Introduction

One of the main distinctions between eukaryotic and prokaryotic cells is that the former can compartmentalize materials and biological processes within membrane-bound regions known as organelles. Perhaps the most well studied of these compartments is the nucleus, a double-membrane-bound organelle that contains the genetic material of the cell. One of the main functions of the nuclear envelope, the boundary between the nucleus and the cytoplasm, is to maintain separation between DNA and cytoplasmic processes. This separation allows for proper maturation of newly transcribed RNA prior to its translation [1], as well as protection of DNA from cytoplasmic damaging agents [2].

Though strict compartmentalization must be maintained, material transfer between the nucleus and cytoplasm is essential. Newly transcribed RNA (ribosomal, transfer, messenger, and non-coding) must be granted passage out of the nucleus, while nucleotides and newly translated nuclear proteins need to enter. Nuclear pores serve as numerous interspersed regulated tunnels between the nucleus and cytoplasm. Each nuclear pore is maintained by a nuclear pore complex (NPC), a massive structure composed of 33 known protein components, called nucleoporins (Fig. 1A), most of which are present in octagonal symmetry around the pore [3]. In aggregate, the total number of nucleoporins that compose the NPC is over 400, making each pore one of the largest protein complexes in the cell [3]. Notably, the half-lives of some nucleoporins are measured in years, an intriguing characteristic when one examines the biological basis of late onset human diseases [4]. This structure, as a whole or on a subunit basis, is responsible for the selective passage of materials in and out of the nucleus [5, 6].

Fig. 1
figure 1

The nuclear pore complex (NPC) is a massive structure comprised of 33 known protein components organized into five main structural domains–the cytoplasmic filaments (yellow), the inner and outer nuclear rings (green), the transmembrane components (purple), the nuclear core components (blue), and the nuclear basket (red) (A). Several nuclear pore components have been found to be disrupted in cases of ALS (colored in black)–Gle1, GP210, NDC1, Nup107, Nup133, Nup153, Nup50, Nup62, Nup98, POM121, RanBP2, and Tpr (B). Nup98 and Nup62 are disrupted in Alzheimer’s disease (C). Cases of Huntington’s disease show disruption of Nup62, Nup88, and Tpr (D). Miscellaneous juvenile neurodegenerative disorders also display disruption in nuclear pore components including ALADIN, Gle1, Nup214, Nup62, Nup88, and RanBP2, many involving a direct mutation in the gene encoding that NPC component (E)

Full size image

Individual nucleoporins also play roles in a variety of nuclear and cytoplasmic processes, independent of their function in maintaining proper transport across the nuclear pore [7,8,9]. Nup98, for example, interacts with DHX9 in the nucleoplasm to regulate the transcription and splicing of certain mRNA transcripts [10]. SEH1, another nucleoporin, performs a cytoplasmic role in the regulation of MTORC1 as part of the GATOR2 complex [11, 12].

Here, we discuss the effect of disease-related nucleoporin mutations on NPC selectivity. Additionally, we examine numerous neurological diseases that present significant alterations in nuclear pore structure and function.

Disease-Related Mutations in Nuclear Pore Complex Components

Many studies over the past few decades have focused on disruptions to the NPC in neurodegeneration. Oftentimes, components of the NPC are mislocalized or dysregulated as a result of an unrelated or even an unknown genetic mutation. However, mutations in genes that directly encode for NPC components can also lead to dysregulation of the nuclear pore and sequential neurodegeneration in both human patients and cell and animal models of disease (Fig. 1E). Below, we review several nucleoporin (Nup) mutations and the associated cell-specific disorders.

RANBP2 (Nup358)

Genetic acute necrotizing encephalopathy (ANE1) is a rare disease resulting from viral infections in patients carrying mutations (most commonly a missense mutation, c.1880CT, p.Thr585Met [13]) in RanBP2 [14], though the precise mechanism of neuronal toxicity is unknown. ANE1 presents as an acute encephalopathy followed by an onset of neurological symptoms including deteriorating consciousness, progression to coma, and possible seizures. RanBP2 (also known as Nup358) is a major component of the cytoplasmic filaments of the NPC. RanBP2 regulates the Ran GTPase cycle which is necessary for nuclear import and export via the active transport receptors, importin β and exportin-1. Ran GTPase activity is required for the release of these receptors from Ran-GTP to initiate and terminate nuclear import and export, respectively [15,16,17,18,19,20]. Knockout of RanBP2 in mouse spinal cord motor neurons leads to the subcellular disorganization of the known RanBP2 substrates importin β, exportin 1, Ran GTPase, and Ran-GTP [21].

In control mice, importin β and exportin 1 are found predominantly at the nuclear rim, while Ran GTPase and Ran-GTP are found largely in the cytoplasm. However, in the RanBP2 knockout mice, importin β becomes sequestered to the nuclear compartment and is lost at the nuclear rim; exportin 1 is completely lost from the nuclear rim; and expressions of Ran GTPase and Ran-GTP are broadly distributed between the nucleus and cytoplasm. RanBP2 ablation in Thy+ neurons in the central nervous system and spinal cord is sufficient to mimic the disease progression of ALS with the mice displaying hypoactivity, hind limb paralysis, respiratory distress, and premature death. Interestingly RanBP2 is also necessary for cell viability in other cell types. For example, loss of the Ran GTPase binding domains 2 and 3 (RBD2/3) of RanBP2 leads to cone photoreceptor degeneration [22]. Whether these findings relate to the pathogenesis of ANE1 is unclear, though they provide interesting insight into the effect of neuronal RanBP2 loss or dysfunction.

Gle1

Gle1 mutations have been discovered in patients with sporadic and familial ALS [23]. The first mutation identified was a nonsense mutation that introduced a premature stop early on in Exon 2 of Gle1, while the other mutation was a splice site mutation that caused a shift in the reading frame and the replacement of the last 44 amino acids of the protein with 88 different amino acids. The NPC-associated factor Gle1 is a critical component of the mRNA export machinery [24, 25]. Humans have two isoforms of Gle1–Gle1A and Gle1B–that differ only in their 3’ untranslated regions, with Gle1B being more highly expressed and localized to the nuclear rim through an interaction with CG1 and Nup155 [26, 27]. In vitro analysis of these mutations revealed two distinct mechanisms of protein disruption depending on the mutation [23]. The nonsense mutation leads to a loss of the Gle1 mRNA via nonsense mediated decay and an overall loss of Gle1 protein. Alternatively, the splice site mutation, with the addition of the amino acids in its C-terminus, is unable to localize properly to the NPC and instead becomes biased towards the cytoplasm.

Interestingly, knockdown of Gle1 in zebrafish embryos using antisense morpholinos causes morphological anomalies in the jaw as well as cell death in the spinal cord, both phenotypes which are reminiscent of symptoms in patients with lethal congenital contracture syndrome-1 (LCCS1) [23, 28]. LCCS1 is an embryonic degenerative disease marked by the loss of the ventral spinal cord and anterior horn motor neurons [29] and is caused by numerous identified mutations in Gle1, including a splice site mutation and several missense mutations [30]. Distinct from the Gle1 mutations that cause ALS, Gle1 mutations in LCCS1 do not alter the localization of the protein from the NPC [31]. Instead, the disease-causing mutation inserts a proline-phenylalanine-glutamine group into the coil-coil domain of Gle1 which prevents its ability to oligomerize and therefore inhibits its role in mRNA export. A closely related disease, lethal arthrogryposis with anterior horn cell disease (LAAHD), has similar motor neuron loss in the anterior horn of the spinal cord and is also caused by a mutation in Gle1 [32].

ALADIN

The nuclear pore component, ALADIN, is anchored to the cytoplasmic side of the NPC through its interaction with NDC1, a nucleoporin required for NPC assembly [33, 34]. A myriad of mutations, including a number of point mutations, frameshift mutations, and nonsense mutations, in the AAAS gene that encodes ALADIN cause the disease triple A syndrome. Triple A syndrome is a multisystemic disorder that includes progressive neurodegeneration and neural impairments [35]. In cells harboring a triple A syndrome-causing AAAS mutation, ALADIN itself predominantly mislocalizes to the cytoplasm; however, there are no other overt morphological abnormalities to the nuclei, nuclear envelope, or nuclear pore complex [36,37,38]. How then does the NPC contribute to disease pathology in triple A syndrome? Fibroblasts from triple A syndrome patients have increased sensitivity to oxidative stress [39]. To further elucidate how mislocalized ALADIN could cause this susceptibility to oxidative stress, another group discovered that there is a decrease in the nuclear accumulation of import cargoes reliant on the karyopherin-α/β import pathway in triple A syndrome fibroblasts [40]. Additionally, these cells have increased sensitivity to oxidative stress and a decreased ability to repair damaged DNA, likely due to the negatively affected nuclear import. Indeed, ALADIN has been shown to interact with and regulate the translocation of ferratin heavy chain protein (FTH1) [41], a nuclear protein that has a protective role in oxidative damage of DNA [42,43,44]. Triple A syndrome patient fibroblasts display a complete absence of nuclear FTH [41]. It is therefore likely that AAAS mutations cause mislocalization of ALADIN, inhibit its ability to import FTH1 to the nucleus, and lead to increased stress from oxidative damage and eventual cell death.

Nup214, Nup88, and Nup62

While our understanding of how the NPC contributes directly to cell death in neurodegeneration is expanding, there are still many cases in which the mechanism underlying disease pathology is incompletely characterized, despite our knowledge of the genetic cause. As mentioned above, mutations to RanBP2 are causative of ANE1 [13, 14, 45]. How the viral infection exacerbates the potential nucleocytoplasmic transport defects due to a lack of RanBP2 function is poorly understood. Similarly, missense and frameshift mutations in Nup214, another component of the NPC cytoplasmic filaments, are causative of acute febrile encephalopathy [46, 47], a disease that is similar to ANE1 and is also very poorly understood mechanistically. A familial missense mutation causing infantile bilateral striatal necrosis (IBSN) was mapped to Nup62 [48]. How the mutation in Nup62 contributes to the disease is unknown; however, it could be through its role in maintaining chromosomal integrity [49]. Mutations in yet another member of the NPC cytoplasmic filaments, Nup88, cause fetal akinesia deformation sequence (FADS), a disease characterized by impaired fetal movement [50]. In a nup88-null zebrafish line, there is reduced association of Nup62 with the NPC as well as a loss of interaction between Nup88 and Nup214. While these impaired interactions may not be the sole cause of FADS, it is possible that it contributes to disease.

Huntington’s Disease

Huntington’s disease (HD) is a disorder with three distinct sets of symptoms including motor deficits, cognitive decline, and psychiatric abnormalities due to progressive degeneration mainly in the striatum and basal ganglia [51], although other distant brain regions also undergo alterations. HD is caused by a CAG trinucleotide repeat expansion in the huntingtin gene (HTT) which leads to a polyglutamine (polyQ) expansion in the N-terminal of the Huntingtin protein [52]. An interesting proteomics study in 2001 discovered that many NPC proteins, particularly Nup62, co-aggregate with polyQ aggregates [53]. These results initiated a hypothesis in which HD mutations may confer toxicity through altered NPC function (Fig. 1D).

Indeed, a subset of Nups (Nup62, Nup88, and RanGAP1) colocalize with mutant huntingtin (mHTT) intracellular aggregates in the striatum and cortex of an HD mouse model which is consistent with observations in postmortem tissue from HD patients [54, 55]. In the mouse cortex and striatum, mHTT exacerbated age-related nuclear dysfunction, including disruption of the nuclear envelope and accumulation of intranuclear mRNA [55]. As the mice age, accumulation of RanGAP1 and mHTT shifts to the perinuclear and cytoplasmic regions of the cell which is also observed in HD patient cortices. This cytoplasmic accumulation of mHTT could cause additional mislocalization of proteins containing disordered and low complexity sequences in neurons, including multiple factors of the nuclear transport machinery, similar to what has been demonstrated in immortalized cell line studies [56]. Whether such mislocalization of these NPC components leads to functional alterations in the NPC is yet uninvestigated.

Additionally, induced pluripotent stem cell (iPSC)-derived neurons from HD patients also display mislocalized RanGAP1 and Nup62 as well as a significant reduction in the nucleocytoplasmic ratio of endogenous Ran, suggesting deficient active transport between the nucleus and cytoplasm in HD. This observation was replicated in primary cortical neurons transduced with mHTT as well as in a Drosophila HD model–several Nups, including RanGAP1, Nup62, and Nup88, are mislocalized and these cells display deficits in nucleocytoplasmic transport. Overexpressing Ran and RanGAP1 confer neuroprotection in both of these systems [54] suggesting that restoring proper transport could be an effective way to mitigate mHTT-induced degeneration.

It is clear that mHTT causes disruptions in NPC components, but evidence connecting these observations to cellular toxicity is lacking. Several studies suggest that mHTT accumulation in the nucleus impairs nuclear export of mRNA and protein. mHTT was found to have preferential binding to RanGAP1 and the mRNA export factor RAE1 [57] which could account for the mRNA accumulation observed in the nucleus in previously described HD models [54, 55]. Interestingly, the N-terminal of wild-type HTT interacts with Tpr, another member of the NPC that facilitates mRNA and protein export [58]. The polyQ expansion in mHTT was shown to cause reduced nuclear export of the protein, accumulation within the nucleus, and subsequent nuclear membrane distortions. mHTT was also found to have increased phosphorylation at Serine-16 in its N-terminus in striatal neurons of an HD mouse model [59]. The increased phosphorylation promotes nuclear accumulation and decreases the interaction between HTT protein and Tpr, possibly accounting for the nuclear accumulations of both mRNA and mHTT protein. The N-terminal of HTT also includes a nuclear export signal which allows the protein to be transported out of the nucleus via exportin 1 [60]. It is therefore unclear whether the nuclear export of cleaved HTT is entirely functional, as the N-terminal fragment of HTT is well exported when overexpressed in HEK293T cells [60] but is also observed within intranuclear accumulations in mouse AD models [61].

Alzheimer’s Disease

Alzheimer’s disease (AD) is a progressive degenerative disease that contributes to 60–80% of the dementia cases [62]. Infamous hallmarks of AD include aggregation of amyloid-beta (Aβ) protein and/or phosphorylated Tau in neurons [63]. Early electron microscopy studies from AD biopsies revealed neurons containing neurofibrillary tangles (NFTs)–insoluble protein tangles consisting mostly of misfolded Tau–have irregular nuclear structure with the NFTs often associated with the nuclear lamina and nuclear pore complexes [64]. Further investigation of AD biopsies and postmortem tissues revealed that irregular nuclear pores often associate with NFTs [65]. While there are no apparent differences in Nup expression or distribution in these samples, hippocampal neurons with and without NFTs have increased accumulation of nuclear transport factor 2 (NTF2) which transports cargoes from the cytoplasm to the nucleus. Additionally, importin α1 mislocalizes to cytoplasmic inclusions in the hippocampus of AD patients [66]. Previous studies revealed inhibition of classical nuclear localization sequence (NLS)-mediated cargo transport via importin α1 as a result of cell stress [67, 68] and it is therefore likely that the importin α1 accumulation is a result of other unknown factors causing cell stress.

Drosophila models of Tau pathology highlighted the importance of nuclear integrity in the context of neurodegeneration. Transgenic expression of human Tau in Drosophila causes nuclear envelope invaginations and a significant decrease in the Drosophila B-type Lamin protein [69], and interestingly similar pathology is seen in AD brain samples [70]. Most disrupted nuclei are associated with extra-nuclear pathogenic Tau aggregates, suggesting a possible correlation between Tau accumulations and disruption of nuclear structure. This Tau-induced Lamin dysfunction impairs mRNA trafficking [71] and is sufficient to drive cell-cycle activation and subsequent apoptotic death [69]. While these studies provide us with a platform to further our understanding of nuclear dysfunction in tauopathies, the mechanisms connecting mRNA and protein trafficking directly to neurodegeneration are still poorly understood.

Using a combination of human, mouse, and cell models of AD, it was recently shown that phosphor-Tau can interfere with NPC integrity and nucleocytoplasmic transport through a variety of mechanisms [72]. As such, a few NPC components have been shown to specifically be altered in AD (Fig. 1C). Hippocampal tissue of AD patients displays a shallower nuclear-to-cytoplasmic Ran gradient than tissue from unaffected individuals. In primary neuron cultures, pathogenic Tau is sufficient to disrupt the Ran gradient. Using nuclear import and export reporters in a live cell assay, Eftekharzadeh and colleagues also demonstrated that neurons containing Tau protein show disruption in both nuclear import and export. In addition to an altered nucleocytoplasmic Ran gradient, Nup98 is mislocalized to the cytoplasm of neurons that contain misfolded Tau and phosphor-Tau in AD postmortem hippocampal tissue as well as in brain tissue from Tau-overexpressing transgenic mice. Nup98 appears to colocalize with the NFTs in the cytoplasm through a direct interaction with Tau and this phenomenon is specific to Nup98, and to a lesser extent Nup62, while other Nups remain unaffected. The mouse lines used in this study express the human Tau transgene that can be suppressed by doxycycline (DOX). Treating these mice with DOX leads to a reduction in soluble Tau, relatively few NFTs, and interestingly a rescue of Nup98 expression at the nuclear membrane, indicating a role for Tau in facilitating Nup98 localization. In primary mouse neuronal cultures, Nup98 facilitates Tau aggregation and also co-aggregates with Tau via a direct interaction [72]. Additionally, Nup98 and Tau colocalize to cytoplasmic accumulations in postmortem AD brain samples. Another group discovered that the repeat binding domain of Tau mediates its interaction with Nup98 [73]. Interestingly, the phosphorylation of Tau, but not its oligomerization, facilitates the association with Nup98, and once bound, the interaction is quite strong. Although unconfirmed, it is likely that the accumulation of phosphor-Tau and Nup98 at the outer surface of the nuclear membrane is a significant contributor to decreased nucleocytoplasmic transport.

ALS/FTD

Amyotrophic lateral sclerosis (ALS) is a progressive and invariably fatal neurodegenerative disease that affects upper and lower motor neurons [74]. Around 10% of cases are familial (fALS) via inheritance of a mutation in one of many known genes, while the remainder are sporadic (sALS), meaning the genetic cause is unknown [74, 75]. The first reported defect in NPC function arose from three independent investigations into the pathogenesis of the C9ORF72 mutation, a repeat expansion that accounts for a significant portion of fALS and sALS [74, 76,77,78]. These groups observed disruption of the Ran GTPase gradient in fixed cells [76] and reporter-based dysfunction of nuclear import [76], and identified numerous components of the nuclear transport machinery that modified toxicity resulting from the mutation [76,77,78]. Dipeptide repeat proteins (DPRs), one type of toxic product resulting from the C9ORF72 mutation, were subsequently linked to alterations in nucleocytoplasmic transport [79, 80]. Two groups demonstrated that arginine-rich DPRs bind to and inhibit the function of the nuclear import receptor, importin β [79, 80]. These experiments were performed using a variety of in vitro permeabilized cell and biochemical assays with synthetic DPRs. Interestingly, an independent investigation that utilized transiently overexpressed DPRs concluded that arginine-rich DPRs had no effect on nuclear transport function in numerous cell types, including Hela Kyoto, SH-SY5Y, and human-induced pluripotent stem cell-derived motor neurons (iPSNs) [81]. This discrepancy may be attributed to the identity of the nuclear localization signals present within the tested nuclear transport cargoes and reporters, and highlights the need to test nuclear transport of multiple classes of cargoes rather than relying on a single reporter. While all three groups assessed various aspects of nuclear transport, none of these investigations were focused on the effect of DPRs on nuclear pore complex structure itself.

While the C9ORF72 mutation is the most common known cause of fALS, mutations in FUS, TARDBP, and SOD1 are also causative of this disease. Pathology arising from these mutations has also been linked to deficits in the function of the nuclear pore complex. For example, overexpression of disease-linked TDP-43 (encoded by TARDBP) variants in Drosophila melanogaster photoreceptors or N2a cells leads to abnormal nuclear envelope morphology, while iPSNs harboring endogenous TDP-43 mutations demonstrate reduced nuclear import of an NLS-tdTomato-NES reporter [82]. Overexpression of various nucleoporins (Nup50, Nup93, Nup98-96, Nup107, and Nup214) rescue photoreceptor and motor defects in their fruit fly model; however, their specific neuroprotective function in this setting remains unclear [82]. Additionally, in female mice expressing an ALS-linked variant of SOD1, various nucleoporins and RanGAP1 exhibit altered spatiotemporal distributions compared to wild-type littermates [83]. While this may indicate at a potential role for altered NPC function in SOD1-ALS, no functional assays were performed to demonstrate whether these alterations led to any functional consequence. While minimal defects in the Ran GTPase gradient were observed in iPSNs harboring FUS mutations, there were evident abnormalities in nuclear lamina partially as a result of Nup62 association with FUS [84]. This interaction between FUS and Nup62 has the potential to form insoluble aggregates in the cytoplasm in vitro, so it is therefore possible that these aggregates sequester and trap Nup62 and other Nups in the cytoplasm, further disrupting the nuclear pore structure. Although numerous groups have reported association of various nucleoporins within aggregates, stress granules, or other cytoplasmic accumulations [80, 82, 85], it was not immediately clear whether this association was subsequent to loss of these nucleoporins from the NPC itself. A recent study demonstrated that in iPSNs harboring the C9ORF72 mutation, eight specific nucleoporins have decreased expression within the NPC (TPR, Nup98, NDC1, POM121, Nup107, Nup133, Nup50, GP210) [86]. This seminal work utilized structured illumination microscopy to comprehensively assess the presence of the 23 nucleoporins within NPCs [86]. Impressively, the same eight nucleoporins are depleted from NPCs in postmortem motor cortex [86], demonstrating the power of endogenous model systems. This loss of nucleoporins is linked to pathogenic hexanucleotide repeat RNA that results from the C9ORF72 mutation and aberrant degradation of nuclear pore components via an ESCRT-III pathway [87, 88], rather than sequestration in stress granules or other cytoplasmic accumulations [86].

Interestingly, loss of certain nucleoporins has also been observed in up to 90% of sporadic ALS patient-derived iPSNs. In iPSNs derived from sporadic ALS patients, five nucleoporins are consistently depleted from the NPC (Nup50, Nup153, TPR, POM121, Nup133) [87]. The loss of these nucleoporins has similarly been linked to degradation via an ESCRT-III pathway. While the primary injury resulting in this aberrant degradation is theoretically different in C9ORF72 ALS and sALS (only ~10% of sALS cases harbor the C9ORF72 mutation) [74], it is possible that the initiation of ESCRT-III overactivity and subsequent loss of nucleoporins occurs via a common event. Future investigations into the similarities between NPC dysfunction in fALS and sALS may yield interesting and highly impactful results.

Importantly, this NPC injury in sALS iPSNs appears to result in the loss of nuclear TDP-43 and its RNA metabolism function, as substantiated by an inverse correlation between aberrant ESCRT-II function and TDP-43 nuclear depletion [87]. Nuclear clearance and cytoplasmic accumulation of TDP-43 is a common pathological hallmark of ALS [89, 90]. Approximately ninety-seven percent of postmortem cases exhibit motor neurons with reduced nuclear and increased cytoplasmic presence of TDP-43, though only a small percentage of neurons within each case demonstrate this pathology [74].

Due to the prominence of this protein pathology, recent investigations have focused on understanding the downstream effects of TDP-43 nuclear clearance. Indeed, recent RNA sequencing studies have uncovered hundreds of transcripts that are altered in abundance upon artificial global TDP-43 depletion, the most presently notable of which is Stathmin-2 [91, 92]. Although Stathmin-2 plays a significant role in the regeneration of axons [91, 92], it is yet unclear whether therapeutic targeting of a single dysregulated TDP-43 target will prove effective in patients.

On the other hand, reversal of NPC injury itself may prove to prevent or diminish TDP-43 nuclear clearance in affected motor neurons, effectively restoring all TDP-43 function. Antisense oligonucleotides (ASO) designed to deplete CHMP7, the identified pathogenic component of the ESCRT-III pathway in ALS, completely repaired the NPC injury and normalized aberrant TDP-43 function [87]. For these reasons, CHMP7 ASO may be an attractive candidate therapy in sporadic ALS.

Discussion

Nuclear structure and function are vital for the proper segregation of nuclear and cytoplasmic processes and materials. The NPC maintains precise passage of materials between these two compartments. When the NPC is disrupted functionally or structurally, a common consequence, especially in the context of neurons, is cellular degeneration.

NPC function can be disrupted in a variety of ways, the most obvious of which is via mutations in NPC components themselves. We discussed that mutations in different nucleoporins lead to distinct cellular consequences. Mutations in RANBP2 and GLE1, for example, have a significant impact on nuclear transport function, while those in ALADIN appear to induce toxicity via increased sensitivity to oxidative stress and DNA damage. This emphasizes that despite its main role as a nucleocytoplasmic gate, the NPC consists of numerous individual proteins, each of which have functions independent of their roles within the NPC itself. Such diversity in function of the numerous NPC components makes it understandably difficult to determine if (and how) mislocalization of these proteins leads to cellular toxicity.

Yet, it remains tempting to presume that altered cellular distribution of one or even many nucleoporins necessarily leads to altered NPC function as a whole. For example, one group overstates their observations as “nuclear pore destruction” in a single postmortem case of juvenile-onset ALS, entirely substantiated by altered nuclear rim staining of Nup62 [93]. In the burgeoning field of NPC-centric investigations of neurodegeneration, high-quality imaging and reproducibility of such phenotypes will be required to substantiate anecdotal observations, e.g., functional impact of mislocalized or absent NPC components on their distal functions, as well as their contribution to NPC function as a whole.

Numerous groups have utilized fluorescent reporter-based assays to quantitate nuclear import and export function [76, 84, 86]. These “shuttle reporters,” typically a GFP- or RFP-based fluorophore appended to an NLS and NES, serve as exogenous indicators of transport into and out of the nucleus [94]. Introduction of such reporters into a perturbed system (e.g., in iPSN harboring a disease mutation) can allow investigators to compare import and export rates across various conditions. However, not all shuttle reporters interrogate both nuclear import and export function. A shuttle reporter, such as the S-GFP reporter [85], with a strong NES and weak NLS, for example, will be mostly cytoplasmic under baseline conditions. This provides significant dynamic range to assess alterations in nuclear export, but alterations to nuclear import function may not be assessable. Meanwhile the S-tdTomato reporter, used in the same study [85], can effectively report on the opposite (it is useful for monitoring changes in nuclear import, but not export). Therefore, it is apparent that shuttle reporters should be carefully chosen to answer a specific question about NPC function, rather than as a catch-all attempt to interrogate multiple pore functions.

As reliable methods to investigate NPC structure and function have become more available and widely used by the neurodegeneration community, we have learned that loss of at least one nucleoporin is differentially associated with alterations to others. Reduction of neuronal POM121, for example, directly leads to a reduction in the NPC presence of 6 or more other nucleoporins [86]. Meanwhile, mislocalization of ALADIN in the context of triple A syndrome leaves the rest of the NPC structurally unaffected [36,37,38]. Future investigations into mislocalized NPC components would do well to test for inter-nucleoporin effects rather than assuming that observed alterations are restricted to a single component of this massive protein complex.

Conclusion

Extensive literature supports a role for NPC dysfunction in the pathogenesis of neurodegenerative diseases ranging from Huntington’s disease to Alzheimer’s disease to amyotrophic lateral sclerosis. However, with the exception of ALS, it is still unclear whether NPC dysfunction is a toxicity-initiating event, or a consequence of some other primary injury. Regardless of whether the event is disease initiation or downstream response to injury–the loss of the NPC and nucleocytoplasmic transport itself is an injurious event and may also be a future target for repair. Additional studies that comprehensively and mechanistically interrogate the structure and function of the NPC in relevant disease models will be crucial to further our attempts to identify useful therapeutic targets in these diseases.