains unclear whether lamin A/C fluctuations are causal or simply a consequence of the widespread genomic abnormalities that occur in cancer cells. Some recent studies have suggested that lower lamin A levels confer a selective advantage to some cancer cells, as the decreased expression of lamin A improves nuclear deformability and cell migration through three-dimensional environments (Rowat et al. 2013; Wolf et al. 2013). In contrast, Kong et al. (2012) modulated lamin levels in prostate cancer cells and found that the overexpression of lamin A/C enhances cell proliferation, migration and invasion, suggesting a causal role of A-type lamins on cancer cell propagation. In conclusion, a clear consensus for the role of A-type lamins in cancer cells might be difficult to find: the diverse types of cancers might respond differently to aberrant lamin A levels, depending on the selective advantage (proliferation, invasiveness) that is required.

Nuclear lamina remodelling during cellular senescence

Cellular senescence is an irreversible cell cycle exit that can be triggered by various forms of cellular stress, including irreparable DNA damage, oxidative, mitotic or oncogenic stress. Senescent cells are characterised by the activation of p53 and Rb tumour suppressor pathways and can be found in premalignant tumours, including benign nevi. Senescent cells also accumulate with age in various tissues, thereby limiting their regenerative potential. The entry into senescence is associated with morphological changes, the expression of senescence-associated β-galactosidase activity (SA-β–gal) and changes in gene expression and chromatin organisation, including the generation of senescence-associated heterochromatin foci and the senescence-associated secretory phenotype (Rodier and Campisi 2011). In addition, the nuclear lamina remodels dramatically as cells senesce (Dreesen et al. 2013a, 2013b; Freund et al. 2012; Shimi et al. 2011). These changes include a reduction of lamin B1 and the lamina associated polypeptide 2α (LAP2α). The loss of lamin B1 levels is specific to senescent cells, whereas LAP2α levels also decline in quiescent cells. Loss of lamin B1 can additionally be used to identify senescent cells in vivo: lamin B1 levels decline in mice treated with a senescence-inducing dose of ionising radiation (Freund et al. 2012), in a mouse model for the accelerated ageing syndrome progeria (McKenna et al. 2014) and during the chronological ageing of human skin (Dreesen et al. 2013a, 2013b). In addition to the loss of lamin B1, a reduced expression of lamin B1 receptor and a relocalisation of lamin A/C from the perinuclear region suggest that the nuclear envelope is structurally compromised in senescent cells (Ivanov et al. 2013). Taken together, these results demonstrate that the nuclear envelope is dramatically remodelled in senescent cells and that loss of lamin B1 can be used as a marker to identify senescent cells in vitro and in vivo.

These data raise the question as to whether the down-regulation of lamin B1 causes the widespread chromatin changes observed in senescent cells, i.e., whether the reduction of lamin B1 is a cause or a consequence of cellular senescence. Several papers have addressed this question with differing results. Whilst some groups reported that small-hairpin-RNA-mediated downregulation of lamin B1 triggers cellular senescence and apoptosis (Harborth et al. 2001; Shimi et al. 2011; Shah et al. 2013), other studies suggested that merely reducing lamin B1 levels is not sufficient to trigger senescence (Dreesen et al. 2013a, 2013b; Sadaie et al. 2013). Undoubtedly, lamin B1 levels decline dramatically during senescence but the causality of this reduction on the senescence phenotype remains debated. Mice that lack lamin B1 and lamin B2 specifically in epidermal keratinocyte or hepatocytes do not show any overt problems in the development and maintenance of skin, or any defects in liver development, histology or function. Similarly, most organs of lamin B1 knockout mice develop to term and are apparently normal, despite the fact that these mice die shortly after birth and exhibit bone, lung and cranial abnormalities. These results demonstrate that lamin B1 is not essential for cell proliferation during organ development in mice. In contrast, when grown under in vitro cell culture conditions, lamin-B1-deficient mouse embryonic fibroblasts exhibit nuclear abnormalities, impaired proliferation and premature senescence. This therefore suggests that lamin B1 deficiency impairs cellular function(s), which in conjunction with additional stress, such as growth under sparse conditions or in vitro culture, triggers senescence (Vergnes et al. 2004; Dreesen et al. 2013a, 2013b).

Concluding remarks

Originally considered as a static and rigid piece of hardware, the nuclear lamina is now known as a highly plastic, modular and versatile part of the nucleus. The sheer number of the known different diseases and mutations linked to the nuclear lamina reveals the complexity of its cellular functions (Worman and Bonne 2007). In addition, as discussed here, the relative abundance of the various lamina components appears to play an essential role in controlling the tissue-specific nature of some laminopathies.

One of the cellular hallmarks of HGPS is the presence of abnormally shaped nuclei, coupled with the accumulation of a mutant lamin A (progerin) within the nuclear lamina. Despite progerin being expressed in nearly all somatic cells, it is a segmental ageing disorder in which the brain is not affected. A major breakthrough has been the discovery of the brain-specific expression of miR-9, which targets LMNA transcripts and therefore prevents progerin expression (Nissan et al. 2012; Jung et al. 2012, 2014). These results highlight the importance of nuclear lamina composition, the way that it varies among different tissues and the manner in which this can affect disease etiology. The peculiar lamina composition in the brain might also account for the phenotypic specificity of ADLD: our studies suggest that the lower abundance of A-type lamin renders the central nervous system particularly vulnerable to elevated levels of lamin B1 (Dreesen et al. 2013a, b). These investigations emphasise the essential role of the nuclear lamina during development and cell differentiation and the highly complex chain of events leading from a single point mutation to the disease phenotype. Importantly, they indicate that, whereas mutation dictates the starting point of this detrimental chain of events, the widely diverse composition of the nuclear lamina in each tissue might ultimately define the tissue-specific nature of laminopathies.