Adult stem cell maintenance and tissue regeneration in the ageing context: the role for A-type lamins as intrinsic modulators of ageing in adult stem cells and their niches

Abstract

Adult stem cells have been identified in most mammalian tissues of the adult body and are known to support the continuous repair and regeneration of tissues. A generalized decline in tissue regenerative responses associated with age is believed to result from a depletion and/or a loss of function of adult stem cells, which itself may be a driving cause of many age-related disease pathologies. Here we review the striking similarities between tissue phenotypes seen in many degenerative conditions associated with old age and those reported in age-related nuclear envelope disorders caused by mutations in the LMNA gene. The concept is beginning to emerge that nuclear filament proteins, A-type lamins, may act as signalling receptors in the nucleus required for receiving and/or transducing upstream cytosolic signals in a number of pathways central to adult stem cell maintenance as well as adaptive responses to stress. We propose that during ageing and in diseases caused by lamin A mutations, dysfunction of the A-type lamin stress-resistant signalling network in adult stem cells, their progenitors and/or stem cell niches leads to a loss of protection against growth-related stress. This in turn triggers an inappropriate activation or a complete failure of self-renewal pathways with the consequent initiation of stress-induced senescence. As such, A-type lamins should be regarded as intrinsic modulators of ageing within adult stem cells and their niches that are essential for survival to old age.

Introduction

Adult stem cell maintenance involves a fine balance between genetic and molecular cell mechanisms, external factors in the local and systemic niches in the body and multiple signalling pathways which interface in the regulation of adult stem cell homeostasis. Age-dependent decreases in tissue regenerative responses have been linked to a decreased number of adult stem cells, their dysfunction in self-renewal and lineage potential, and/or the inhibitory activity of the local and systemic factors in the aged stem cell niches (Conboy & Rando, 2005). Hence, gaining insights into the complex regulation of adult stem cell maintenance has been regarded as the key to understanding the processes of generalized tissue degeneration associated with old age. Nuclear filament proteins, A-type lamins, have been linked to a number of progeroid syndromes and degenerative adult-onset diseases, and are proposed to be regulators of adult mesenchymal stem cell (MSC) homeostasis (Hutchison & Worman, 2004; Gotzmann & Foisner, 2006). In this review, we will introduce some of the most important mechanisms critical in determining the effects of ageing on tissue regeneration and adult stem cell maintenance. Following from this, we will discuss some of the latest findings implicating the involvement of A-type lamins in the homeostasis of adult tissues, particularly in a number of signalling pathways central to adult stem cell maintenance and tissue regeneration as well as adaptive responses to stress. In so doing, we will put forward a hypothesis for A-type lamins as intrinsic modulators of ageing in adult stem cells and their niches and thus of organismal lifespan. Therefore, human diseases and animal models with lamin A mutations may offer important clues to understanding the maintenance of adult stem cells and provide links to the physiology of adult stem cell ageing.

Adult stem cells and their niches

The ability of stem cells to self-renew is crucial in the maintenance and repair of adult tissues throughout life as much as it is in development. During development, embryonic stem cells give rise to germ cells and to a wide range of other more specialized somatic stem cells that maintain tissues within the adult organism. Adult stem cells can replace cells within tissues that have either high turnover such as blood, vascular endothelium and epithelia of skin, intestine, and respiratory tract, or those that have low turnover but a high regenerative potential upon injury or disease such as skeletal muscle, liver, pancreas and bone (Rando, 2006). In adult organisms, stem cells have been predominantly found in the bone marrow but they have also been identified from an ever increasing array of tissues including skeletal muscle, adipose tissue, hair follicle, peripheral circulation, intestinal epithelium, myocardium, lung, breast, kidney and brain (Wagers & Weissman, 2004). Bone marrow stem cells include pluripotent haematopoietic stem cells (HSCs) and MSCs (Serakinci & Keith, 2006). HSCs can give rise to all different blood cell lineages such as the myeloid and lymphoid cell lineages whilst MSCs give rise to stromal cells which in turn give rise to multiple mesoderm-type cell lineages such as osteogenic, adipogenic, chondrogenic and myogenic lineages. In addition, MSCs have been known to give rise to non-mesoderm-type cell lineages including fibroblastic cells, neuronal-like cells, cardiomyocytes and vascular cells such as smooth muscle-like and endothelial-like precursor cells (Barry & Murphy, 2004).

The process of self-renewal allows a stem cell to replicate and thus generate either two (symmetric division in embryonic stem cells) or one (asymmetric division in germinal and adult stem cells) daughter stem cell whilst preserving their broad developmental potential (Molofsky et al. 2004). This potential refers to the competence of daughter stem cells to express each potential fate whilst at the same time inhibiting their lineage commitment and differentiation. The offspring of stem cells are termed progenitor cells and can be classified as either early or late according to the degree of their lineage commitment. Stem cells have varying degrees of developmental potential from the pluripotency of embryonic stem cells, which give rise to all cell types of the embryo proper to the multipotentiality (MSCs) and unipotentiality (endothelial progenitor cells) of adult stem cells, which can differentiate into either a subset of cell types or only the cell type specific to the tissue from which they are derived, respectively. However, recent evidence suggests that many ‘unipotential’ adult stem cells have the ability to differentiate or trans-differentiate into cell types other than their tissue of origin (Qian et al. 2007).

Stem cells normally reside in the stem cell microenvironment or niche, which can modulate how stem cells participate in tissue regeneration, maintenance and repair (Scadden, 2006). Stem cell niches are diverse and comprise a variety of cells, extracellular matrices and paracrine hormonal and endocrine signals, which all interact with each other. This important interplay between stem cells and their niches mediates appropriate responses to the metabolic needs of tissues and the organism as a whole, especially under conditions of physiological challenge such as increased levels of oxidative stress following inflammation or injury. As such, the stem cell niche can have a dual role in stem cell regulation. Whilst on the one hand it may act as a protective barrier against damaging stimuli that lead to stem cell depletion or excessive proliferation, on the other hand, it can negatively modify the external stimuli and alter the function of stem cells leading to disease pathologies. Interestingly, a hallmark of stem cell populations in young animals is their quiescent state. Yet, in older animals, an increased number of stem cells is found in the cell cycle (Morrison et al. 1996). The increased stem cell cycle kinetics observed in older animals is thought to be a consequence of chronic inflammation or infection that occurs with age. Therefore, maintaining a balance between stem cell activity and periods of quiescence is a feature of the functional niche (Moore & Lemischka, 2006).

Ageing of ‘disposable soma’ and loss of adult stem cell potential

The finite lifespan of adult somatic cells and their loss in the adult human body is balanced through the production of new daughter cells from adult (somatic) stem cells. In simple multicellular organisms such as Caenorhabditis elegans and Drosophila melanogaster, which are almost entirely made up of post-mitotic cells, the maximum lifespan of an organism as a whole is determined by the lifespan of their ‘disposable soma’. Interestingly, however, the midgut in D. melanogaster has recently been found to be actively replenished by adult intestinal stem cells (Micchelli & Perrimon, 2006; Ohlstein & Spradling, 2006). It has been speculated that the acquisition of adult stem cells during evolution in more complex organisms has resulted in a major extension of organismal lifespan and thus the role of adult stem cells primarily lies in the rejuvenation of ageing somatic tissues (Kamminga & de Haan, 2006). As such, ageing of an organism could be considered as a gradual loss of adult stem cell potential. How would this occur? One theory that successfully bridges both evolutionary and genetic perspectives on the ageing process, the ‘disposable soma’ theory, provides an unanticipated explanation. In nature, there is a trade-off between somatic maintenance and repair on the one hand and reproduction costs on the other so that those organisms that invest more in the former have longer lifespans than those that invest in the latter (Kirkwood, 1977). Bearing this in mind, it seems very probable that the more complex organisms have evolved to live longer lives by investing more in maintaining and repairing their ‘disposable soma’ through adult stem cell regeneration. Following on from this argument, given that stem cell hyper-activation has been linked to cancers (Pelicci, 2004), it is envisaged that short-lived organisms would show an increased incidence of germ-cell-derived tumours whilst long-lived organisms would show an increased incidence of tumours in a wide range of somatic tissues.

The ageing of organisms is characterized by declining tissue repair and regeneration in response to injury which is thought to be a driving cause of many age-related tissue pathologies (Rando, 2006). The cause of this decline seems to be tissue-specific and is believed to be a consequence of both intrinsic stem cell ageing and external alterations of both the local factors within the stem cell niche and the systemic factors within the systemic organ environment. The connection between the replicative potential of stem cells and tissue ageing is not fully understood. Whilst a substantial number of adult stem cells seem to be sustained in later life (Collins et al. 2007), their replicative activity and thus regenerative potential, at least in mouse models, has been shown to decline sharply with age (Schlessinger & Van Zant, 2001; Shefer et al. 2006). Moreover, recent studies performed in HSCs and muscle satellite cells of aged mice show that in tissues of high turnover and/or regenerative potential, the loss of the replicative ability of stem cells may not be the main cause of tissue degeneration. Apparently, it was not the loss of replicative ability that directly contributed towards a decrease in regenerative potential in these aged adult stem cell types but the latter was a consequence of external alterations within the aged local stem cell niches and systemic organ environments in which these cell types resided (Conboy et al. 2005). Namely, these authors show that parabiotic pairings of young and old mice significantly improved both muscle and liver regeneration in the aged mice by restoring the regenerative capacity of the pre-existing aged satellite cells and liver progenitors, respectively, through exposure to the systemic factors of young mice. In addition to such inhibitory effects of aged systemic niches, it was also found that the replicative ability and myogenic capacity of satellite cells isolated from an injured young muscle was inhibited when co-cultured with the aged myofibre, demonstrating the negative influence of local aged muscle niches on satellite cell regenerative capacity (Carlson & Conboy, 2007). Remarkably, the similarity in the inhibitory properties of systemic and local organ niches indicates that either the inhibitory factors produced by the local aged niches have circulatory/endocrine activity or that the age-specific systemic inhibitory factors become deposited in the old tissues.

Cellular senescence: a window into stem cell ageing

A number of inefficient intrinsic cellular processes can limit lifespan, including errors in DNA repair, failures in protection against oxidative damage and protein cross-linking, the accumulation of defective mitochondria, the extent of protein or cell turnover, and failures in tumour protection mechanisms (von Zglinicki, 2002; Wright & Shay, 2002). One model that explains the functional decline of various organ systems with increasing age is that the cells essential for tissue function and regeneration reach a finite lifespan termed ‘cellular senescence’ and as such accumulate in the organism. Although cellular senescence has been mainly studied as an in vitro cellular phenomenon concerning finite replicative lifespan (Hayflick & Moorhead, 1961), it is proposed and more recently demonstrated in vivo that senescent cells can account for over 15% of the cell populations in the mitotic tissues of ageing primates (Herbig et al. 2006; Jeyapalan et al. 2007). Premature senescence is also a hallmark of the so-called premature ageing syndromes or segmental progerias which show an early or exacerbated manifestation of one or a few aspects of ageing such as Hutchinson Gilford Progeria Syndrome (HGPS) and Werner Syndrome (WS) (Bridger & Kill, 2004; Kipling et al. 2004). It was proposed that premature ageing in HGPS individuals may be a consequence of premature stem cell exhaustion (Halaschek-Wiener & Brooks-Wilson, 2007). The premature ageing syndromes in humans and mice could therefore provide insights into the mechanisms of accelerated ageing in stem cell populations. Recently, it has been demonstrated that the Klotho mouse models of accelerated ageing show a decrease in stem cell number and an increase in progenitor stem cell senescence in various tissues and organs (Liu et al. 2007). The Klotho gene encodes a single-pass transmembrane protein but can also function as a humoral factor when its extracellular domain becomes cleaved and secreted, in which case it binds to the cell surface receptors and inhibits both the insulin/IGF-1 (insulin growth factor-1) and Wnt (Wingless/INT) signalling pathways.

Senescent cells accumulate at many tissue sites associated with age-related pathologies such as atherosclerotic lesions, skin ulcers, arthritic joints and hyper-proliferative regions in the prostate and liver (Faragher & Kipling, 1998). How may senescent cells affect stem cells and their niches? Stem cells are prime targets of carcinogenesis and the ability of stem cells to undergo senescence is considered an important tumour-suppressive mechanism which prevents the accumulation of oncogenic mutations in the self-renewing compartments (Beausejour & Campisi, 2006). However, the accumulation of senescent cells in tissues over time would not only deplete the pool of mitotically competent stem and progenitor cells but would also change the surrounding stem cell niche and thus compromise tissue repair and renewal (Campisi, 2001, 2005). For example, interstitial fibroblasts synthesize stromal support for all renewable epithelial tissues. Senescent fibroblasts secrete epithelial growth factors, inflammatory cytokines and matrix metalloproteinases that alter the tissue structure and cause local inflammation (Campisi, 2005). This is thought to be an important causative factor in the altered tissue remodelling seen in atherosclerosis and hyperplastic epithelial lesions. As a result, senescent cells may be an example of ‘antagonistic pleiotropy’ involved in inhibiting stem cell hyper-proliferation as the protective tumour-suppressive mechanism in the young organisms whilst fostering a microenvironment that promotes age-associated diseases and/or neoplastic transformation in the old organisms (Krtolica et al. 2001).

Age-associated tissue degeneration

Ageing is associated with a number of degenerative disorders and inflammatory conditions including sarcopenia, osteoarthritis and osteopenia, lipo-atrophy with insulin resistance and diabetes, and cardiovascular conditions such as hypertension, atherosclerosis and cardio-myopathy (see Fig. 1). These disorders are believed to reflect an altered proliferative homeostasis between cell loss and cell replacement through adult stem cells in aged tissues and include a number of processes ranging from tissue fibrosis, sclerosis and hyperplasia to atrophy (Fehrer & Lepperdinger, 2005; Martin, 2007). Most commonly, the atrophy of specialized differentiated cells within aged tissues is associated with either a compensatory hyperplasia of the same cell types or an interstitial fibrosis in tissues such as heart and muscle. Whilst fibrotic tissue is critical in rejoining broken tissue during muscle or heart regeneration, its engagement post-injury interferes with tissue regenerative potential and prevents proper innervation (Mourkioti & Rosenthal, 2005).

Fig. 1.

Schematic diagram drawing parallels between mesenchymal tissues affected in a number of age-associated degenerative disorders (labelled in pink) and human diseases associated with mutations in the LMNA gene (labelled in blue). The full names for LMNA-associated diseases are given in the body of the article.

In aged skin, there is a reduction of both epidermal and dermal cell compartments leading to age-associated dermo-atrophy. As a result of age-associated alterations in these compartments, a characteristic wrinkling of skin occurs, reflecting changes in the collagen and elastin fibre matrix as well as a loss of hydration (Boukamp, 2005). Environmental stressors such as exposure to cigarette smoke and UV radiation also lead to alterations in skin connective tissue and accumulation of elastotic material in the dermis (Uitto & Bernstein, 1998). There is a decline in skin dermal fibroblast proliferation, motility and adhesion properties whilst their contractile behaviour increases with age. Cutaneous wound healing is impaired in terms of inflammatory profile, proliferation and remodelling leading to the chronic wound healing states associated with conditions such as diabetic ulcers, pressure sores and venous stasis ulcers (Ashcroft et al. 1995). A decreased number of hair follicle epidermal stem cells is thought to be involved in this wound closure defect (Ito et al. 2005).

Visceral obesity with a loss of subcutaneous adipose tissue and insulin resistance (i.e. metabolic syndrome) is another characteristic of old age (Karagiannides et al. 2001). Ageing is characterized by tissue damage through cytotoxic lipids and altered fatty acid handling by adipose cells (Guo et al. 2007). Adipose progenitor stem cells, the pre-adipocytes, are thought to play a crucial role in altered fat tissue function and age-associated lipo-atrophy (Kirkland & Dobson, 1997). The pre-adipocytes in aged rats fail to undergo adipogenesis in response to a fatty acid diet and show an increased tendency towards cell death. Yet, in aged tissues such as muscle and liver there is an abnormal accumulation of lipid droplets while adipose connective tissue is found to replace muscle fibres, heart tissue and bone matrix (Kirkland et al. 2002). Redistribution of lipid to extra-adipose sites with ageing could result from a loss of lipid storage capacity in adipose cells, altered fatty acid handling by non-adipose cells and/or de-differentiation of mesenchymal precursors into a partial adipocyte phenotype.

Senile osteoporosis and osteopenia are characterized by a progressive and irreversible decline in bone mass. Age-associated low bone density results from the lower activity of bone-producing osteoblasts as compared to bone-reabsorbing osteoclasts, the cell types involved in remodelling and maintenance of bone tissue (Hartmann, 2006). These cell types regulate bone homeostasis by secreting factors that regulate the activities of one another, which are believed to be altered in aged bone. Interestingly, a loss of bone density associated with osteoporosis and osteopenia is accompanied by an increase in bone marrow adipose tissue (Mueller & Glowacki, 2001). This is thought to result from the alterations in differentiation pathways of bone marrow MSCs (Duque et al. 2004). In addition, an inappropriate ossification of other cell types such as interstitial fibroblasts in ligaments, tendons and muscles in the aged individuals leads to a stiffness in the affected areas and limits movement in the affected joints (Mourkioti & Rosenthal, 2005).

Muscle ageing is associated with a loss of muscle mass, strength and velocity of contraction known as sarcopenia. A progressive muscle denervation, loss of muscle fibres, decreased synthesis of myofibrillar components and the accumulation of connective tissue are some of the known cellular changes that accompany sarcopenia (Shi & Garry, 2006). Skeletal muscle consists of a single differentiated cell type, the contractile myofibre, which upon injury activates satellite cells, the resident stem cells located beneath the basal lamina surrounding each myofibre. With age, there is a gradual decline in the regenerative response of satellite cells to damage, often accompanied by a fibrous scar and/or adipose connective tissue (Renault et al. 2002; Lees et al. 2006). The alterations in local stem cell niches in aged muscles are found to affect directly the regenerative ability of muscle satellite cells (Carlson & Conboy, 2007).

Age-associated dysfunction in vasculature is a main cause of cardiovascular diseases such as atherosclerosis and hypertension (Shantsila et al. 2007). The vascular wall and endothelium undergo constant processes of injury and repair to both mechanical and chemical injury. Microvascular wall alterations of widening and atrophy occur with age together with poor angiogenesis. It is thought that the vascular progenitor stem cells from the bone marrow, peripheral blood or vascular adventitia can all give rise to endothelial cells and smooth muscle cells which play an important role in vascular integrity and endothelial repair (Hill et al. 2003; Qian et al. 2007). During ageing and in vascular pathological conditions such as coronary artery disease, these processes are thought to become impaired due to a decreased number of circulating endothelial progenitor cells (Boos et al. 2006; Tao et al. 2006).

Age-related deficits in heart function often lead to ischaemic cardio-myopathy and heart failure in old age (Torella et al. 2006). In the adult mammalian heart, cellular cardiomyocyte hypertrophy and proliferation of cardiac fibroblasts primarily characterizes the regenerative response to injury (Ballard & Edelberg, 2007). Moreover, cardiac stem cells (CSCs) have been identified in the heart, which are able to differentiate into cardiomyocytes and vascular cells and replace damaged myocardium and vascular tissues (Beltrami et al. 2003). CSCs are believed to be subject to age-dependent changes that impair their function and contribute to deregulated repair mechanisms following injury in the aged heart leading to cardiomyopathy (Torella et al. 2004). In addition, an abnormal ossification of heart valves due to osteogenic differentiation and mineralization of interstitial fibroblasts is a common cause of their age-dependent degeneration (Moioli et al. 2007).

The aging brain is characterized by a significant decrease in weight and volume, particularly after the age of 50. This atrophy is thought to result from the loss of neurons and a decline in myelination of axons by the myelinating glial cells, oligodendrocytes, in response to injury (Franklin et al. 2002). In contrast, gliosis, which exhibits both astrocytic and microglial markers, is a prominent feature of the aged brain and neuropathic conditions such as stroke, which involves production of a dense fibrous tissue in the areas of damage (Conde & Streit, 2006). Recent evidence supports the view that neurogenesis in the adult mammalian brain occurs in discrete regions via multipotent adult neuronal stem cells (NSCs), which give rise to all three main neuronal cell types, the neurons, astrocytes and oligodendrocytes (Johansson et al. 1999; Rietze et al. 2001). Although the identity of these NSCs is still a matter of debate, the astroglial origin for NSCs in the adult brain has received much support (Doetsch et al. 1999; Sanai et al. 2004). There is a decreased rate of neurogenesis with increasing age in the adult brains of mammals (Kuhn et al. 1996; Amrein et al. 2004), which may contribute to memory deficits and learning impediments known to occur in advanced age (Shors et al. 2001; Winocur et al. 2006).

A-type lamins as guardians of the ageing soma

According to ‘disposable soma’ theory the strongest candidates for longevity genes are those regulating somatic maintenance and repair, including the cellular responses to stress. Recently, the lamin A gene has been linked to longevity and was proposed to be a guardian of somatic cells during their lifetime (Hutchison & Worman, 2004). Mutations in the lamin A gene cause a spectrum of 20 age-related human disorders termed laminopathies, affecting the maintenance of one or more tissues of mesenchymal origin including skeletal muscle, tendons, adipose, skin and skin appendages, bone, peripheral neurons, myocardium and vasculature. These disorders represent ‘a functional continuum of related conditions rather than separate diseases’ (Bonne & Levy, 2003) with the same underlying cause, which is modulated by genetic and/or environmental factors (Stewart et al. 2007). Remarkably, many tissues affected by mutations in the lamin A gene are also affected in many degenerative conditions common to old age including sarcopenia, osteoporosis, atherosclerosis, cardio-myopathy, dermo-atrophy, neuropathy and lipo-atrophy (see Fig. 1). The compromised tissue functions in laminopathy diseases are proposed to be a consequence of decreased cellular proliferation (Pekovic et al. 2007), a failure to maintain a differentiated state and/or a loss of tissue repair during regeneration (Mounkes & Stewart, 2004; Bakay et al. 2006). Mouse models of lamin A knock-out, mutated lamin A knock-in or transgenic mice manifesting either muscular dystrophy (Sullivan et al. 1999; Arimura et al. 2005), premature ageing (Mounkes et al. 2003; Yang et al. 2005; Varga et al. 2006) or dilated cardiomyopathy (Mounkes et al. 2005), respectively, have shortened lifespans and die prematurely. Moreover, mouse models null for the pre-lamin A processing enzyme Zmpste24 also have shortened lifespans and show progeria-like pathologies of bone and muscle (Bergo et al. 2002; Pendas et al. 2002) whilst a down-regulation of Ce-lamin in C. elegans leads to a 16% decrease in lifespan (Haithcock et al. 2005).

Laminopathies can be loosely divided into four categories: those affecting primarily one specific tissue such as either striated muscle, adipose and bone tissues or peripheral neurons, and those affecting several tissues in a systemic manner as it occurs in the premature ageing syndromes and their related progeroid-like disorders (Broers et al. 2006; Worman & Bonne, 2007). However, it is important to stress that even disorders affecting predominantly one or two tissues often have a variable degree of involvement in other tissues (see Fig. 1). Disorders of striated muscle include: the autosomal dominant and recessive forms of Emery–Dreyfuss muscular dystrophy (AD- & AR-EDMD, Bonne et al. 1999; Raffaele Di Barletta et al. 2000), Limb-Girdle Muscular Dystrophy with atrioventricular conduction disturbances type 1B (LGMD1B, Muchir et al. 2000) and Dilated Cardiomyopathy with conduction system disease (DCM-CD; Fatkin et al. 1999). Other atypical striated muscle conditions include: Dilated Cardiomyopathy with early-onset cardiac fibrosis (DCM-CF, van Tintelen et al. 2007), Dilated Cardiomyopathy with apical left ventricular aneurysm (DCM-VA, Forissier et al. 2003), Early-onset Atrial Fibrillation (AF, Sebillon et al. 2003), Amyotrophic Quadricipital Myopathy with dilated cardiomyopathy (QM-DCM, Charniot et al. 2003) and Dropped Head Syndrome (DHS, D’Amico et al. 2005). Disorders of peripheral neurons include: the autosomal recessive and dominant forms of Charcot-Marie-Tooth disorder type 2 (AR- and AD-CMT2, De Sandre-Giovannoli et al. 2002; Chaouch et al. 2003; Goizet et al. 2004; Benedetti et al. 2005). Disorders affecting adipose tissue include: Dunnigan-type Familial Partial Lipodystrophy (FPLD, Cao & Hegele 2000; Shackleton et al. 2000) or FPLD with variable skeletal and heart involvement (Garg et al. 2002; van der Kooi et al. 2002; Vantyghem et al. 2004), Polycystic ovaries with type A insulin resistance syndrome (POS, Young et al. 2005) and Mandibuloacral Dysplasia (MAD, Novelli et al. 2002). The two premature ageing syndromes include: HGPS (De Sandre-Giovanolli et al. 2003; Cao & Hegele 2003; Eriksson et al. 2003) and Atypical Werner Syndrome (AWS, Chen et al. 2003a,b). The progeroid-like disorders include: Restrictive Dermopathy (RD, Navarro et al. 2004), congenital Seip–Berardinelli Syndrome (Seip, Csoka et al. 2004), Lethal Foetal Akinesis (LFA, Muchir et al. 2003) and Systemic laminopathy with generalized lipoatrophy, insulin-resistant diabetes, leukomelanodermic papules, liver steatosis and hypertrophic cardiomyopathy (LILLC, Caux et al. 2003). Moreover, mutations in the pre-lamin A processing enzyme Zmpste24 have been found in patients with RD and MAD (Agarwal et al. 2003; Navarro et al. 2005). Whilst most laminopathies are caused by missense mutations spread throughout the protein, most cases of HGPS are caused by a de novo silent mutation which activates a cryptic splice site within exon 11 of the LMNA gene and leads to the production of an alternatively spliced truncated lamin A variant lacking 150 nucleotides from its 3′ end (De Sandre-Giovanoli et al. 2003; Eriksson et al. 2003). In addition, there are other rare single base mutations within exon 11 or other exons of the LMNA gene that have been reported to cause atypical HGPS (Rankin & Ellard, 2006).

The nuclear lamina as a stress-resistant filamentous network

The vertebrate nuclear lamins are type V intermediate filament proteins of A-type and B-type that together with a diverse array of inner nuclear membrane proteins form a stress-resistant elastic network termed the nuclear lamina, which lies beneath the inner nuclear membrane and also veils the nucleus (Hutchison & Worman, 2004; Mounkes & Stewart, 2004; Smith et al. 2005; Broers et al. 2006). The nuclear envelope (NE) consists of the outer and inner nuclear membranes (ONM & INM) (the former being continuous with the rough endoplasmic reticulum, ER), the nuclear pore complexes (NPCs) and the nuclear lamina (NL), which together bridge the cytoskeleton and the various cytosolic organelles in the cytoplasm with the chromatin, the nuclear bodies and the nucleoskeleton inside the nucleus (Salpingidou et al. 2007) (see Fig. 2). Although it has been historically assumed that the nuclear lamina is an inert scaffolding structure which essentially functions to provide structural support to the nucleus and protect the peripheral chromatin, it is now known that the nuclear lamina is directly or indirectly involved in a number of fundamental molecular processes ranging from DNA replication and RNA transcription (Spann et al. 1997, 2002) to genome silencing and DNA repair (Liu et al. 2005, 2006; Reddy et al. 2008).

Fig. 2.

Schematic diagram depicting an overview of the major structural components of the nuclear envelope and their interactions and organization in the nucleus. These include the outer and inner nuclear membranes (ONM & INM) [the former of which is continuous with the endoplasmic reticulum (ER) lumen lined with ribosomes], the nuclear pore complexes (NPCs) and the nuclear lamina (NL) composed of A-type and B-type lamin oligomers, and INM & ONM proteins. Some selected INM proteins include: lamin B receptor (LBR), lamina-associated protein 2β (LAP2β), MAN1, emerin and Sun1. Sun1 acts as a tether for Nesprin 2 in the ONM through interactions which span the perinuclear space and link cytoskeleton to the nuclear membrane. Chromatin domains are also anchored to the nuclear lamina. A-type lamins and their binding partners (emerin, MAN1, LAP2α, Sun1) form functional complexes at the nuclear periphery and/or in the nucleoplasm, which have a role in gene regulation and signalling pathways. Not depicted are complexes involving B-type lamin binding partners such as LAP2β and LBR. Some selected nucleoplasmic partners include LAP2α, PP2A, Rb, E2F, b-catenin (b-cat), c-fos, Smads and SREBP1. Question marks indicate suggested but not yet proven A-type lamin interactions with p53, NF-κβ and ERK1/2.

Lamin monomers consist of a central alpha-helical rod domain flanked by a short amino-terminal head domain and a larger carboxy-terminal globular tail and become assembled in the nucleus through a process of dimerization, polymerization and higher-order assembly (Stuurman et al. 1998). It is suspected that lamins form a variety of oligomers and polymers with distinct binding properties both at the nuclear envelope and in the nucleoplasm (Broers et al. 1999; Zastrow et al. 2004). In vertebrates, A-type lamins are produced by an alternative splicing of the LMNA gene and include several somatic (A, C, A delta10) and germ cell-specific (C2) isoforms whilst B-type lamins are produced from two distinct genes, LMNB1 (B1) and LMNB2 (somatic B2 and germ cells-specific B3) (Hutchison, 2002). In D. melanogaster, only one A-type (lamin C) and one B-type lamin have been identified (Dm0), whilst in C. elegans only one B-type lamin (with A-type lamin features) has been found (Ce-lamin). Interestingly, lamin C is first expressed in the gut tissue of Drosophila(Riemer et al. 1995), which is so far the only tissue found in this organism able to replenish itself by adult intestinal stem cells (Ohlstein & Spradling, 2006; Micchelli & Perrimon, 2006). In C. elegans, the B-type lamin, Ce-lamin, is crucial in germ cell development (Margalit et al. 2005). As the complexity of lamin proteins increased during evolution (Cohen et al. 2001), it is tempting to speculate that B-type lamins evolved to protect germ cells in all multicellular organisms whilst A-type lamin isoforms co-evolved with the appearance of adult stem cells in a wide range of tissues in long-lived species.

A-type and B-type lamins differ in their post-translational modifications of the C-terminal CAAX motif (C, cysteine; A, any aliphatic amino acid; X, any amino acid) which is an isoprenylation signal found in many proteins including Ras, Ras-related proteins and G protein subunits, and is required for their attachment to the inner nuclear membrane (Rusinol & Sinensky, 2006). Mature lamin A is synthesized as a precursor prelamin A containing the CAAX motif, which similarly to B-type lamins, undergoes a sequence of reversible modifications including cysteine isoprenylation, –AAX proteolytic cleavage and cysteine carboxy-methylation (Rusinol & Sinensky, 2006). Whilst B-type lamins retain the modified C-terminal cysteine and as such are tightly coupled to the inner nuclear membrane, the maturation of lamin A requires a unique 15 amino acid cleavage at a conserved hexapeptide cleavage site which removes the downstream C-terminal modified cysteine. In contrast, an A-type lamin isoform, lamin C, lacks the C-terminal CAAX motif, and is thought to become attached to the inner nuclear membrane via interaction with prelamin A (Hutchison et al. 2001). In a majority of individuals with HGPS, the splicing mutations in the LMNA gene lead to the 50 amino acid truncated form of prelamin A, termed progerin, that has lost the second endoprotease cleavage site and thus permanently acquired the C-terminal modified cysteine which interferes with mitotic disassembly (Cao et al. 2007; Dechat et al. 2007).

One of the hallmarks of laminopathy cells are dramatic deformations of the nuclear morphology ranging from dysmorphic nuclear shapes to envelope raffling, blebbing, loss peripheral heterochromatin and/or INM proteins from one pole of the nucleus, and nuclear fragility upon stress (Hutchison, 2002; Goldman et al. 2004). Mutations in A-type lamins not only lead to nuclear envelope fragility but also to nucleoplasmic lamin disorganization (Broers et al. 2005; Wiesel et al. 2008), and can affect both the structural organization of lamin dimers and filaments as well as protein–protein and protein–DNA interactions (Dhe-Paganon et al. 2002; Krimm et al. 2002).

A-type lamins have a diverse spectrum of interacting partners ranging from the nuclear membrane proteins such as B-type lamins (Schirmer et al. 2001), emerin (Clements et al. 2000; Lee et al. 2001; Vaughan et al. 2001; Holaska & Wilson, 2007) and MAN1 (Mansharamani & Wilson, 2005), cytoskeletal linker proteins such as nesprins (Mislow et al. 2002; Zhang et al. 2005; Libotte et al. 2005) and SUN (Sad/UNC-84) domain proteins (Haque et al. 2006; Padmakumar et al. 2005) to gene regulatory proteins such as Rb (Ozaki et al. 1994; Mancini et al. 1994), LAP2α (lamina-associated polypeptide 2α) (Dechat et al. 2000), Smads (van Berlo et al. 2005), SREBP1 (sterol-regulatory element-binding protein) (Lloyd et al. 2002), c-fos (Ivorra et al. 2006) and MyoD (Bakay et al. 2006) (see Fig. 2). Mutations in A-type lamin binding partners including emerin, nesprins 1/2, LAP2α, MAN1 and B-type lamins have been implicated in X-linked EDMD (Bione et al. 1994), AD-EDMD (Zhang et al. 2007), DCM (Taylor et al. 2005), disorders of increased bone density (Osteopoikilosis, Busche–Ollendorf Syndrome and Melorheostosis) (Hellemans et al. 2004), AD-Leukodystrophy (Padiath et al. 2006) and Acquired Partial Lipodystrophy (Hegele et al. 2006). In the absence of A-type lamins, emerin (Sullivan et al. 1999; Vaughan et al. 2001) and nesprins (Muchir et al. 2000; Libotte et al. 2005) mislocalize to the ER whilst LAP2α forms aggregates in the nucleus (Muchir et al. 2000; Pekovic et al. 2007) and Rb is targeted to the proteasome and/or subnuclear splicing domains (Johnson et al. 2004; Pekovic et al. 2007). In addition, an absence of A-type lamins leads to the premature nuclear entry of transcription factors such as Smads (van Berlo et al. 2005) and c-fos (Ivorra et al. 2006).

A-type lamins as defenders against growth-related postnatal stress

B-type lamins are dynamically expressed throughout all stages of embryonic development and in all cells of the adult organism (Broers et al. 1997) and are essential for growth, development and survival (Harborth et al. 2001; Vergnes et al. 2004). In contrast, A-type lamin expression is developmentally regulated. Some of the transcriptional regulators that control the promoter activity of the LMNA gene include AP1 (activating protein 1), Sp1/3 (Muralikrishna & Parnaik, 2001), hepatocyte nuclear factor-3β, retinoic X receptor β (Arora et al. 2004) and p53 (Rahman-Roblick et al. 2007). Although A-type lamins are initially present in fertilized oocytes, they are absent during early embryonic development and become expressed in a tissue-specific manner during the middle stages of embryonic development as well as during postnatal development (Stewart & Burke 1987; Broers et al. 1997). Indeed, lamin A/C has recently been reported as a marker of both mouse and human embryonic differentiation (Constantinescu et al. 2006). For example, during mouse embryonic development, lamins A/C appear first during myogenesis of the eye and trunk muscles followed by their appearance during epidermis stratification whereas the expression in skeletal muscle, heart, liver, lung and brain is postnatal (Stewart & Burke, 1987; Rober et al. 1989). Consistent with these observations, in adult organisms, A-type lamins continue to be expressed in most differentiated adult cells except for undifferentiated lymphoid cells of the spleen, thymus, blood and bone marrow (Stewart & Burke, 1987). In addition, neural and neuro-endocrine cells of adult mice are also absent for A-type lamins (Broers et al. 1997). However, caution should be exercised with some of these conclusions, as some of these earlier studies relied solely on immunofluorescence microscopy and the absence of A-type lamins could be a result of epitope masking (Tunnah et al. 2005).

Recently, Takamori and co-workers gained a valuable insight into the expression of A-type lamins during neurogenesis in the adult rat brain. Remarkably, whilst lamins A/C were expressed in the primary neuronal progenitor stem cells as well as in the mature neurons of the adult rat brain, they were not or only marginally present in the early and late progenitor cells, respectively (Takamori et al. 2007). Therefore, in the adult mammalian tissues, lamin A/C expression seems to be reduced or absent in undifferentiated or proliferative cells, but is observed in differentiated or non-proliferative cells such as quiescent adult stem cells. Moreover, the primary embryonic fibroblasts from LMNA knockout (ko) mice models show no differences in proliferation patterns from wild-type cells across their lifespan (Sullivan et al. 1999). However, fibroblasts from early postnatal LMNA ko mice show severe proliferation defects and rapidly die (Mounkes & Stewart, 2004). Taken together, it becomes apparent that A-type lamins are required for postnatal growth and the maintenance of quiescence and differentiation. Given the argument that the rate of ageing of more complex tissues depends on the maintenance of their post-mitotic cells through adult stem cell regeneration, this leads us to suggest that the longer lifespans in more complex organisms may be linked to the protection of adult stem cell function by A-type lamins.

The nuclear lamina and signalling networks in adult stem cell homeostasis

The control of stem cell homeostasis within tissues such as the balance between proliferation, differentiation, apoptosis and senescence is strictly linked to the regulation of tissue repair and regeneration. It has been proposed that disease-causing mutations in A-type lamins perturb the balance between proliferation and differentiation in adult stem cells, leading to less efficient tissue regeneration (Gotzmann & Foisner, 2006; Vlcek & Foisner, 2007). There are at least three signalling pathways central to adult stem cell regeneration which have been linked to the functions of A-type lamins and their binding partners, including the Rb/E2F and Rb/MyoD pathways, the Wnt/β-catenin pathway and the TGF-β/Smad pathway (see Figs 3 and 5A). Therefore, A-type lamins can be viewed as ‘signalling receptors’ of the nucleus which can receive and transduce signals from the extracellular matrix and cytosol through covalent or non-covalent changes. This has been extensively studied in the satellite muscle cells involved in postnatal muscle growth and regeneration as well as the pre-adipocytes and dermal fibroblasts with A-type lamin, emerin, MAN1 or LAP2α mutations.

Fig. 3.

Schematic diagram illustrating how A-type lamins in association with other INM proteins modulate different signalling pathways that participate in either adult stem cell maintenance (TGF-β pathway) or growth-related stress responses (Ras/MAPK pathway). In the TGF-β pathway, TGF-β ligands in the extracellular space bind to dimerized TGF-β serine/threonine kinase receptors (type I and type II R) found on the plasma membrane. Ligand-receptor binding results in the phosphorylation of the receptor Smads 2/3, which disassociate from the receptor and form a heteromeric complex with the co-mediator Smad 4 in the cytoplasm, which can then translocate across the nuclear membrane. A-type lamins alone or together with the INM protein MAN1 bind to the Smad complex in the nucleus and inhibit their activation of transcription factors (TF) and target genes. In the Ras/MAPK pathway, growth factors in the extracellular matrix bind to heterotrimeric G protein-coupled receptors (GPCR) or receptor tyrosine kinases (not shown) on the plasma membrane. This leads to the activation of the G protein Ras (in a complex with Grb2 and SOS proteins) through GDP/GTP exchange, which is then free to bind to serine/threonine protein kinase Raf and/or G protein Rac (not shown). In the cytosol, downstream kinase cascade from Raf includes phosphorylation of protein kinase MEK 1/2, which leads to activation and enhanced phosphorylation of ERK 1/2 dimers and their subsequent translocation across the nuclear membrane. A-type lamins alone or together with some unknown INM proteins may bind to phosphorylated ERK 1/2 and inhibit their activation of transcription factors (TF) and gene expression. Grb, growth factor receptor-bound protein; SOS, son-of-sevenless; GDP/GTP, guanosine di/triphosphate; MEK, mitogen- or extracellular signal-related kinase.

Fig. 5.

Schematic representation of the involvement of A-type lamins in the signalling pathways central to adult stem cell regeneration (A) and adaptive responses to stress (B), which control the balance between proliferation, differentiation, senescence and apoptosis. (A) A-type lamins and their binding partners regulate at least three regenerative pathways including Rb/E2F and Rb/MyoD, Wnt/β-catenin and TGF-β/Smad. A number of extracellular stimuli including growth factors, cytokines and mitogens trigger the activation of either the Wnt-Frizzled receptor (and co-receptor LRP5/6), which protects their cytosolic target β-catenin against proteosomal degradation, or activation of the TGF-β receptor kinases 1/2 that phosphorylate receptor-associated Smads 2/3 in the cytosol. This allows β-catenin or Smads 2/3 (in complex with Smad 4) to translocate to the nucleus from the cytosol and control expression of genes. A-type lamin-binding partner emerin binds to β-catenin in the nucleus and promotes its nuclear export, which represses activation of TCF/LEF-responsive genes. A-type lamins alone or in concert with their binding partner MAN1 bind to Smads 2/3 in the nucleus and repress Smad-responsive genes by inducing their de-phosphorylation through nuclear phosphatase PP2A. A-type lamins and their binding partner LAP2α bind to transcription factor Rb and regulate the expression of E2F-dependent genes during S-phase or the activation of MyoD-dependent genes during muscle differentiation. (B) In response to mechanical stress, DNA damage, increased ROS levels or inflammatory cytokines in the extracellular space, cells activate the stress-associated signalling pathways through a number of cytosolic receptors such as TNF-α, PDGF-β and G protein coupled receptors, which activate cytosolic I-kβ kinase complex and its target NF-kβ, ATM kinase and its target p53 or one of the kinase branches of the Ras/MAPK cascade (ERK1/2, JNK/SAPK or p38), respectively. A-type lamins alone or in concert with other unknown binding partners regulate the nuclear entry, phosphorylation, DNA binding and/or transcriptional activity of these nuclear targets and thus their control of gene expression. Note that it is not known which cytosolic receptors lead to the activation of Rb in the cytosol but Rb is a downstream nuclear effector of a number of pathways including TGF-β, Ras/MAPK and p53.

Rb pathways

Regulation of the cell cycle via the Rb/E2F pathway involves the cell cycle-regulated activity of cyclins, cyclin-dependent kinases (CDK), CDK inhibitors p16^ink4 and p21^cip1 and their transcription factor targets, the retinoblastoma protein Rb and E2F (Galderisi et al. 2006). A-type lamins and their binding partner LAP2α bind to and tether Rb/E2F transcription factor complexes in the nucleus during G1 phase of the cell cycle, which is essential for its role in repressing activation of E2F-dependent S-phase genes (Markiewicz et al. 2002; Dorner et al. 2006). Hyper-phosphorylation of Rb during the G1/S phase transition by cyclin/CDKs, which are activated by mitogenic signalling, releases active E2F and leads to S-phase progression (Mittnacht et al. 1997). This regulation of Rb by the nucleoplasmic A-type lamin/LAP2α complexes is exerted through either protecting Rb from proteosomal degradation (Johnson et al. 2004) and targeting to the subnuclear splicing speckles (Pekovic et al. 2007), or through controlling its rate of phosphorylation via protein phosphatases such as PP2A (van Berlo et al. 2005). Loss of Rb regulation by either RNAi knock-down of LAP2α or A-type lamins in human and mouse cells leads to accelerated S-phase entry, which is followed by either inadequate growth arrest following withdrawal of mitogens (Dorner et al. 2006) or S-phase delay and G2 arrest (Pekovic et al. 2007). Similarly, in the absence of Rb, skeletal muscle cells progress slowly through S phase but then arrest in S and G2 phases of the cell cycle, most likely due to incomplete replication and/or DNA damage (Novitch et al. 1996). A deregulated Rb/E2F pathway in laminopathy cells could therefore lead to premature stem cell exhaustion through increased proliferation and/or the accumulation of DNA damage.

Rb can cooperate with a number of tissue-specific transcription factors in order to regulate tissue-specific differentiation. The Rb/MyoD pathway has a well-established role in myogenesis of satellite myoblasts whereby the interaction of Rb with the myogenic transcription factor MyoD leads to the activation of MyoD-dependent target genes and initiation of muscle differentiation (Novitch et al. 1996). A number of recent studies have shed new light on the role of A-type lamins and their binding partner emerin in satellite cell differentiation and regeneration upon injury. Mouse satellite cells null for A-type lamins or emerin have impaired differentiation potential as a result of delayed ability to exit the cell cycle and show a decreased expression of proteins critical in muscle differentiation including Rb, MyoD, desmin and M-cadherin (Frock et al. 2006). Similarly, two further studies have implicated a deregulated Rb/MyoD pathway as a cause of decreased differentiation potential in A-type lamin or emerin null regenerating muscle using mRNA expression profiling (Bakay et al. 2006; Melcon et al. 2006). Interestingly, other myogenic factors such as Myf5 and Pax7 are up-regulated in lamin A/C null cells, indicating that they still retain myogenic properties (Frock et al. 2006). Over-expression of Myf5 in the absence of MyoD is thought to be a marker for self-renewing satellite cells or ‘reserve cells’ in the adult muscle (Yablonka-Reuveni et al. 1999). In this regard, lamin A/C and emerin null myoblasts resemble the behaviour of MyoD null myoblasts which undergo an enhanced self-renewal at the expense of producing progeny that undergoes differentiation (Megeney et al. 1996). Two previous studies on mouse myoblasts transfected with mutant lamin A demonstrated the inhibition of muscle differentiation and decreased expression of late myogenic factors such as myogenin and cathepsin B (Favreau et al. 2004; Markiewicz et al. 2005, see Fig. 4). These myoblasts exhibited a failure to exit the cell cycle as seen by persistently high levels of hyper-phosphorylated Rb and proliferation marker PCNA.

Fig. 4.

C2C12 mouse myoblasts stably transfected with GFP-R453W lamin A linked to muscular dystrophy fail to undergo muscle differentiation and up-regulation of muscle-specific gene products. C2C12 mouse myoblasts stably transfected with GFP-wt lamin A or GFP-R453W lamin A were grown for 3 days after which they were induced to differentiate for 4 days by replacing their growth media (DMEM/10% FCS) with differentiation media (DMEM/2% horse serum). Coverslips with cells were taken after 24, 72 and 96 h and processed for confocal microscopy using the antibody against late muscle differentiation marker myogenin and DNA-intercalating dye DAPI. Images were collected on a Zeiss confocal microscope and projected as blue/red colour merged micrographs in which myogenin is in red and DAPI is in blue. DMEM, Dulbecco's modified Eagle's medium.

In preadipocytes, Rb stimulates adipocyte differentiation through modulating adipogenic factors such as CAAT/enhancer binding protein-α (C/EBP-α) and peroxisome proliferator-activated receptor-γ (PPAR-γ) (Classon et al. 2000). Prelamin A-binding protein SREBP1 also activates PPAR-γ when its activated form gets translocated to the nucleus (Lloyd et al. 2002; Capanni et al. 2005). In fibroblasts of FPLD, MAD and AWS patients which accumulate prelamin A, SREBP1 is sequestered to the nuclear envelope which prevents the activation of PPAR-γ and thus inhibits adipogenesis. Similarly, the over-expression of both wild-type and the lipodystrophy-causing mutant lamin A inhibits lipid accumulation and the expression of adipogenic markers (Boguslavsky et al. 2006). Moreover, the over-expression of A-type lamin-binding partner LAP2α in pre-adipocytes initiates early events in the adipogenic pathway but inhibits lipid accumulation and terminal adipogenesis (Dorner et al. 2006). In contrast, embryonic fibroblasts lacking A-type lamins undergo adipogenesis more readily than controls and these lamin A null mice do not develop lipodystrophy (Sullivan et al. 1999). These studies suggest that down-regulation of A-type lamins in physiological conditions may be necessary to induce normal adipogenesis. Given that in cells from lipodystrophy patients A-type lamin levels are normal, a loss of adipogenic potential may be attributed to their increased stability in cells. The ability of A-type lamins to dynamically reorganize themselves (Markiewicz et al. 2005; Mariappan & Parnaik, 2005) and/or to modulate their stability (E. Markiewicz, personal communication) may be important regulators of mesenchymal differentiation.

Wnt/β-catenin pathway

The canonical Wnt/β-catenin pathway is one of the most important regulators of mesenchymal tissue proliferation and differentiation. It acts via the Wnt-Frizzled receptor and co-receptor LRP5/6 (low-density lipoprotein receptor-related protein) through inactivating a GSK3-β/Axin/APC destruction complex which regulates β-catenin ubiquitination and proteosomal degradation (GSKβ-glycogen synthase kinaseβ; APC-adenomatous polyposis coli) (Hartmann, 2006). This leads to a stabilized cytoplasmic pool of β-catenin which can then translocate to the nucleus resulting in the co-activation of transcription factors TCF/LEF (T cell factor/lymphoid enhancer factor). Emerin binds to β-catenin via its C-terminal APC-like domain which restricts the nuclear accumulation of β-catenin in a CRM1-export-dependent manner (CRM1-exportin chromosome region maintenance1) and thus represses the transcriptional activation of TCF/LEF-responsive genes (Markiewicz et al. 2006). Indeed, a number of genes associated with Wnt signalling were found to be up-regulated in X-EDMD patient fibroblasts (Tsukahara et al. 2002). The downstream nuclear targets of β-catenin include a number of proliferative genes such as cyclin D1 and c-myc which are both up-regulated in emerin-null human fibroblasts leading to auto-stimulatory growth phenotype (Markiewicz et al. 2006). Abnormal β-catenin activity has been implicated in a number of mesenchymal fibro-proliferative disorders (Bowley et al. 2007). As cardiac and skeletal muscles accumulate fatty fibrotic tissue in patients with emerin and LMNA mutations, the aberrant Wnt signalling is suggested to lead to this phenotype in X-EDMD patients through accelerated cell turnover (Markiewicz et al. 2006). Remarkably, increased Wnt/β-catenin signalling has been found to induce the conversion of aged muscle satellite cells into fibroblastic cell lineage during muscle regeneration leading to increased fibrosis associated with poor regenerative response in the aged muscle (Brack et al. 2007). However, in recent studies of both human and mouse EDMD models of muscle regeneration, abnormal Wnt signalling has not yet been implicated.

TGF-β/Smad pathway

The canonical tumour growth factor-β (TGF-β)/Smad pathway is another important regulator of mesenchymal tissue homeostasis (Verrecchia et al. 2006). It acts through the TGF-β receptor kinases 1/2 that phosphorylate the receptor-associated Smads 2 and 3 which then interact with the co-mediator Smad 4 and as such can be translocated to the nucleus where they control the expression of various genes such as collagen (see Fig. 3). TGF-β can also induce an activation of nuclear phosphatase PP2A whereby it requires Rb protein for the execution of its growth-suppressing role (Herrera et al. 1996; Derynck & Zhang, 2003). A-type lamins have been found to bind to both the TGF-β-induced Smads 2/3 and Rb and to repress TGF-β/Smad-dependent gene expression in mouse embryonic fibroblasts (MEFs) by promoting their dephosphorylation via PP2A (van Berlo et al. 2005). Interestingly, in senile osteoporosis, decreases in the expression of TGF-β signalling in somatic cells and MSCs of aged mice (Moerman et al. 2004; Han et al. 2005) induce a lack of osteoblasts at the expense of adipocytes in the bone matrix (Chan & Duque, 2002). This resembles the situation in MSCs of lmna-deficient mice, which show a preference towards adipocyte as opposed to osteoblast differentiation (Mounkes et al. 2003). Moreover, an A-type lamin-binding partner MAN1, which interacts with R-Smads and antagonizes TGF-β signalling (Lin et al. 2005; Pan et al. 2005) when mutated, leads to the over-stimulation of this pathway and increased bone density. Over-expression of TGF-β signalling is also implicated in tissue fibrosis in muscle and heart due to increased extracellular matrix production (Khan & Sheppard, 2006). It is thus suggested that the increased proliferation rate and collagen production as a result of deregulated TGF-β/Smad signalling in lmna-null mice has implications for the tissue fibrosis seen in their heart and muscle (van Berlo et al. 2005).

A-type lamins in regulation of stress-activated signalling pathways

In addition to pathways associated with tissue regeneration, A-type lamins have been linked to the signalling pathways associated with cellular responses to stress including MAPK signalling pathways, the p53/p21 and Rb/p16 pathways and the NF-κβ pathway, which are either chronically up-regulated or fail to activate upon stress in the cells and tissues with lamin A/C mutations (see Figs 3 and 5B).

MAPK pathways

Genome-wide profiling of hearts in a mouse model of AD-EDMD revealed increases in the expression of several genes associated with MAPK signalling such as c-jun, Elk1, pERK1/2 and pJNK as well as the genes involved in fibrosis and inflammation (Muchir et al. 2007). Activation of MAPK (mitogen-activated protein kinase) signalling either through ERK or JNK in hearts has been associated with the development of cardiomyopathy in mouse models (Nicol et al. 2001; Petrich et al. 2003). MAPK signalling is initiated from a number of receptor and non-receptor tyrosine kinases or heterotrimeric receptors coupled to small G proteins such as Ras (Iwasa et al. 2003). There are a number of effector pathways downstream of Ras, the best known being the mitogenic signalling pathway, the MAPK cascade (see Fig. 3). MAPK pathways consist of three distinct signalling branches including ERKs (extracellular signal-related kinases), JNK/SAPKs (Jun N-terminal kinases/stress-activated protein kinases) and p38 which, when activated by phosphorylation, translocate to the nucleus and transactivate transcription factors involved in gene regulation. Whilst the ERK pathway is activated by mitotic stimuli, the latter two pathways are activated by cellular stresses such as UV light, ROS (reactive oxygen species) and inflammatory cytokines. Ras pathway over-activation leads to premature senescence in cells with intact cell cycle controls (Serrano et al. 1997) by activating the stress-kinases JNK/SAP and p38 (Hutter et al. 2002; Wang et al. 2002; Iwasa et al. 2003). Thus, the activation of JNK in lamin A/C mutant hearts could lead to cardiomyopathy by inducing premature senescence of cardiomyocytes or resident cardiac stem cells. Interestingly, downstream targets of Ras-mediated signalling involve the regulation of cyclin D1, pRb phosphorylation and its partner MyoD (Mittnacht et al. 1997). Suppression of Ras signalling using MAPK inhibitors leads to the trans-activation of Rb/MyoD pathway and restores myogenic differentiation in Rb-null muscle cells (Lee et al. 1999). It would be worth investigating whether increased Ras/MAPK activity also prevents proper cardiomyocyte differentiation in cardiac cells with lamin A/C or emerin mutations through modulating the Rb/MyoD pathway as has recently been demonstrated for myogenic differentiation of C2C12 myoblasts expressing EDMD-linked lamin A mutant (Favreau et al. 2008).

p53/p21 and Rb/p16 pathways

The p53/p21 and Rb/p16 pathways are essential gatekeepers of senescent arrest in both mouse and human models (Shay et al. 1991). Activation of these pathways in somatic stem cell compartments limits the regenerative capacity of their respective tissues. Mouse models with high expression or activity of p53 have a shorter lifespan and show signs of premature ageing (Tyner et al. 2002; Maier et al. 2004) and over-expression of CDK inhibitor p16 in stem cells leads to premature exhaustion of their activity (Janzen et al. 2006). The p53/p21 pathway is induced in response to a variety of signals following platelet-derived growth factor-β (PDGF-β) receptor activation of ataxia telangiectasia mutated (ATM) kinases which sense DNA damage (Chen et al. 2003a). Hyper-activation of the p53 pathway has recently been reported in Zmpste24-null and lmna-null mouse models of accelerated ageing (Varela et al. 2005), which is triggered by increased DNA damage and/or a reduced level of anti-oxidant enzymes (Liu et al. 2005; Varela et al. 2005). Moreover, oxidative stress has been implicated in the pathophysiology of lamin A-linked Amyotrophic Quadricipital Syndrome with cardiac involvement and in human FPLD fibroblasts with lamin A/C mutations which undergo p16-dependent senescent arrest (Caron et al. 2007; Charniot et al. 2007). Interestingly, lmna-null and pre-lamin A processing mutant mouse cells which have low levels of Rb cannot arrest in response to DNA damage or increased activity of CDK inhibitor p16^ink4(Johnson et al. 2004; Nitta et al. 2006). These data imply that A-type lamin/Rb complexes may be important in both DNA damage-induced and p16-dependent senescent arrest. How would this regulation be exerted? Rb has a well-known role in activating growth arrest following DNA damage and preventing the accumulation of DNA damage in the face of unscheduled DNA synthesis, which would otherwise give rise to double-stranded breaks (DSBs) and genomic instability (Bosco et al. 2004). Impaired DNA repair, genomic instability and premature senescence have also been demonstrated in Zmpste24-null MEFs, bone marrow cells and human HGPS fibroblasts (Liu et al. 2005).

NF-κβ pathway

The nuclear factor-κβ (NF-κβ) pathway is involved in the regulation of diverse cellular processes including survival, apoptosis, cell growth, differentiation, immunity and cellular responses to stress (Jones et al. 2003). It can be activated by the canonical cytokine-mediated pathways such as tumour necrosis factor-α (TNF-α) or by signal transduction cascades involving the ERK1/2 branch of MAPK signalling. Many of these cascades activate NF-κβ by activating the inhibitor I-κβ kinase complex, which then allows its translocation to the nucleus and activation of target genes. NF-κβ functions as a key regulator of cardiac gene expression in many physiological states such as adaptive cardiac hypertrophy as well patho-physiological states such as dilated cardiomyopathy and atherosclerosis (Jones et al. 2003). Following cytokine stimulation or mechanical stress, both lamin A/C-null and emerin-null mouse fibroblasts show impaired mechano-sensitive transduction pathways involving NF-κβ at a step downstream of its nuclear entry and DNA binding (Lammerding et al. 2004, 2005). Moreover, in response to mechanical loads, lamin A/C-null cells are not able to resist forces of compression and are prone to apoptosis (Broers et al. 2004; Lammerding et al. 2004), which is not the case for emerin null cells (Lammerding et al. 2005). However, it is not yet known how precisely lamin A/C and emerin are involved in the regulation of NF-κβ transactivation and whether each protein acts independently or in combination with one another or other INM proteins. It has been suggested that the lack of dephosphorylation of NF-κβ subunit relA (reticuloendotheliosis viral oncogene A) by A-type lamin-binding phosphatase PP2A may explain its altered activity in lamin A/C- or emerin-null cells (van Berlo et al. 2005).

A-type lamins at the crossroads between tissue regeneration and stem cell ageing: a summary

Adult stem cell maintenance involves a fine balance between intrinsic mechanisms regulating genome integrity and repair, external factors in the local and systemic niches in the body, and multiple signalling pathways which incorporate these mechanisms as their integral components and regulate the decision as to whether stem cells engage in self-renewal, lineage specification, growth arrest or stress response. Age-dependent tissue degeneration is associated with the alterations in these intrinsic and extrinsic factors in adult stem cells and their niches, which leads to decreased regeneration of tissues upon injury (Conboy & Rando, 2005). As a result, tissue-degeneration in old age is accompanied by a compensatory hyperplasia of the same tissue-specific cell types or interstitial fibrosis (Martin, 2007). One model that explains the decline in adult stem cell regeneration with age in the context of both intrinsic and extrinsic mechanisms concerns the accumulation of senescent cells with age. Senescent cells arise as a result of a stress response to a number of inefficient intrinsic cellular mechanisms that alter the tissue microenvironment, which becomes inhibitory to stem cell activity (Campisi, 2005).

Recently, A-type lamins have been linked to the maintenance and regeneration of a number of mesenchymal tissues and have been proposed to be regulators of mesenchymal stem cell regeneration (Gotzmann & Foisner, 2006). Both lamin A and progerin have been implicated in physiological ageing (Duque & Rivas, 2006; McClintock et al. 2006; Scaffidi & Misteli, 2006; Afilalo et al. 2007; Cao et al. 2007; Ukekawa et al. 2007; Candelario et al. 2008; Huang et al. 2008). Tissues affected by lamin A/C mutations often show degenerative changes accompanied by increased fibrosis and/or lipid accumulation (Hutchison & Worman, 2004). A number of degenerative mesenchymal diseases associated with increasing age show significant clinical overlap with those diseases caused by lamin A/C mutations. How could A-type lamins regulate adult stem cell function and their onset/rate of ageing? A-type lamins interact with a diverse spectrum of interacting partners, which together form a stress-resistant signalling network in the nucleus (Lammerding et al. 2005) that can receive external signals from the extracellular matrix and cytosol and transduce them into intrinsic control of gene expression. The signalling pathways regulating adult stem cell homeostasis as well as cellular responses to stress have been linked to the function of A-type lamins as ‘signalling receptors’ of the nucleus. Regulation of these pathways by A-type lamins in combination with their distinct binding partners may provide an additional level of complexity which may determine the extent of cross-talks between different pathways. For instance, it would be important to investigate whether TGF-β signalling (van Berlo et al. 2005) cooperates with Wnt signalling (Markiewicz et al. 2006) or the MAPK stress signalling (Muchir et al. 2007) in lamin A/C- and emerin-null cells to induce rapid growth and increased fibrosis in hearts and skeletal muscles of patients with EDMD. Wnt signalling in the ageing context has recently been implicated in the abnormal fibrogenic conversion of the aged muscle satellite cells during muscle regeneration (Brack et al. 2007). Moreover, increased TGF-β signalling via the MAPK kinase can lead to premature senescence in human fibroblasts (Kim et al. 2004), which is known to induce β-catenin activation (Sato, 2006). Therefore, the increased fibrosis in the heart and muscle of patients with lamin A/C and emerin mutations may result from the induction of premature senescence in resident stem cells as a result of the abnormal cross-talk between these pathways.

Furthermore, in cases where the affected tissues show severe degeneration, damaged myofibres may send the signals that trigger increased satellite cell renewal and thus lead to a vicious circle of premature stem cell exhaustion. The accumulation of DNA damage in lamin A/C mutant tissues as a result of either decreased NF-κβ-dependent adaptive mechanical responses (Lammerding et al. 2004), inappropriate DNA damage repair (Liu et al. 2005; Manju et al. 2006), inability of cells to prevent unscheduled DNA replication via Rb (Johnson et al. 2004; Nitta et al. 2006) or increased oxidative stress via MAPK stress signalling (Caron et al. 2007; Charniot et al. 2007; Muchir et al. 2007) could all mimic an injury environment in the stem cell niches and lead to constant cycles of stem cell regeneration in those tissues and their depletion. This scenario would lead to stress-induced premature stem cell senescence by up-regulation of p53/p21 signalling and/or the Rb/p16 pathway depending on whether the latter is still intact. In conclusion, a loss of appropriate regulation of a number of regenerative pathways intrinsic to adult stem cell maintenance may underpin the premature demise of tissues with lamin A/C mutations. This may lead to a loss of protection against growth-related stress as well as defects in a number of cellular processes in adult stem cells and/or their progenitors including growth, differentiation, apoptosis and repair, which may in turn switch on the chronic stress signalling pathways that serve to maintain organismal survival but at the same time promote the ageing phenotype. Therefore, A-type lamins should be considered as intrinsic modulators of ageing within the adult stem cells and their niches.

Update

Since this article was submitted for publication, two new complementary studies pertinent to this topic have been published. The first study shows that the expression of progerin in immortalized human MSCs negatively affects their molecular identity and differentiation potential by interfering with the Notch signalling pathway, which is essential in stem cell regulation and maintenance (Scaffidi & Misteli, 2008). The second study demonstrates that the Zmpste24-null mouse model of accelerated ageing with prelamin A accumulation shows a defective proliferative potential of bulge cells in the hair follicle niche due to a decreased Wnt signalling pathway, which leads to dysfunction of epidermal stem cell renewal (Espada et al. 2008). These findings strongly support the hypothesis that lamin A regulates stem cell maintenance via a range of regenerative signalling pathways and provide the insight that the regulation of adult stem cell ageing may occur at a number of levels that intersect with lamin A (wt lamin A, progerin and prelamin A) including adult stem cells, their progenitors and/or stem cell niches.