Nuclear pore complex and nucleocytoplasmic transport disruption in neurodegeneration

Abstract

Nuclear pore complexes (NPCs) play a critical role in maintaining the equilibrium between the nucleus and cytoplasm, enabling bidirectional transport across the nuclear envelope, and are essential for proper nuclear organization and gene regulation. Perturbations in the regulatory mechanisms governing NPCs and nuclear envelope homeostasis have been implicated in the pathogenesis of several neurodegenerative diseases. The ESCRT-III pathway emerges as a critical player in the surveillance and preservation of well-assembled, functional NPCs, as well as nuclear envelope sealing. Recent studies have provided insights into the involvement of nuclear ESCRT-III in the selective reduction of specific nucleoporins associated with neurodegenerative pathologies. Thus, maintaining quality control of the nuclear envelope and NPCs represents a pivotal element in the pathological cascade leading to neurodegenerative diseases. This review describes the constituents of the nuclear-cytoplasmic transport machinery, encompassing the nuclear envelope, NPC, and ESCRT proteins, and how their structural and functional alterations contribute to the development of neurodegenerative diseases.

Graphical Abstract

In this review, the authors detail and discuss recent literature documenting nuclear pore complex, nucleocytoplasmic transport, and nuclear envelope alterations in neurodegenerative disease.

Introduction

The nuclear pore complex (NPC) is a pivotal player in the regulation of material exchange between the nucleus and cytoplasm. This sophisticated structure is made up of around 30 nucleoporins (Nups) that are organized into subcomplexes, forming eight modular units in total [1–7]. The NPC monitors the entry and exit of a range of cargos, including proteins and transcripts, into and out of the nucleus. Many Nups are responsible for providing structural integrity, scaffolding support, and anchoring of the NPC into the nuclear envelope. In contrast, roughly one third of Nups contain phenylalanine-glycine (FG) repeats to form the permeability barrier of the NPC. This barrier of intrinsically disordered FG repeat Nups is maintained by weak hydrophobic interactions between FG repeats themselves and functions to prevent or significantly slow the passive diffusion of larger cargo molecules (>60 kDa) into the nucleus [8–14]. However, this is not a strict size exclusion cutoff and can vary due to a number of factors including FG-Nup density and charge within the central channel, overall Nup composition within NPCs, and perhaps even the properties of the macromolecules undergoing passive transport themselves [12, 15]. It is unknown at this time whether the “rules” governing passive diffusion through the NPC can also vary amongst cell types. On the other hand, the well-defined active nucleocytoplasmic transport (NCT) process through the NPC is powered by the RanGTP-RanGDP gradient across the nuclear membrane. This gradient is established by the RanGEF RCC1, the RanGTPase activating protein RanGAP, and the nuclear transport factor NTF2. Specifically, RCC1, bound to chromatin ensures nuclear RanGTP levels remain high and RanGAP localized in the cytoplasm and at the cytoplasmic face of the NPC functions in cooperation with the RanGTP binding proteins RANBP1 and RANBP2 to drive GTP hydrolysis thereby maintaining RanGDP levels in the cytoplasm. This hydrolyzed RanGDP complex is transported back into the nucleus via NTF2 thereby facilitating interactions with RCC1 [13, 16–25]. To provide specificity and directionality of active nucleocytoplasmic transport, nuclear transport receptors (NTRs), broadly termed importins and exportins, recognize specific nuclear localization (NLS) and nuclear export sequences (NES) and interact with FG Nups to direct the flow of cargos through the NPC [13, 23, 24, 26–31].

Recent studies have highlighted alterations to the NPC, NTRs, and NCT as early and significant contributors to neurodegenerative disease pathophysiology [32]. Perhaps one of the earliest indications of disrupted nuclear-cytoplasmic compartmentalization in neurodegenerative disease stems from landmark studies identifying nuclear clearing and cytoplasmic mislocalization/aggregation of the nuclear RNA binding protein TDP-43 as a prominent pathology hallmark in neurodegeneration [33, 34]. More recently, evidence more directly implicating the NPC and NTRs in impaired nuclear-cytoplasmic compartmentalization in neurodegeneration has been reported. Specifically, reduction of specific Nups from the NPC, cytoplasmic aggregation of Nups and NTRs, as well as nuclear envelope abnormalities have all been documented in Amyotrophic Lateral Sclerosis (ALS), Frontotemporal Dementia (FTD), and Huntington’s Disease (HD) [35–48]}. In this review we detail and discuss alterations to the NPC, NE, and NTR machinery and highlight emerging mechanisms by which these disruptions occur and impact pathophysiological events in neurodegenerative disease.

Nuclear pore complex assembly and maintenance

There are two modes of NPC assembly: interphase assembly and postmitotic assembly. Interphase NPC assembly takes place within the continuous double-lipid nuclear envelope bilayer and involves the asymmetric insertion of NPCs into the nuclear membrane. This process begins with the invagination of the inner nuclear membrane towards the outer nuclear membrane and thus it occurs through an “inside out” mechanism. The nuclear ring then assembles within this newly created “dome” within the nuclear membrane ultimately resulting in a fusion of the outer and inner nuclear membrane on either side of the NPC [4, 49–53]. This process is initiated during telophase following the completion of nuclear membrane sealing when the postmitotic nucleus expands its surface to facilitate nuclear import and chromosome decompaction. There are multiple proposed models by which interphase NPC assembly may occur: 1) The newly synthesized NPC components could be inserted into existing NPCs, forming enlarged mother NPCs that split into daughter NPCs, 2) the new NPC components could be newly incorporated into a region of the nuclear envelope devoid of preexisting NPCs, or 3) NPCs could be assimilated into vesicular intermediates that can later be fused to the nuclear envelope. Notably, interphase NPC assembly is an order of magnitude slower than postmitotic assembly, and it recruits components in a different order. For example, POM121 is recruited before the Nup107–160 complex in interphase assembly, whereas this order is reversed in postmitotic assembly. Additionally, interphase NPC assembly may require the function of the ESCRT-III/VPS4 complex to facilitate membrane sealing and maintain nuclear-cytoplasmic compartmentalization during the membrane fusion steps. [49, 50, 54–58]. Postmitotic NPC assembly occurs after cell division and is faster than interphase NPC assembly [49]. This process is initiated by the dephosphorylation of specific Nups such as Nup98 by CDK1 to facilitate NPC disassembly [59]. NPC assembly then starts with the binding of the nuclear ring component ELYS to chromatin in mid-anaphase, followed by the recruitment of the Nup107–160 complex and other transmembrane Nups, including NDC1 and POM121. Ultimately, this leads to the recruitment of other inner ring components, such as Nup205, Nup188, and Nup93, followed by the Nup62 complex, which forms the central channel component ultimately establishing the passive diffusion permeability barrier and active transport competent NPCs [49, 57, 58, 60].

In proliferating cells, Nups undergo active replacement through selective autophagy-lysosome pathways or the ubiquitin-proteasome pathway [61–64]. However, core scaffold components of the NPC have long lifespans in both dividing yeast cells and postmitotic cells [65–68]. The loss of mitotic renewal in terminally differentiated cells highlights the need for alternative turnover pathways for NPCs. Studies using C2C12 myoblast-derived mouse muscle cells shed light on NPC maintenance in non-dividing cells showing two distinct turnover mechanisms: myotubes exhibited a piecemeal replacement of individual Nups components at different rates, while quiescent cells replaced entire NPCs with newly assembled ones [67]. It is likely that when Nups are incorrectly assembled or malfunctioning, it can impair the structural and functional integrity of NPCs. Interestingly, the deterioration of NPCs with age and the loss of nuclear integrity have been implicated in the age-related decline of postmitotic cell function and the development of degenerative diseases [32, 65, 69, 70]. However, the mechanisms that underlie NPC maintenance in long-lived non-dividing human neurons remain unknown at this time.

Amyotrophic Lateral Sclerosis and Frontotemporal Dementia

Amyotrophic Lateral Sclerosis (ALS) is a progressive neurodegenerative disease culminating in the degeneration of upper and lower motor neurons and local interneurons as well as glial cell dysfunction within the motor circuitry [71, 72]. Disease causative mutations in over 20 genes have been linked to ALS. However, only about 10% of ALS cases are inherited (fALS) and the remaining 90% of cases are termed sporadic ALS (sALS) [73–75]. The most frequently observed mutation is a hexanucleotide repeat expansion (HRE) in the first intron of the C9orf72 gene accounting for about 40% of fALS and 8% of sALS [76, 77]. Interestingly the C9orf72 HRE is also the most common cause of familial forms of frontotemporal dementia (FTD), the second most common form of dementia characterized by progressive dysfunction and loss of neuronal populations within the frontal and temporal cortices. In fact, in some families, these mutations can cause ALS, FTD, or in a subset of patients both ALS and FTD [76–82]. Consistent with C9orf72 centric clinical genetics, up to 50% of ALS patients can display a mild cognitive impairment and about 15% are diagnosed with FTD [82, 83].

Pathologically, 97% of ALS cases, about 50% of FTD, and up to 50% of Alzheimer’s Disease (AD) cases display nuclear clearing and subsequent cytoplasmic mislocalization and/or aggregation of the normally nuclear RNA binding protein TAR DNA Binding Protein (TDP-43) in end-stage autopsy tissue [79, 84–88]. Thus, despite the distinct clinical manifestations and differences in CNS regions implicated in disease, ALS and FTD are considered to be on a spectrum of neurodegenerative disease due to shared genetic and pathological underpinnings. Thus, these commonalities suggest that these age-related neurodegenerative disease may also share pathophysiological underpinnings. Below, we will discuss recent documentations of NPC and NCT alterations in ALS and FTD pathophysiology.

Nuclear pore complexes, Nuclear Transport Receptors, and Nucleocytoplasmic Transport in ALS and FTD

In 2015, a landmark study was the first to detail disruptions in nucleocytoplasmic transport in C9orf72 ALS. Using induced pluripotent stem cell (iPSC) derived neurons (iPSNs), fly models, and postmortem human tissues, the authors demonstrated that expression of the C9orf72 HRE disrupts the localization of Ran GTPase and functionally compromise active nuclear import [46]. Subsequent work has demonstrated that dipeptide repeat proteins (DPRs) non-canonically translated from repeat RNA species may also impact NCT [89–91] perhaps through direct interactions with NTRs that impede cargo loading/unloading at the NPC [44]. However, we note that the impact of DPRs on NCT remains controversial. For example, one report suggested that artificial PolyPR peptides can localize to the central channel of the NPC in Xenopus oocytes and impede nuclear import and export capacity in U2OS cells [90]. Notably, the localization of PolyPR peptides to the NPC central channel or NPC itself has not been observed in authentic human disease tissue [76, 77, 92–95]. Moreover, a recent report has suggested that PolyGR and PolyPR DPRs do not interfere with NCT in HeLa and SH-SY5Y cells as well as iPSNs [96]. Thus, these studies suggest that the experimental results may be significantly impacted by the model system and specific DPR constructs or peptides utilized. Further, a recent study has suggested that DPR-induced phenotypes can be altered by commonly utilized fluorescent tags [97]. Fewer studies have centered around the contribution of repeat RNA itself to disrupted NPCs and NCT. However, two reports clearly demonstrate that pathologic G₄C₂ repeat RNA compromises Nup expression and localization within the NPC [37] and sequesters RanGAP1 to impede NCT [46]. Importantly, RanGap1 aggregation and reduction of Nups from the NPC were also observed in postmortem C9orf72 patient CNS tissues [37, 46] suggesting that they occur in human disease.

A number of studies have now documented the mislocalization and/or nuclear and cytoplasmic aggregation of a number of elements of the nuclear transport machinery including Nup62, RanGap1, and additional Nups and NTRs in familial and sporadic forms of ALS [35, 42, 46–48, 91, 98–101]. Interestingly, the cytoplasmic localization/mislocalization of Nups and NTRs has been linked to RNA binding protein aggregation [42, 48, 102]. Specifically, the RNA binding proteins TDP-43 and Fused in Sarcoma (FUS), which have been genetically linked to a small number of ALS cases [103, 104], form characteristic insoluble aggregates in disease at least in part due to aberrant phase transitions [84, 105]. It is hypothesized that these RNA binding protein aggregates sequester Nups and NTRs which may contribute to neuronal demise. Nup62 and TDP-43 inclusions are frequently found in ALS/FTD, suggesting that cytoplasmic Nup62 may contribute to TDP-43 pathology [42]. Further, Nup62 and FUS directly interact and Nup62 impacts FUS droplet formation [48, 106]. Moreover, nuclear import receptors have been shown to reverse aberrant phase separation and aggregation of TDP-43, FUS, and DPRs [102, 107–110] suggesting not only pathological but potentially therapeutic implications for NTR – RNA binding protein interactions in neurodegeneration.

Although cytoplasmic aggregates have been shown to interfere with the NCT of protein and RNA [111], it is unknown whether specific Nup and NTR mislocalization and/or aggregation observed in ALS is sufficient to directly impact NCT and impact overall NPC composition. For example, it remains unknown what proportion of Nup or NTR molecules must be sequestered away from their normal function to directly impact NCT. It also remains unclear as to how many independent Nups and NTRs need to be pathologically impacted to impede NCT and whether others may be able to functionally compensate at least until a threshold of sequestration is reached. In addition, cytoplasmic aggregation may reflect a pathological sequestration of cytoplasmic pools of Nups that are either newly synthesized or have normal cytoplasmic functions and thus may not directly affect the overall composition of the NPC itself.

In contrast, recent work has used super resolution imaging strategies to examine the expression and localization of 23 individual Nup proteins in iPSNs and postmortem human CNS tissues. These comprehensive studies demonstrated a reduction of roughly eight specific Nups (POM121, GP210, NDC1, Nup133, Nup107, Nup50, TPR, Nup153 (sALS), and Nup98 (C9orf72 ALS/FTD) from the nucleus, presumably NPCs, in both sALS and C9orf72 ALS/FTD [36, 37]. A series of rapid protein degradation and overexpression studies concluded that NPC injury events were initiated by a reduction of the transmembrane Nup POM121. Moreover, the authors of these studies showed that the collective reduction of Nups from the NPC was sufficient to impact the subcellular distribution of Ran GTPase, functional active nuclear import, the localization and function of TDP-43, and overall neuronal survival [36, 37].

NPCs and their Nup constituents are extremely long-lived, especially in non-diving post-mitotic cells such as neurons [66–68]. However, it has been shown that individual Nup proteins within NPCs can exchange and turnover at varying rates in a piecemeal manner [67, 112, 113]. Interestingly, some Nups contain a NLS or NES that functions to facilitate their insertion into the NPC or nuclear envelope [114, 115]. Thus, it is possible that deficits in functional NCT compromise Nup exchange and turnover to contribute to both the reduction of specific Nups from the NPC and nuclear and cytoplasmic Nup accumulations observed in neurodegenerative disease (Table 1, Figure 1).

Table 1:

Summary of Nup, NTR, and nuclear envelope alterations observed in model systems of neurodegenerative disease.

Neurodegenerative Disease	NPC Subdomain, NTR, or NE Localization	Protein Affected	Alteration	Model System	Reference
Amyotrophic Lateral Sclerosis	Cytoplasmic Ring and Filaments	Nup88	Altered nuclear envelope distribution	SOD1 A4V ALS and sALS human spinal cord	Kinoshita et al., 2009
			Cytoplasmatic mislocalization and aggregation	Neuro-2A cells expressing TDP-43 CTF	Chou et al., 2018
		Nup214	Reduced protein expression	Drosophila brains expressing 30 GGGGCC repeats	Dubey et al., 2023
			Reduced expression and altered distribution at the nuclear envelope (mAb414)	Primary mouse motor neurons and human lymphoblast cells expressing PFN1 mutants	Giampetruzzi et al., 2019
			Cytoplasmatic mislocalization and aggregation	Neuro-2A cells expressing TDP-43 CTF and TDP-43 Q331K	Chou et al., 2018
		Nup358 (RanBP2)	Reduced expression and altered distribution at the nuclear envelope (mAb414)	Primary mouse motor neurons and human lymphoblast cells expressing PFN1 mutants	Giampetruzzi et al., 2019
			Cytoplasmatic mislocalization and aggregation	Neuro-2A cells expressing TDP-43 CTF and TDP-43 Q331K	Chou et al., 2018
	Outer Ring	Nup107	Reduced nuclear expresion	C9orf72 iPSNs	Coyne et al., 2020
			Reduced nuclear expression	C9orf72 human motor cortex and thoracic spinal cord	Coyne et al., 2020
			Perinuclear aggregation	C9orf72 human motor cortex	Zhang et al., 2015
			Perinuclear aggregation	Drosophila salivary gland expressing 58 GGGGCC repeats	Freibauman et al., 2015
			Mislocalization and aggregation	Neuro-2A cells expressing TDP-43 CTF and TDP-43 Q331K	Chou et al., 2018
		Nup133	Reduced nuclear expression	C9orf72 human induced pluripotent stem cell derived neurons	Coyne et al., 2020
			Reduced nuclear expression	C9orf72 human motor cortex and thoracic spinal cord	Coyne et al., 2020
			Reduced nuclear expression	sALS iPSNs	Coyne et al., 2021
		Nup160	Cytoplasmatic mislocalization and aggregation	Neuro-2A cells expressing TDP-43 CTF	Chou et al., 2018
	Inner Ring	Nup35	Cytoplasmatic mislocalization and aggregation	Neuro-2A cells expressing TDP-43 CTF	Chou et al., 2018
		Nup93	Reduced nuclear expression and increased cytoplasmic expression	Drosophila brains expressing 30 GGGGCC repeats	Dubey et al., 2023
			Cytoplasmatic mislocalization and aggregation	Neuro-2A cells expressing TDP-43 CTF, TDP-43 WT and TDP-43 Q331K	Chou et al., 2018
		Nup155	Cytoplasmatic mislocalization and aggregation	Neuro-2A cells expressing TDP-43 CTF	Chou et al., 2018
		Nup205	Altered nuclear localization	C9orf72 human motor cortex	Zhang et al., 2015
			Alterated nuclear distribution	C9orf72 and TDP-43 mutart fibroblasts	Chou et al., 2018
			Abnormal localization and perinuclear and cytoplasmic aggregation	C9orf72 human motor cortex	Chou et al., 2018
			Perinuclear aggregation	C9orf72 iPSNs	Zhang et al., 2015
			Decreased immunoreactivty and cytoplasmic aggregation	sALS and TDP mutant human motor cortex	Chou et al., 2018
			Cytoplasmatic mislocalization and aggregation	Neuro-2A cells expressing TDP-43 CTF	Chou et al., 2018
	Central Channel	Nup58	Cytoplasmatic mislocalization and aggregation	Neuro-2A cells expressing TDP-43 CTF	Chou et al., 2018
		Nup62	Reduced nuclear expression	iPSC neurons with lentivral-mediated GR50 expression	Gleixner et al., 2022
			Reduced expression at the nuclear envelope (mAb414)	Primary mouse motor neurons and human lymphoblast cells expressing PFN1 mutants	Giampetruzzi et al., 2019
			Decreased immunoreactivty and altered nuclear envelope distribution	sALS human spinal cord	Aisawa et al., 2019
			Decreased nuclear expression and cytoplasmic aggregation	sALS, C9orf72, and TDP-43 mutant human spinal cord	Khalil et al., 2022
			Altered nuclear envelope distribution	SOD1 A4V ALS and sALS human spinal cord	Kinoshita et al., 2009
			Altered nuclear envelope distribution	SOD1 G93A mouse spinal cord	Kinoshita et al., 2009
			Altered nuclear envelope distribution and cytoplasmic mislocalization	FUS mutant iPSNs	Lin et al., 2021
			Disrupted nuclear envelope distribution, mislocalization, and aggregation	C9orf72 and sALS human motor cortex	Gleixner et al., 2022
			Cytoplasmatic mislocalization and aggregation and reduced nuclear expression	HEK293 cells expressing GR50	Gleixner et al., 2022
			Cytoplasmatic mislocalization and aggregation	Neuro-2A cells expressing TDP-43 CTF, TDP-43 WT and TDP-43 Q331K	Chou et al., 2018
		Nup98	Reduced nuclear expression	C9orf72 iPSNs	Coyne et al., 2020
			Reduced nuclear expression	C9orf72 human motor cortex and thoracic spinal cord	Coyne et al., 2020
			Reduced expression and cytoplasmic mislocalization	Drosophila brains expressing 30 GGGGCC repeats	Dubey et al., 2023
			Mislocalization and aggregation	Neuro-2A cells expressing TDP-43 CTF, TDP-43 WT and TDP-43 Q331K	Chou et al., 2018
	Nuclear Basket	Nup50	Reduced nuclear expression	sALS iPSNs	Coyne et al., 2021
			Reduced nuclear expression	C9orf72 iPSNs	Coyne et al., 2020
			Reduced nuclear expression	C9orf72 human motor cortex and thoracic spinal cord	Coyne et al., 2020
			Reduced expression and cytoplasmic mislocalization	Drosophila brains expressing 30 GGGGCC repeats	Dubey et al., 2023
			Altered nuclear envelope distribution	Drosophila expressing poly-GA 64 or poly-GR 64	Solomon et al., 2018
		Nup153	Reduced nuclear expression	sALS iPSNs	Coyne et al., 2021
			Reduced expression and altered distribution at the nuclear envelope (mAb414)	PrimaiY mouse motor neurons and human lymphoblast cells expressing PFN1 mutants	Giampetruzzi et al., 2019
			Abnormal nuclear distribution	SOD1 A4V ALS and sALS human spinal cord	Kinoshita et al., 2009
			Mislocalization and aggregation	Neuro-2A cells expressing TDP-43 CTF and TDP-43 WT	Chou et al., 2018
		TPR	Reduced nuclear expression	C9orf72 iPSNs	Coyne et al., 2020
			Reduced nuclear expression	sALS iPSNs	Coyne et al., 2021
			Reduced total protein level and cytoplasmic mislocalization	Drosophila brains expressing 30 GGGGCC repeats	Dubey et al., 2023
	Transmembrane Ring	GP210	Reduced nuclear expression	C9orf72 iPSNs	Coyne et al., 2020
			Mislocalization	Neuro-2A cells expressing TDP-43 CTF and TDP-43 Q331K	Chou et al., 2018
		NDC1	Reduced nuclear expression	C9orf72 iPSNs	Coyne et al., 2020
		POM121	Reduced nuclear expression	sALS iPSNs	Coyne et al., 2021
			Reduced nuclear expression	C9orf72 iPSNs	Coyne et al., 2020
			Reduced nuclear expression	C9orf72 human motor cortex and thoracic spinal cord	Coyne et al., 2020
			Reduced expression at nuclear envelope	Primary mouse motor neurons and human lymphoblast cells expressing PFN1 mutants	Giampetruzzi et al., 2019
			Mislocalization and aggregation	Neuro-2A cells expressing TDP-43 CTF	Chou et al., 2018
	Nuclear Transport Receptors and Nucleocytoplasmic Transport Proteins	Aladin	Mislocalization	Neuro-2A cells expressing TDP-43 CTF	Chou et al., 2018
		Gle1	Cytoplasmatic mislocalization and aggregation	Neuro-2A cells expressing TDP-43 CTF and TDP-43 Q331K	Chou et al., 2018
		KPNA2	Reduced expression, increased cytoplasmic distribution	sALS human spinal cord	Nishimura et al. 2010
Nuclear depletion and cytoplasmic accumulation			Drosophila expressing poly-GA64/poly-GR64/ΔNLS-TDP-43, human TDP-43/TDP-43Q331K in salivary gland cells	Solomon et al., 2018
KPNA3		Nuclear depletion and cytoplasmic accumulation	Salivary gland cells of Drosophila expressing ΔNLS-TDP-43, human TDP-43 WT or TDP-43 Q331K	Solomon et al., 2018
KPNA6		Reduced expression	sALS human spinal cord	Nishimura et al. 2010
KPNB1		Reduced nuclear expression	sALS human spinal cord	Kinoshita et al., 2017
		Reduced nuclear and nuclear envelope expression and cytoplasmic aggregation	C9orf72 and TDP-43 mutart human spinal cord	Khalil et al., 2022
		Altered nuclear envelope distribution, increased cytoplasmic localization	sALS human spinal cord	Aizawa et al., 2019
RanGap1		Nuclear accumulation	C9orf72 iPSNs	Zhang et al., 2015
		Nuclear accumulation	Drosophila S2 cells expressing (G4C2)30	Zhang et al., 2015
		Abnormal nuclear localization	C9orf72 human motor cortex	Zhang et al., 2015
		Altered nuclear envelope distribution	Mice expressing AAV (G4C2)149	Chew et al., 2019
		Altered distribution	Mouse primal cortical neurons expressing TDP 43 CTF or TDP-43-mtNLS	Chou et al., 2018
		Increased nuclear localization	TDP-43 A315T mouse model	Gautam et al., 2019
Ran GTPase		Reduced nuclear/cytoplasmic ratio	C9orf72 iPSNs	Zhang et al., 2015
		Reduced nuclear/cytoplasmic ratio	C9orf72 iPSNs	Coyne et al., 2020
		Reduced nuclear/cytoplasmic ratio	Drosophila S2 cells expressing (G4C2)30	Zhang et al., 2015
		Reduced nuclear/cytoplasmic ratio	sALS iPSNs	Coyne et al., 2021
		Nuclear reduction	C9orf72 iPSNs	Jovičić et al., 2015
THOC2		Cytoplasmic mislocalization	HEK293T cells expressing TDP-43 fragments	Woerner et al., 2016
TPNO1		Cytoplasmic aggregation	TDP-43 A315T mouse model	Gautam et al., 2019
XPO5		Mislocalization and aggregation	Neuro-2A cells expressing TDP-43 CTF	Chou et al., 2018
Nuclear Envelope	Lamin B1	Abnormal nuclear envelope distribution	C9orf72 and TDP-43 mutant patient fibroblasts	Chou et al., 2018
		Abnormal nuclear envelope distribution	TDP-43 mutant iPSNs	Chou et al., 2018
		Abnormal nuclear envelope distribution	Mouse primary cortical neurons expressing TDP-43 CTF, TDP-43 Q331K, or TDP-43 M337V	Chou et al., 2018
	Lamin C	Abnormal nuclear envelope distribution	Drosophila salivary gland expressing 58 GGGGCC repeats	Freibaum et al., 2015
Frontotemporal Dementi a/Alzheimer’s Disease	Central Channel	Nup62	Perinuclear co-aggregation with phospho-Tau	FTD/AD human hippocampus	Eftekharzadeh et al., 2018
		Nup98	Perinuclear co-aggregation with phospho-Tau	Mice expressing Tau P301L	Eftekharzadeh et al., 2018
			Perinuclear co-aggregation with phospho-Tau	FTD/AD human hippocampus	Eftekharzadeh et al., 2018
			Perinuclear aggregation	Tau mutant iPSNs	Paonessa et al., 2019
	Nuclear Transport Receptors and Nucleocytoplasmic Transport Proteins	KPNA2	Reduced expression	FTD-TDP human frontal cortex	Nishimura et al. 2010
		KPNA3	Reduced nuclear expression and cytoplasmic accumulation	C9orf72 and sFTD	Solomon et al., 2018
		KPNA4	Reduced nuclear expression and cytoplasmic accumulation, reduced solubility	C9orf72 and sFTD human frontal cortex	Solomon et al., 2018
Huntington’s Disease	Cytoplasmatic Ring and Filaments	Nup88	Abnormal nuclear distribution and co-aggregation with mHTT	mHTT mouse model	Grima et al., 2017
			Co-aggregation with mHTT	HEK293 cells expressing PolyQ	Suhr et al., 2009
		Nup214	Co-aggregation with mHTT	HEK293 cells expressing PolyQ	Suhr et al., 2009
		Nup358	Co-aggregation with mHTT	HEK293 cells expressing PolyQ	Suhr et al., 2009
	Central Channel	Nup62	Abnormal nuclear distribution and co-aggregation with mHTT	mHTT mouse model	Grima et al., 2017
			Abnormal local izaiton	human striatum and cortex	Grima et al., 2017
			Cytoplasmic mislocalization	HD iPSNs	Grima et al., 2017
			Co-aggregation with mHTT	HEK293 cells expressing PolyQ	Suhr et al., 2009
	Nuclear Basket	Nup153	Co-aggregation with mHTT	HEK293 cells expressing PolyQ	Suhr et al., 2009
	Transmembrane Ring	POM121	Co-aggregation with mHTT	HEK293 cells expressing PolyQ	Suhr et al., 2009
	Nuclear Transport Receptors and Nucleocytoplasmic Transport Proteins	Gle1	Abnormal nuclear distribution and co-aggregation with mHTT	mHTT mouse model	Gasset-Rosa et al., 2017
		RanGAP1	Co-aggregation with mHTT	mHTT mouse model	Grima et al., 2017
			Intranuclear accumulation	human striatum and cortex	Grima et al., 2017
			Cytoplasmic mislocalization	HD iPSNs	Grima et al., 2017
			Abnormal nuclear distribution and co-aggregation with mHTT	mHTT mouse model	Gasset-Rosa et al., 2017
			Abnormal nuclear distribution	human cortex	Gasset-Rosa et al., 2017
		Ran GTPase	Reduced nuclear/cytoplasmic ratio	HD iPSNs	Grima et al., 2017
			Reduced nuclear/cytoplasmic ratio	primary mouse cortical neurons expressing mHTT	Grima et al., 2017
		KPNA2	Co-aggregation with mHTT	primary mouse neurons expressing mHTT	Woerner et al., 2016
		KPNA4	Co-aggregation with mHTT	primary mouse neurons expressing mHTT	Woerner et al., 2016
	Nuclear Envelope	Lamin B1	Abnormal nuclear distribution and morphology	mHTT mouse model	Gasset-Rosa et al., 2017

Figure 1: Nuclear pore complex, nuclear envelope, and nuclear transport receptor pathology in neurodegeneration.

Graphical depiction of Nup, NTR, and nuclear envelope pathologies documented in neurodegenerative disease. Normally, intact and functional NPCs maintain proper nucleocytoplasmic compartmentalization. NTRs (red circles) facilitate the active nucleocytoplasmic trafficking of macromolecules between the nucleus and cytoplasm. The ESCRT-III protein CHMP7 (purple circles) can passively diffuse into the nucleus to maintain NPC and nuclear envelope homeostasis (in cooperation with other ESCRT-III proteins). However, nuclear levels remain low due to active nuclear export. TDP-43 (green circles) is predominantly localized to the nucleus where it functions to regulate transcription and splicing of its mRNA targets. TDP-43 is also involved in mRNA transport and translation, though its cytoplasmic localization is typically minimal. In ALS the abnormal nuclear accumulation of CHMP7 is sufficient to initiate a reduction of Nups from the NPC leading to their subsequent degradation via the proteasome and lysosome. This collective reduction of multiple Nups from the neuronal NPC impacts functional nucleocytoplasmic transport (dashed arrows) which in turn, at least in part, contributes to nuclear loss of TDP-43 function and localization. Loss of nuclear TDP-43 localization and function impacts expression and splicing of target RNAs potentially contributing to loss of function for their associated proteins. We note that the majority of studies to date have been largely descriptive and thus, the molecular mechanisms that give rise to Nup, NPC, NTR, and nuclear envelope alterations in disease remain largely unknown. Nonetheless, defects in nucleocytoplasmic transport (dashed arrows), nuclear mRNA accumulation (presumably through impaired RNA export), and cytoplasmic Nup, NTR, and protein aggregation have been documented in models of ALS, FTD, and HD. Additionally, in Tau-based FTD models, nuclear envelope invaginations and perinuclear Nup – Tau co-aggregations have been observed. In HD, intranuclear accumulations comprised of mutant Htt protein and Nups or NTRs have been reported. Figure created with BioRender.

It is reasonable to hypothesize that this pathologic event compromises multiple functions of the NPC, given that these Nups span multiple subcomplexes of the NPC. However, this phenomenon has yet to be studied in detail and it is unclear specifically how reduction of each of these individual Nups impacts neuronal function and disease pathophysiology. Intriguingly, genetic variants in Nup50 have recently been identified as a risk factor for sALS [116]. While some of these variants result in a truncated protein and elicit a reduction in Nup50 protein levels, others are postulated to impact binding to Importin α or Nup153 based on their location within the Nup50 protein sequence. Thus, it is hypothesized that loss of Nup50 function is a significant contributor to neuronal death in ALS. Indeed, Megat and colleagues demonstrated that knockdown of Nup50 is sufficient to elicit locomotor dysfunction, neurodegeneration, and reduced cellular survival in multiple cellular and in vivo model systems [116]. Collectively, these studies clearly demonstrate a significant contribution of NPC injury to disease pathophysiology and importantly restoring POM121 expression was shown to be neuroprotective in human neurons [37] highlighting the potential therapeutic benefit of targeting NPC injury events in neurodegeneration.

The cytoskeleton and nuclear pore complexes and nucleocytoplasmic transport in ALS and FTD

Mutations in Profilin 1 (PFN1) and the microtubule associated protein Tau (MAPT, Tau), proteins associated with the actin and microtubule cytoskeleton respectively, have been identified in a subset of ALS and FTD cases [117, 118]. Moreover, about 45% of FTD cases are characterized by aggregation of wildtype Tau [84, 119]. Cytoskeletal integrity plays a pivotal role in maintaining positioning, stability, and integrity of the nucleus via connections with the LINC complex [120–125]. Thus, pathologic alterations to the cytoskeleton are likely to detrimentally impact nuclear integrity and stability. Indeed, ALS associated mutations in PFN1 can impact actin assembly dynamics [126, 127]. In addition, expression of PFN1 mutants in primary rodent neurons resulted in the mislocalization of Nups and NTRs as well as impaired functional nuclear import and abnormal nuclear envelope structure. Importantly, pharmacological modulation of actin polymerization was sufficient to restore nuclear import in both PFN1 and C9orf72 ALS/FTD models of disease [41] supporting a role for actin dynamics in the maintenance of NPC function and integrity. Furthermore, Tau deposition results in abnormal microtubule dynamics that deform the nuclear membrane causing nuclear envelope invaginations [128]. Interestingly, these nuclear lamina invaginations are highly reminiscent of those observed in laminopathies, a group of diseases caused by mutations in genes related to the nuclear lamina such as LMNA, LMNB1, and LMNB2 or nuclear envelope such as EMD, LAP2, LBR, and SYNE-1 [129–135], suggesting that nuclear maintenance of nuclear membrane integrity is critical for cellular function and survival. Morphological changes in the nuclear envelope and alterations to the NPC and NCT have also been reported during aging [65, 136–140]. As such, it is hypothesized that age-related neurodegenerative disease pathologies are the result of “accelerated” aging phenomenon. However, this has not yet been tested directly. Nonetheless, nuclear envelope invaginations impact nuclear lamina structure and impact gene expression and mRNA export in Drosophila models of tauopathy [141–143]. Moreover, pathologic phosphorylated Tau co-aggregates with Nup98 in the perinuclear space and FTD-associated mutant Tau impairs nuclear import in mice and human neurons [38, 128]. Interestingly, reduction of Nup98 restores the subcellular distribution of Ran GTPase [38] suggesting that Tau-mediated Nup98 aggregation may directly impact NCT capacity in FTD. Collectively, these studies demonstrate a central role of cytoskeletal disruptions as a contributor to impaired NCT and Nup and NTR mislocalization in ALS and FTD.

Huntington’s Disease

Trinucleotide CAG repeat expansions of greater than 40 repeats in the first exon of the huntingtin (HTT) gene are causative of the autosomal dominant neurodegenerative disorder Huntington’s Disease (HD) that results in degeneration of striatal medium spiny neurons to give rise to choreiform movement and dementia. Interestingly, longer expansion lengths have been associated with increased disease severity and earlier onset [144, 145]. While much of the research efforts detailed below have focused on pathologic PolyQ proteins, it is noted that CAG repeats can undergo non-canonical translation to produce a multitude of pathologic protein species including PolyA, PolyS, PolyL, and PolyC which have also been demonstrated to accumulate in human brain [146]. Highly aggregation-prone PolyQ proteins form inclusions in the striatum and cortex and can sequester a number of proteins including Nups and NTRs [43, 146–148]. In mice, iPSNs, and postmortem human brain tissue, mutant huntingtin (mHTT) co-aggregates with Nup62, Nup88, Gle1, and RanGAP1 and causes nuclear envelope abnormalities, altered subcellular distribution of Ran GTPase, nuclear mRNA accumulation and deficiencies in active nucleocytoplasmic transport [43, 148–150]. Notably, overexpression of Ran GTPase prevented HTT-induced neurodegeneration in Drosophila [150] demonstrating a role for NCT capacity in mediating disease pathophysiology in HD.

The ESCRT pathway and its relationship to neurodegenerative disease

The Endosomal Sorting Complexes Required for Transport (ESCRT) pathway is involved in the sorting, trafficking, and degradation of various proteins, lipids, and membrane-bound vesicles. The ESCRT pathway is conserved among eukaryotes, from yeast to humans and broadly functions in membrane remodeling events and is crucial for various cellular processes such as cytokinesis, virus budding, multivesicular body (MVB) biogenesis, autophagy and endo-lysosomal degradation, and more recently, nuclear pore assembly surveillance [151–153]. The ESCRT pathway is composed of five distinct protein complexes: ESCRT-0, ESCRT-I, ESCRT-II, ESCRT-III, and the Vps4 complex, which function together in a stepwise manner. Briefly as it relates to protein degradation, ESCRT-0 helps sort ubiquitinated cargo and initiates MVB formation. ESCRT-0 recruits ESCRT-I, which in turn recruits ESCRT-II to form a larger complex on the endosomal membrane. This complex recruits ESCRT-III in order to facilitate the recruitment of the AAA ATPase VPS4 and the final membrane scission event. VPS4 then catalyzes the disassembly of the ESCRT-III complex, allowing the ESCRT-III proteins to be recycled and proteins contained within MVBs and endosomes to be degraded via the lysosome [151–154].

Defects in the endo-lysosomal protein degradation pathway are associated with the pathogenesis of many neurodegenerative diseases [155–157]. Given the involvement of ESCRT proteins in this pathway, loss of function of ESCRTs can lead to toxic accumulation of cargo in cells, causing stress and ultimately neuronal death. For example, forebrain-specific ablation of ESCRT-0 induces hippocampal neuronal cell loss accompanied by the accumulation of ubiquitinated proteins, including α-synuclein and huntingtin [158]. Moreover, mutations in the ESCRT-III protein CHMP2B have been identified in FTD and a lower motor neuron predominant form of ALS [159–164]. CHMP2B is an integral component of the ESCRT machinery, enabling scission in MVB formation and sorting ubiquitinated cargo for endosomal-lysosomal trafficking. Studies have shown that expression of mutant CHMP2B leads to cell death progressive neurodegeneration in mice [165]. Although early overexpression-based studies suggest that disease-associated CHMP2B mutations disrupt endo-lysosomal trafficking and autophagic degradation [159, 161, 164, 166], the pathophysiological cellular disruptions resulting from endogenous levels of mutant CHMP2B expression remain largely unknown. Interestingly, CHMP2B has recently been demonstrated to regulate the phosphorylation of TDP-43 via CK1. Moreover, knockdown of CHMP2B prevented toxicity associated with TDP-43 overexpression in both Drosophila and mammalian cell models [167]. This observation may challenge initial documentation that TDP-43 pathology is not a defining feature of CHMP2B FTD [159, 160, 163, 168, 169]. Thus, exploring the relationship between CHMP2B, TDP-43, and the ESCRT machinery may provide valuable insights into the underlying mechanisms of FTD pathogenesis.

The ESCRT-III pathway and NPC and nuclear envelope surveillance

Recently, a role for the ESCRT pathway in nuclear envelope and NPC surveillance has been identified during cell division and NPC assembly and insertion into the nuclear membrane [170–177]. As it relates to NPC surveillance, ESCRT-III and VPS4 function to prevent aberrant NPC assembly, which can lead to the formation of a previously undiscovered quality control compartment termed Storage of Improperly assembled Nuclear pore Complexes (SINC) within the nuclear envelope in yeast. In the SINC, defective NPC assembly intermediates are recognized and cleared through a surveillance pathway that involves the ESCRT-III protein Chm7, inner nuclear membrane LEM family proteins Heh2/Heh1 and potentially upstream factors in the ubiquitin-proteasome system in yeast. Chm7 and Vps4 are recruited to the NPC assembly site through direct interactions with Heh2/Heh1 and there are multiple potential models for how they act to remove aberrant NPC assembly intermediates. In one model, Vps4 plays a role in stripping ESCRT-III from the membrane in a complex with Nups. In the second model, NPC assembly intermediates might be cleared through a vesicular intermediate in the nuclear envelope lumen, similar to the mechanism employed by ESCRT-III in multivesicular body biogenesis [170–172]. Moreover, upon nuclear envelope breakdown or rupture, the mammalian homolog of Chm7, CHMP7, binds to the inner nuclear membrane protein to LEMD2. This event facilitates CHMP7 polymerization which ultimately initiates a cascade of molecular events culminating in VPS4-mediated resealing of the nuclear envelope [173, 174, 178]. Collectively, these findings suggest that the ESCRT machinery plays a critical role in removing aberrant NPC biogenesis intermediates during NPC assembly and insertion and resealing of nuclear envelopes during cell division. However, whether these mechanisms relate to postmitotic NPC and nuclear envelope maintenance remains understudied.

Recent work in sALS and C9orf72 ALS/FTD iPSNs and postmortem human CNS tissues has identified a critical and pivotal role for aberrant nuclear accumulation of CHMP7 in initiating and facilitating NPC injury events including the reduction of specific Nups from neuronal nuclei and NPCs [36]. Under basal cellular conditions, CHMP7 passively diffuses in the nucleus and is actively exported by the NTR CRM1/XPO1 [170, 179] thereby maintaining low nuclear levels of CHMP7. In sALS and C9orf72 ALS/FTD human neurons, CHMP7 is aberrantly accumulated or retained, a pathological phenomenon that is sufficient to initiate a reduction in specific Nups from the human neuronal nucleus and NPC which in turn impacts functional NCT and the localization and function of TDP-43 [36]. In support of these studies, aberrant degradation of Nup214, Nup98, and Nup50 were prevented by the knockdown of Vps4 and additional ESCRT-III pathway proteins including CHMP2B and CHMP4B in a Drosophila model of C9orf72 ALS/FTD [180]. Additionally, knockdown of ESCRT proteins prevented amyloid precursor protein (APP)-mediated neurodegeneration in a Drosophila model of AD [181], and overexpression of ECSRT proteins protected cellular models from the detrimental impacts of Tau aggregation [182].

Interestingly the Nups targeted in Drosophila and human models of C9orf72 ALS/FTD are distinct [37, 180] and importantly, flies do not have gene encoding POM121 which has previously been demonstrated to be an essential component of human NPC injury events [37]. Moreover, pathological alterations in CHMP2B and CHMP4B have not been observed in human C9orf72 ALS/FTD and sALS neuronal nuclei [183] underscoring critical differences between Drosophila models and authentic human disease. Nonetheless, these studies collectively highlight the ESCRT-III mediated nuclear surveillance as an important contributor to disease pathogenesis. Intriguingly, antisense oligonucleotide (ASO)-mediated knockdown of CHMP7 following the emergence of NPC injury was sufficient to restore the nuclear and NPC localization and expression of Nups and prevent TDP-43 dysfunction and mislocalization in sALS and C9orf72 ALS/FTD iPSNs [36] highlighting the therapeutic targetability of ESCRT-III proteins for the alleviation of pathophysiologic cascades in ALS.

Conclusions and Perspectives

Nup, NTR, and nuclear envelope abnormalities (Table 1, Figure 1) are now widely regarded as a prominent pathologic hallmark of neurodegenerative diseases. However, although a critical and necessary first step, studies to date have remained largely descriptive. Thus, although likely to have wide ranging implications for neuronal function including impaired nucleocytoplasmic transport, RNA binding protein mislocalization and aggregation, and potentially more global dysregulation of gene, RNA, and protein expression and localization, the precise pathophysiological impacts of Nup and NTR alterations and their individual contributions to disease pathophysiology remain largely unknown. Additionally, the molecular mechanisms that lead to documented Nup, NTR, and nuclear envelope pathologies in neurodegenerative disease are unclear at this time.

Recent work has highlighted the ESCRT-III nuclear surveillance pathway as a significant contributor to NPC injury events in ALS. Importantly, therapeutic targeting of aberrant ECSRT-III nuclear surveillance events alleviates pathogenic cascades in ALS iPSNs. Thus, these studies highlight NPC injury events and the proteins that mediate these events as potential therapeutic targets for neurodegenerative disease. However, as discussed throughout this review, a number of unanswered questions remain. What is the precise functional contribution of each Nup and NTR abnormality to pathophysiologic events in disease and how are these similar and different in cell type specific neurodegenerative diseases? What is the mechanistic and/or pathologic link between nuclear Nup reduction and Nup and NTR aggregation or do these represent independent events that could provide multiple opportunities for therapeutic intervention? What are the molecular mechanisms that lead to Nup and NTR dysfunction in cell type specific neurodegenerative diseases and in particular sporadic forms of disease? What is the utility of different model systems (e.g. endogenous vs overexpression, neuronal cells vs non-neuronal cells, human vs mouse vs flies) for replicating authentic human disease pathologies? Future studies aimed at answering these and additional questions will be essential for understanding the widespread contribution of NPC and NE injury to neurodegeneration and the identification of novel therapeutic targets for mitigating these events and their downstream consequences.

Abbreviations

NPC

nuclear pore complex

Nups

nucleoporins

FG

phenylalinine-glycine amino acids

NCT

nucleocytoplasmic transport

NTR

nuclear transport receptor

NLS

nuclear localization sequence

NES

nuclear export sequence

ALS

Amyotrophic Lateral Sclerosis

fALS

familial Amyotrophic Lateral Sclerosis

sALS

sporadic Amyotrophic Lateral Sclerosis

FTD

Frontotemporal Dementia

HD

Huntington’s Disease

AD

Alzheimer’s Disease

ESCRT

Endosomal Sorting Complex Required for Transport

HRE

hexanucleotide repeat expansion

iPSC

induced pluripotent stem cell

iPSN

induced pluripotent stem cell derived neuron

DPR

dipeptide repeat protein

MVB

multivesicular body