Interdependent changes of nuclear lamins, nuclear pore complexes, and ploidy regulate cellular regeneration and stress response in the heart

ABSTRACT

In adult mammals, many heart muscle cells (cardiomyocytes) are polyploid, do not proliferate (post-mitotic), and, consequently, cannot contribute to heart regeneration. In contrast, fetal and neonatal heart muscle cells are diploid, proliferate, and contribute to heart regeneration. We have identified interdependent changes of the nuclear lamina, nuclear pore complexes, and DNA-content (ploidy) in heart muscle cell maturation. These results offer new perspectives on how cells alter their nuclear transport and, with that, their gene regulation in response to extracellular signals. We present how changes of the nuclear lamina alter nuclear pore complexes in heart muscle cells. The consequences of these changes for cellular regeneration and stress response in the heart are discussed.

KEYWORDS:

• Cardiac hypertrophy
• cardiac remodeling
• cardiomyocyte
• heart regeneration
• nuclear lamina
• nuclear pore
• nuclear transport

Introduction

A fundamental question in biology is how cells interpret chemical signals in their environment to carry out cellular function, which at the molecular level, is known as cellular signal transduction. Confounding this question is the fact that the same extracellular signal can elicit fundamentally different cellular responses. An excellent example of differing responses is heart muscle cells, known as cardiomyocytes. In the immature state (before birth in most mammals), they respond to proliferative signals, and, yet, as post-mitotic, mature cells, they no longer proliferate in r esponse to the same signals. The clinical significance of this lack of response of post-mitotic cardiomyocytes to proliferative signals is that this limits heart regeneration [1].

If medicine is to advance to direct heart regeneration, it will be important to understand the mechanisms that induce or block cellular signal transduction in the context of cardiomyocyte proliferation. Cellular signal transduction begins at the plasma membrane. Once signals have been propagated across the plasma membrane, they must cross another barrier, the nuclear envelope, to reach the interior of the nucleus and alter gene activity. Little is known about how cells regulate signal propagation across the nuclear envelope. Here, we review the structural changes in the nuclear lamina and nuclear pore complexes that occur during cardiomyocyte maturation, and their broad effects on nuclear import and gene expression.

Nuclear envelope

The nuclear envelope is composed of two phospholipid bilayers, which acts as a barrier between the cytoplasm and nucleoplasm, and encloses the genetic material of the cell. To enable transport between the cytoplasm and nuclear compartments, the nuclear envelope has nuclear pore complexes [2]. Nuclear pore complexes (NPCs) are barrel-shaped structures that are located in the fold between the inner and outer membranes of the nuclear envelope (NE) [2,3]. The fundamental structure of nuclear pore complexes consists of ~30 nucleoporins (NUPs), which are arranged with eightfold symmetry [2,3] (Figure 1). On the cytoplasmic side, specialized filamentous structures funnel cargo through the central core into the nucleus, while on the nucleoplasmic side, a structure, called the nuclear basket, mediates transport through the central channel to the cytoplasm [2,3].

Figure 1. Schematic of nuclear pore complexes, the nuclear lamina, and their association with heart disease. Structural modules are colored according to the scale at the top left. The outer nuclear membrane (ONM) and inner nuclear membrane (INM) of the nuclear envelope are indicated as gray lines. The lamina meshwork is shown as red-brown (B-type lamins) and blue (A-type lamins) ribbons. Individual structural components are listed according to their association with heart diseases.

At the nucleoplasmic side of the nuclear envelope is the nuclear lamina (Figure 1). The nuclear lamina is a meshwork of type V intermediate filaments, composed of A- and B-type lamins [4,5]. The lamina confers biomechanical rigidity to the nucleus [6], transduces biomechanical forces from the cytoskeleton to the nucleoskeleton, and participates in chromatin organization and regulation [7–10]. The Lamin B1/B2 meshwork is closer to the inner nuclear membrane than the Lamin A meshwork [11], which lacks the farnesyl group.

Nuclear pore complexes, the nuclear lamina, and heart disease

In the early 2000’s, genetic and mechanistic connections were established between nucleoporins, nuclear lamins, and heart disease [12,13]. Nucleoporin gene mutations have been found in cardiac diseases, such as atrial fibrillation, cardiomyopathy, and heart failure [12,14,15] (Figure 1). More specifically, mutations in the gene encoding NUP155 have been implicated in atrial fibrillation and cardiac hypertrophy due to impaired nuclear import, chromatin structure, and gene expression [12,16]. Increased expression of Nup107 and decreased NUP35 protein levels were reported in post-myocardial infarction and cardiac arrhythmias in murine cardiomyocytes [17,18]. Protein levels of NDC1, NUP160, NUP153, and NUP93 are dysregulated in patients with heart failure and cardiomyopathy [14,15,19]. Mutations in NUP37, NUP43, and NUP188 are associated with atrial fibrillation and sudden cardiac death [20]. In summary, although alterations in nuclear pore complexes and nucleoporins cause cardiac diseases (Figure 1), the underlying cell biological mechanisms are largely unknown.

Mutations in nuclear lamin genes also have a recognized role in the development of cardiac pathologies [20–22]. In mammals, the LMNA gene encodes A-type Lamins, and the LMNB1 and LMNB2 genes encode Lamin B1 and Lamin B2, respectively [23–25]. Over 400 identified mutations of the Lamin A gene result in a spectrum of diseases, known as laminopathies [26]. Laminopathies affect multiple organ systems, including the cardiovascular system. With respect to the heart, the most common laminopathy is dilated cardiomyopathy, which causes weakening of the cardiac muscles [22] (Figure 1). Dilated cardiomyopathy due to Lamin A mutations has a poor prognosis, with a high rate of sudden death [27]. Fewer connections between B-type lamin mutations and human diseases have been identified [26].

The established associations between the nuclear pore complexes, nuclear lamina, and heart disease indicate the need for identifying and characterizing the biological mechanisms that are altered in nuclear lamina and nucleoporin-related cardiac pathologies. This knowledge could advance the understanding of heart disease and regeneration and could provide new therapeutic targets.

Relationship between lamin B2 levels, cardiomyocyte M-phase progression, and number of nuclear pore complexes

Cell cycle entry and division of mononucleated diploid cardiomyocytes is a fundamental mechanism of cardiac development and regeneration (Figure 2a). Maturation of cardiomyocytes in mammalian heart development is characterized by formation of bi- and multi-nucleated cells and polyploid nuclei, which are post-mitotic [1]. Thus, with the increasing prevalence of post-mitotic cardiomyocytes, cardiac regenerative capability decreases [1] (Figure 2b). As a result, neonatal mice can regenerate myocardium. However, this ability is lost in adults [28–30].

Figure 2. Cardiomyocyte maturation in rodents encompasses changes in the nuclear lamina and nuclear pore complexes (NPCs). (a) Changes during cardiomyocyte maturation include decrease in proliferative capacity, increase formation of binucleated cardiomyocytes by cytokinesis failure, increase in nuclear ploidy by karyokinesis failure, and increase in cardiomyocyte volume by cellular hypertrophy. (b) Decreased cardiomyocyte proliferation decreases the potential for scarless repair. (c) B-type Lamin mRNA and protein expression decreases, while A-type Lamin mRNA expression increases after birth and declines in cardiomyocytes in old mice. NPC numbers decrease with advancing cardiomyocyte maturation. The left black dashed vertical line marks birth and the right black dashed vertical line marks the beginning of adulthood, i.e., when heart and body size stop growing. Colored shapes represent approximations and are drawn to scale where quantitative results are available.

In a previous study, we discovered a connection between changes in the nuclear lamina and cell cycle progression and regeneration in post-mitotic cardiomyocytes [31]. More specifically, we have found that Lmnb2 expression in mouse decreases during cardiomyocyte maturation from the fetal to the adult post-mitotic stage [31] (Figure 2c). Decreased Lamin B2 levels lead to an increase in ploidy, a finding confirmed by Lmnb2 gene inactivation, which inhibited prometaphase-to-metaphase transition [31]. This provides a barrier for attachment of spindle microtubules to centromeres, reducing chromosome attachment and separation, in turn resulting in karyokinesis failure and generation of polyploid daughter nuclei. As polypoid cardiomyocytes are post-mitotic, Lmnb2 deletion in mice decreases cardiomyocyte cell cycle activity and reduces neonatal heart regeneration [31] (Figure 3). In addition to Lamin B2, other Lamins also show interesting development-related expression changes. During the transition from fetal to neonatal murine cardiomyocytes, Lmnb1 expression decreases and remains at very low levels throughout the lifespan [31], whereas Lamin A/C is upregulated after birth [31], remains high in young adults and declines in old mice (Figure 2c) [32], suggesting that B-type and A-type Lamins are differentially regulated during heart development.

Figure 3. Cardiomyocyte maturation and experimental decrease of Lamin B2 reduce nuclear import of signal molecules via decreasing nuclear pore complex numbers, thus reducing heart regeneration and adverse tissue remodeling. Immature (fetal and neonatal, top diagram) cardiomyocytes express Lamin B2 to ensure M-phase progression and formation of mononucleated, diploid daughter cells (2N). In fetal and neonatal cardiomyocytes, which have high nuclear pore complex (NPC) numbers, high levels of nuclear import of signal transducers facilitate activation of cell cycle genes, allowing myocardial regeneration. Experimentally inactivating Lamin B2 expression in immature cardiomyocytes leads to increased ploidy (4N) and decreased NPC numbers, resulting in lower cardiomyocyte proliferation and impaired myocardial regeneration in response to injury. Similarly, during cardiomyocyte maturation, Lamin B2 expression and NPC numbers decrease naturally. In adult cardiomyocytes (lower diagram), lower NPC numbers restrict nuclear import of signaling molecules, leading to reduced proliferation as well as decreased expression of remodeling genes, thus protecting adult cardiomyocytes against adverse heart remodeling at the cost of reduced heart regeneration [31,33].

We considered that the transition to the post-mitotic state may involve changes in the nuclear envelope and focused on nuclear pore complexes (NPC). Confocal microscopy has shown that fetal rat cardiomyocyte nuclei have an average of 1,900 NPCs, which decreases to 1,000 NPCs per nucleus in neonatal cardiomyocytes (48% decrease), and further to 700 NPCs per nucleus in adult cardiomyocytes (30% decrease) (Table 1), totaling a 63% reduction from fetal to the adult stage [33]. By comparison, electron microscopy of replicas of freeze-fractured nuclei has shown that the NPC density decreases by about 40% from 7.3 ± 0.5 NPCs/µm [2] in fetal nuclei to 4.6 ± 0.2 NPCs/µm [2] in neonatal nuclei [33]. The consistency in NPC densities quantified with two different imaging techniques demonstrated that cardiomyocytes decrease their NPC numbers during maturation (Figure 2c).

Table 1. Nuclear pore complex numbers vary between different species and cell types.

Cell type/line	Experimental condition	NPCs/nucleus	Imaging method	Reference
Rat heart muscle cells	Fetal (E14.5)	1,900	CM, STORM	[33]
	Neonatal (P2)	1,000
	Adult (P60)	700
Rat heart muscle cells	Neonatal (P2)		CM	[33]
	Mononucleated	1,300
	Binucleated	900
	Adult (P60)
	Mononucleated	800
	Binucleated	560
Mouse heart muscle cells	Neonatal (P2)	1,500	STORM	[33]
African frog heart cells	NA	3,500	EM	[65]
Northern Crested newt heart cells	NA	12,000	EM	[65]
Eastern newt lens	NA	16,000	EM	[65]
Rat liver	6 weeks	450	SIM	[71]
	6 months	450
	13 months	530
	2 years	450
Rat brain	6 weeks	430	SIM	[71]
	6 months	380
	13 months	330
	2 years	290
Normal rat kidney (NRK)	Early G1	2,000	Live CM	[102]
	Late G2	2,800
Human T-lymphocytes	Baseline	250–290	EM	[78]
	After stimulation	~1,800
HeLa	Baseline (G1)	2,000	EM	[78]
	End of S-phase	4,000
U2OS	NA	3,000–5,000	SIM	[56]
Yeast	NA	100	EM	[65]

Note: CM, confocal microscopy; STORM, stochastic optical reconstruction microscopy; EM, electron microscopy of freeze-fracture replicas; SIM, structured illumination microscopy.

During M-phase, nuclear lamins and NPCs undergo partial disassembly and re-assembly. Because disassembly of both is triggered by M-phase entry signals, we have investigated the mechanistical link between Lamin B2 levels and NPC numbers. We have found that 1) depleting Lmnb2 with siRNA decreases NPC numbers in neonatal rat cardiomyocytes, 2) overexpression of Lmnb2 increases NPC numbers per nucleus in neonatal rat cardiomyocytes, and 3) cardiomyocyte-specific Lmnb2 gene inactivation in mouse decreases NPC numbers in neonatal cardiomyocytes [33]. These findings demonstrate that Lamin B2 has an essential function in maintaining NPC numbers in cardiomyocytes, although the mechanistic underpinning of this regulation requires further investigation.

It is expected that a change in NPC number would impact nuclear transport, with a higher NPC density leading to a higher rate of nuclear import and export, and a lower density leading to a reduced rate of nuclear transport. As an example, the former relationship has been shown for the transition from maternal-to-zygotic transcription during zebrafish embryo development [34]. An increase in the number of functional NPCs raises the nuclear import of transcriptional regulators above the threshold required for initiating zygotic genome activation [34]. An example of the latter relationship has been observed in rat and mouse cardiomyocytes. Nuclei with lower NPC numbers exhibit less and slower active nuclear import of signaling molecules, including phospho-extracellular signal-regulated kinase (p-ERK), p38 mitogen-activated protein kinase (p38), nuclear factor-κB (NFκB), and c-Fos [33] (Figure 4). This is consistent with the observations that a) neonatal cardiomyocytes have lower nuclear import of fluorescently tagged ERK compared with fetal rat cardiomyocytes; b) cardiomyocytes with inherently lower NPC numbers due to natural cell-to-cell variation have lower nuclear import at the same developmental stage; c) compared to wildtype, nuclear import of ERK is lower in Lmnb2 depleted cardiomyocytes; and finally, d) nuclear import of ERK is reduced after depletion of an essential nucleoporin, Nup155 [33].

Figure 4. In cardiomyocytes, the number of nuclear pores complexes (NPCs) in the nuclear envelope is a developmentally regulated barrier for nuclear signal transduction. Extracellular signals activate receptors on the plasma membrane and initiate intracellular signaling cascades. These signaling molecules, including kinases and transcription factors, upon activation, are transported into the nucleus through NPCs and regulate downstream gene transcription, protein expression, and cellular response. Thus, via regulating signal transduction into the nucleus, the NPC number has a critical function in regulating the cardiomyocyte response to extracellular stimuli. This paradigm was demonstrated at the example of one G protein-coupled receptor (GPCR, endothelin receptor) and one receptor tyrosine kinase (RTK, fibroblast growth factor receptor), with four signaling pathways (ERK, p38, NFκB, c-Fos) and six transcripts (cJun, cFos, Nppa, Nppb, Myh7 and Myh6) [33].

Relationship between NPC number, heart regeneration, and cardiac stress tolerance

The regulation of heart regeneration and stress response by signal transduction pathways inside of cardiomyocytes draws connections to NPC number changes. The Extracellular Regulated Kinase (ERK)-pathway regulates cardiomyocyte cell cycle activity [35–38] and, in post-mitotic cardiomyocytes, stress signaling and cellular enlargement (hypertrophy) [39,40]. To direct these changes, active ERK enters the nucleus, where it alters gene transcription. As such, it is not surprising that cardiomyocytes with lower NPC numbers have less cell cycle activity. It is supported by the evidence that neonatal mice with lower NPC numbers due to Lmnb2 gene inactivation exhibit less myocardial regeneration [31].

On the other hand, several stress genes and hypertrophy factors were examined regarding downstream effects of decreased nuclear import with lower NPC numbers [33]. Stress and hypertrophy genes, including cJun, cFos, Nppa, Nppb, and Myh7, are all expressed at lower levels in cardiomyocytes with fewer NPCs [33] (Figure 4). These findings demonstrate that lower NPC numbers could determine the degree of active nuclear import of signaling molecules and that mature cardiomyocytes, via decreasing Lamin B2 level, use this mechanism to restrict signaling from extracellular stimuli [33].

In terms of consequences on cardiac functional outcomes, adult mice with fewer NPCs in their cardiomyocytes are more protected against adverse myocardial remodeling, when subjected to high blood pressure [33]. Specifically, compared to adult wildtype mice with normal NPC numbers, adult Lmnb2-deleted mice with fewer NPCs respond with less cardiac fibrosis and hypertrophy, when exposed to pressure overload that modeled high blood pressure [33]. The reduced maladaptive tissue remodeling results in better cardiac function. These observations in vivo support the findings in cultured cardiomyocytes: lower NPC numbers from Lmnb2-deleted cardiomyocytes restrict nuclear import of stress signals, leading to reduced cellular and cardiac hypertrophic response and increased resilience against environmental stress cues [33].

Other mechanisms underlying NPC-modulated cell differentiation have been reported. D’Angelo and colleagues have shown that increased Nup210 expression promotes myogenic and neuronal differentiation [41]. NUP210 is absent in proliferating non-differentiated cells. Nup210 knockdown prevents myogenesis as well as neuronal differentiation. Importantly, the NUP210-mediated differentiation appears to be independent of nuclear transport, as recruitment of NUP210 to NPCs does not affect nuclear transport rate or intranuclear accumulation of a red fluorescent transport reporter [41]. This is not surprising as NUP210 recruitment to NPCs happens late in the NPC assembly process and should not affect already established pore density or nuclear transport [42]. Rather, NUP210-elicited control over cell differentiation may function at the level of myogenesis gene induction [41]. Several other NUPs have been reported to play a role in regulating cell differentiation in Drosophila, mice and humans, including NUP133 [43], Seh1 [44], NUP358 [45], NUP155 [12] and NUP98 [46]. Although the role these NUPs could play in cardiomyocyte terminal differentiation is unknown, NPC-directed regulation of cell fate decisions is a promising research direction.

A role for lamins in controlling the numbers of NPC numbers and their positioning

One question to be addressed concerns the mechanisms by which different Lamin B2 levels could alter NPC numbers. Several possibilities have been suggested where NPC formation and maturation encompass several steps, including assembly of pre-pore complexes in the endoplasmic reticulum, trafficking of pre-pore complexes to the nucleus, insertion of pore complexes into the nuclear envelope, and maturation of NPCs embedded in the nuclear envelope [47–51]. Work by us and others has shown that deletion of specific nucleoporins, including Nup155 [33,52], Nup107 [53], Nup153 [54], Atchf1 (Elys) [55], and Tpr [56], decreases NPC density by disrupting assembly. Saliently, loss of any of these nucleoporins decreases nuclear pore complex formation, resulting in lower NPC numbers.

Little is known about the step of pre-pore complex trafficking from the endoplasmic reticulum into the nuclear envelope. The next step of NPC insertion into the envelope involves a non-nucleoporin protein, reticulum domain homology protein (REEP4), which regulates NPC density by binding to the nucleoporin ATCHF1 (ELYS) [57].

In addition, inappropriate cytoplasmic distribution of nuclear membrane proteins, including NUPs, can be caused by LMNA mutations in human muscle and has been modeled in Drosophila muscle [58], suggesting that Lamins are required for appropriate NPC insertion and retention. The distribution of NPCs within the nuclear envelope is not homogenous, and nuclear lamins may have a function in positioning and distributing NPCs in the nuclear envelope. Interestingly, B-type Lamins are generally associated with areas of the nuclear envelope enriched in NPCs, while A-type Lamins are associated with NPC-free islands [59]. Previous studies in mammalian cells [59], Caenorhabditis elegans [8], and Drosophila [60] have shown that nuclear lamins physically interact with NPCs and control NPC positioning. Furthermore, yeast cells, which lack a nuclear lamina, exhibit clustered and highly mobile NPCs [61,62]. Thus, physical interactions between the nuclear lamina and NPCs could provide the mechanisms by which formation and maintenance of both are dynamically coordinated.

Despite clear interactions between NPCs and lamins, studies investigating a possible effect of Lamin A mutations on nuclear transport are inconclusive. A study investigating LMNA mutations that cause skin disease and Hutchinson – Gilford progeria syndrome in humans supports this notion by showing that nuclear import is reduced by the mutant lamin A [63]. However, Ferri and colleagues have shown that mutation in LMNA did not alter nuclear transport [64]. In comparison, effects of Lamin B2 on nuclear transport are supported by the findings that deleting Lmnb2 reduced nuclear import while overexpressing Lmnb2 potentiates nuclear import in murine neonatal cardiomyocyte [33].

Along these lines of investigation, we have obtained evidence that decreasing Lamin B2 protein levels could decrease NPC numbers by reducing pre-pore complex trafficking from the assembly site in the endoplasmic reticulum to the nuclear envelope [33]. This conclusion is supported by evidence that a) the area density ratio of NPC in the nucleus relative to the perinuclear area is lower in Lmnb2-deleted cardiomyocytes, b) the co-localization of NPCs and endoplasmic reticulum markers is lower in Lmnb2-deleted cardiomyocytes [33]. However, future studies are needed to investigate the detailed mechanisms by which Lamin B2 could modulate NPC trafficking and maturation.

A generalizable association between higher NPC numbers and proliferation

We are left with the question of whether the association between higher NPC numbers and cellular proliferation is generalizable. It is clear that NPC numbers differ between species, tissues, and cell types (Table 1). There are vast differences in NPC numbers, with the highest in newt lens tissue with 16,000 per nucleus, while the lowest numbers were found in yeast with only 100 per nucleus [65]. NPC numbers also vary depending on the cell cycle status. On one end of the spectrum are post-mitotic cells, which downregulate NUP expression and maintain their assembled NPCs for their entire lifetime [66]. On the other end of the spectrum are cancer cells, many of which have increased NPC numbers and higher nuclear transport rates [67–69]. Moreover, inhibition of NPC formation selectively kills cancer cells, stops tumor growth, and even induces tumor regression, while normal cells manage to survive by cell cycle arrest [70]. This suggests that formation of new NPCs could be restricted to proliferative cells [66].

A newer study has quantified NPC changes during aging in different tissue types. NPC numbers decrease in aging rat brain cells, whereas in rat liver, NPC numbers are preserved with advancing age (2 years, Table 1) [71]. Even though both liver and brain cells are differentiated, liver cells with preserved NPC numbers during aging are proliferative [72], whereas most brain cells with decreasing NPC numbers during aging are not proliferative [73–75].

To address how proliferative cell types could maintain high NPC numbers, especially through the cell cycle, we have considered the two known NPC assembly pathways, interphase and post-mitotic. They differ in their timing with regard to the cell cycle. Interphase NPC insertion into the NE occurs during interphase and its activity appears constant throughout interphase [76]. The second pathway, post-mitotic NPC insertion, occurs during nuclear envelope reassembly around the daughter nuclei and at mitotic exit [76]. Given that the two pathways have different kinetics of nucleoporin incorporation into NPC [76], it is possible that maturating cardiomyocytes lack one or both assembly mechanisms, eventually leading to lower NPC density.

Proliferating cells go through the process of doubling and dividing DNA contents and organelles. To maintain NPC numbers during proliferation, NPC numbers should oscillate according to specific cell cycle phases. Indeed, it has been reported that the number of NPCs increases as cells proceed through the cell cycle and their density peaks in S-phase [77,78]. In further support, proliferative cells exhibit oscillating NPC numbers at different phases of the cell cycle [78], such as in T-lymphocytes and HeLa cells (Table 1), indicating dynamic NUP levels and NPC trafficking, insertion, and maturation. It is possible that such increases in NPC numbers before cell division may be absent in the cell cycles forming post-mitotic cells, like cardiomyocytes.

This raises questions about the mechanisms by which cardiomyocytes could decrease NPC numbers during maturation. Although limitation of NPC trafficking by decreased Lamin B2 expression may be one mechanism [33], other mechanisms could participate. Specifically, we have shown that several Nup mRNA and protein levels are lower in neonatal versus fetal cardiomyocytes and are further diminished to very low levels in adult cardiomyocytes [33]. These results suggest that Nup gene down-regulation and reduction in NUP assembly into the NPC macrostructure may contribute to cardiomyocyte maturation.

Future studies are required to explore the association between higher NPC numbers and proliferation. Whether the number of NPCs per nucleus is a reliable biological marker of high or low proliferative capacity in cardiomyocytes requires more direct investigation. The relation between experimentally manipulated NPC numbers and cell cycle activity is testable. With a combination of live cell imaging and super-resolution microscopy, NPC number change or lack thereof at different cell cycle phases may help to distinguish proliferative cells from cells heading toward terminal differentiation. This may be of particular significance for studies on post-mitotic and terminally differentiated cells. For instance, in heart regeneration studies, being able to identify cardiomyocytes that could proliferate would provide the possibility to identify undiscovered mechanisms of heart regeneration and to reveal novel therapeutic targets.

NPCs, nuclear lamina, chromatin, and cell-specific transcriptional programs

As stated above, a fundamental question in biology is how cells interpret chemical signals in their environment to carry out cellular functions. A particularly important question centers on how an external signal works to alter nuclear epigenetic information, which in turn modulates gene expression to elicit a cellular response. Besides canonical nuclear transport function, a large body of evidence now indicates functions for nucleoporins (NUP50 [79], NUP62 [80], NUP93 [81–85], NUP98 [46,86,87], NUP107 [80], NUP133 [43], NUP153 [14,54,80,82,88–90], NUP155 [16,91], NUP205 [85], NUP210 [41,92] AHCTF1 (ELYS) [93], SEH1 [94], TPR [95,96]) in gene regulation. These molecular functions occur through nucleoporin binding to chromatin, including at enhancers/super-enhancers [82,93], chromatin organization [90], via interactions with transcription factors [46], chromatin modifiers and chromatin architectural proteins [MEF2C (myocyte enhancer factor 2c) and TRIP6 [92], OLIG2 (oligodendrocyte transcription factor 2) [94], SOX2 (SRY-Box transcription factor 2) [88,89], BRD4 (Bromodomain-containing Protein 4) [84], BRD7 (Bromodomain-containing protein 7) [94], P300/KAT2B (PCAF, lysine acetyltransferase 2B) [14], HDAC4 (histone deacetylase 4) [16,81], KMT2A (lysine methyltransferase 2A) [80], PCR1 (polycomb-repressive complex 1) (RING1B, CBX7) [54], WDR82 (WD Repeat Domain 82) in SETDB1/Compass complex [87], CTCF (CCCTC-binding factor) [83,90], cohesin [90]], and through gene networks [91,94]. One output of these molecular functions is the regulation of cell- and lineage-specific gene programs to direct proliferation, differentiation, and cellular function (NUP50 [79], NUP98 [46], NUP133 [43], NUP153 [54,82,88–90] NUP155 [16,91] NUP160 [97], NUP210 [41], SEH1 [94]).

Several studies point to regulatory roles for nucleoporins in cell-specific transcriptional programming during cardiogenesis. In rat cardiomyocytes, physical interactions between NUP155 and histone deacetylase 4 (HDAC4; a chromatin remodeler for chromatin compaction and gene repression) block chromatin interaction and expression [16]. Upon induction of cardiomyocyte hypertrophic growth, HDAC4 is released, and cardiac genes gain nuclear peripheral location and activated expression, including clusters of sarcomeric and calcium-handling genes [16] (Figure 5). Induction of cardiomyocytes from embryonic stem cells indicates that NUP155 regulates a gene network, moving from a pluripotent transcriptional program to a cardiovascular developmental program, which is aberrantly regulated by NUP155 insufficiency [91]. In zebrafish, Nup210 recruits Mef2C (myocyte enhancer factor 2c; a skeletal and cardiac muscle transcription factor) to NPCs during myoblast differentiation, activating muscle structural, sarcomeric, and cell adhesion genes [92]. In cardiac progenitors, two regulators of genome organization, LBD1 (Lim-domain binding 1) and ISL1 (Lim homeobox 1), promote long-range enhancer-promoter interaction at Mef2c and other transcription factor binding sites, orchestrating a cardiac-specific transcriptional program [98].

Figure 5. Schematic of NPC/nucleoporin directed cardiac-specific transcriptional programming, leading to cellular differentiation. In immature cells (left), histone deacetylases (HDACs) block chromatin interactions and expression. Upon cardiomyocyte differentiation (right), HDACs are released, and clusters of cardiac-specific genes localize at the nuclear rim, where their expression is activated by transcription factors (TF). ONM, outer nuclear membrane; INM, inner nuclear membrane. Arrows (↱) indicate active transcription.

In addition to NPCs, Lamins play a repressive role in a cardiac-specific transcriptional program (Figure 6). In cardiomyocytes derived from human induced pluripotent stem (iPS) cells, LMNA mutations disrupt Lamin B1-associated domain (LAD) interactions with chromatin, leading to misexpression of genes outside the cardiac transcription program, including neuronal-specific genes [99]. Ectopic expression of noncardiac genes with changes in nuclear peripheral to nuclear interior chromatin compartments is also found in cardiomyocytes derived from human iPS cells [100]. Mechanistically, LMNA mutations result in loss of nuclear peripheral occupancy, reduction in repressive H3K9me2, and decreases in Lamin B1 enrichment, indicating that peripheral Lamin A has a distinct role in LAD cell-specific identity by silencing genes for alternative cell fates. In another model for cardiogenesis, mouse embryonic stem cells are differentiated to multipotent progenitors and then to cardiomyocytes. In progenitor cells, HDAC3 localizes with Lamin B in H3K9me2-enriched chromatin, silencing cardiac myocyte genes, including Mef2c [101]. Upon cardiomyocyte differentiation, cardiac-specific genes lose H3K9me2-LAD occupancy, while genes with non-cardiac function gain H3K9me2-LAD occupancy [101] (Figure 6). These data indicating a role for HDAC3 (and catalytic inactive HDAC3) for sequestering/tethering genes to the lamina to carry out repressive programming.

Figure 6. Schematic of Lamin-directed cardiac-specific transcriptional program. In cardiomyocytes, HDACs tether repressed genes with noncardiac function to the nuclear lamina. LMNA mutations disrupt Lamin B1 interactions with chromatin, leading to loss of nuclear rim positioning, decrease in repressive H3K9me2, and activation of genes (arrows) for alternative cell fates. TF, transcription factor; OMN, outer nuclear membrane; INM, inner nuclear membrane [99,101].

Important questions remain about the relationship between NPCs/nucleoporins and Lamins for chromatin regulation in cell/lineage-specific transcriptional programs. Are the functions between these molecules distinct? For example, do NPCs act as a scaffold for transcription hubs (including with enhancers/super-enhancers), while the lamina is a scaffold for gene repression (perhaps involving HDACs)? Does the interface between Lamin-chromatin and NPC-chromatin provide an important function? For example, does the interface possess boundary function, possibly in conjunction with architectural proteins, CTCF and cohesin? Do Lamins and NPCs physically interact to influence chromatin state, and if so, is there a hierarchical relationship between these molecules? For example, do Lamins regulate NPC number to transit from a pluripotent to mature, differentiated transcriptional states? How do the Lamin-mediated regulation on NPC numbers affect developmentally programmed cardiomyocyte terminal differentiation and can it be targeted to delay terminal differentiation in order to improve regeneration? In each of these scenarios, investigations are required to distill the mechanisms of action.

Summary, conclusions, and perspectives

The presented research studies show that maturation of cardiomyocytes into the post-mitotic state encompasses not only the transition to polyploidy but also changes in the nuclear lamina and NPCs, as well as cell-specific transcriptional programs. Studies in cardiomyocytes, liver, brain, and cancer cells suggest a general association between higher NPC numbers and higher proliferative capacity. The functional consequence of lower NPC numbers is lower nuclear import, resulting in altered cardiomyocyte response to extracellular signals. This plays out in adult mouse hearts as higher stress resiliency and in neonatal mice as lower cardiac regenerative capability.

The cardiomyocyte studies indicate new research directions about the processes that establish NPC numbers. Although cardiomyocytes require Lamin B2 for maintaining a higher NPC number, other proteins could be involved. The precise cellular mechanisms by which Lamin B2 functions in maintaining NPC numbers are unknown, including their function in regulating chromatin state. NPC numbers should be quantified, and the mechanisms altering NPC numbers should be characterized and compared between different cell types, degrees of maturity, stress, and aging states. Answering these questions will require further defining the processes of NPC insertion into the nuclear envelope, NPC maturation, and more complete characterization of changes of the nuclear lamina and NPC in aging and disease. Innovations in super-resolution, live, and correlative microscopy should advance progress in these directions, especially to define the interactions between Lamins and NPCs in different sub-cellular localizations. Assessing the functional consequences for nuclear transport will require time-resolved, i.e., live measurements of molecular transport through NPCs at the single nucleus level. Detailed data from different cell types could be integrated into a paradigm of regulated NPC numbers to modulate signaling across the NE.

Author contributions

All authors prepared the manuscript concept, outline, and draft. All authors reviewed and edited the manuscript.

Acknowledgments

We thank members of the Kuhn laboratory and Quasar Padiath (School of Public Health, University of Pittsburgh) for support, helpful discussions, and critical reading of the manuscript.

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Funding

Research leading to this paper was supported by the Richard King Mellon Foundation Institute for Pediatric Research (UPMC Children’s Hospital of Pittsburgh), NIH grants R01HL151415, R01 HL151386, and R01HL155597, and a grant from the UPMC Aging Institute (to B.K.) and R01HD101574 and R01HD109347 (to M.R.W.M.). This project was supported by an NIH-Training Grant (to Y.L., T32HL129949).