Ubiquitin System Mutations in Neurological Diseases

Abstract

Neuronal ubiquitin balance impacts the fate of countless cellular proteins, and its disruption is associated with various neurological disorders. The ubiquitin system is critical for proper neuronal cell state transitions and the clearance of misfolded or aggregated proteins that threaten cellular integrity. This article reviews the state and recent advancements in our understanding of the disruptions to components of the ubiquitin system, in particular E3 ligases and deubiquitylases, in neurodevelopmental and neurodegenerative diseases. Specific focus is given to enzymes with recent progress in their characterization, including identifying enzyme-substrate pairs, the use of stem cell and animal models, and the development of therapeutics for ubiquitin-related diseases.

Ubiquitin signaling in the nervous system

The ubiquitin-conjugation machinery (see Glossary) is a critical regulator in the nervous system, influencing a wide array of neuropathologies, including neurodevelopmental and neurodegenerative diseases [1–4]. This intricate system functions by tagging proteins with ubiquitin (Ub), a small 76 amino acid regulatory protein, marking them for proteasomal degradation, altering their function, or changing their localization within the cell. This process of ubiquitylation is pivotal for various cellular mechanisms, including cell cycle control, endocytic trafficking, DNA repair, and signal transduction, rendering it essential for neuronal operation and development.

Briefly, protein ubiquitylation is an ATP-driven process initiated by a ubiquitin-activating enzyme (E1) that primes ubiquitin and transfers it to the ubiquitin-conjugating enzyme (E2). Ubiquitin is then conjugated by its C-terminal glycine to the amino group of lysine residues through the action of ubiquitin ligases (E3s) [5]. Unconventional ubiquitin modifications not involving an isopeptide bond on proteins, lipids, and sugars are also starting to emerge [6]. With over 600 encoded in the human genome, the E3 ligases mediate substrate specificity for the Ub-conjugation machinery for various degradative and non-degradative signaling [7]. The reverse reaction is catalyzed by roughly 100 deubiquitylating enzymes (DUBs) in the human genome, which remove ubiquitin residues [8].

Unlike rapidly proliferating cells, neurons are post-mitotic and must regulate protein synthesis and degradation with a greater degree of care to prevent the accumulation of misfolded or aggregated proteins. This reliance on protein quality control pathways is underscored by the numerous neuropathologies linked to mutations in the ubiquitin-conjugation machinery and the ubiquitin-proteasome system (Figure 1).

Figure 1. Neurons rely on the ubiquitin system to maintain protein homeostasis.

To maintain proper neurodevelopment and homeostasis, neurons use the ubiquitin proteasome system to regulate protein abundance. The small protein ubiquitin is conjugated to a substrate through the action of E1, E2, and E3 enzymes to mark a substrate for proteasomal degradation or alter its localization or signaling. Ubiquitin is removed from substrates by DUBs. Mutations in E3s or Dubs alter substrate stability and are associated with developmental defects and neurodegenerative disease.

In the realm of neurodevelopmental disorders, disruptions in the ubiquitin-conjugation machinery can lead to significant abnormalities in synaptic development and neurogenesis, contributing to conditions like autism spectrum disorders (ASDs), intellectual disabilities (IDs), and microcephaly [1]. Indeed, specific E3 ubiquitin ligases and deubiquitylating enzymes have been identified as key players in these developmental processes, and mutations can result in either a loss or gain of function in ubiquitin-regulating enzymes, disrupting the protein turnover equilibrium and leading to atypical neurodevelopmental outcomes. Proteomic identification of E3 ligase-substrate pairs has begun to identify the mechanistic connections between ubiquitin system mutations and neurological diseases [9].

Conversely, in neurodegenerative diseases such as Alzheimer’s, Parkinson’s, and Huntington’s disease, the accumulation of ubiquitylated protein aggregates is a hallmark pathological feature. These conditions are marked by a gradual loss of neuronal structure and function, often coupled with the misfolding and aggregation of specific proteins like α-synuclein in Parkinson’s or SOD1 in ALS [10, 11]. Polyubiquitylation of protein aggregates is critical to recruit autophagy receptors for autophagosome formation for lysosomal degradation or for proteasomal targeting [12]. In neurodegenerative disorders, there is frequently an impairment in the proteasomal degradation pathway, resulting in the accumulation of ubiquitin-tagged proteins that aggregate and form inclusion bodies within neurons.

Integral to neuronal health and function, the ubiquitin-conjugation machinery is frequently dysregulated in a spectrum of neurological disorders. Mutations in key E3 ligases and deubiquitylating enzymes, as well as in E1s, E2s, and ubiquitin itself, result in developmental disorders and neurodegenerative diseases. In this review, we will delve into the current knowledge regarding mutations in the ubiquitin system and neuropathological conditions. This knowledge is crucial for the treatment of neuropathologies, but also for understanding the basic cellular mechanisms in neurons regulated by the ubiquitin system.

The ubiquitin system in neurodevelopment

Mammalian neurodevelopment is a complex process requiring intricate cellular coordination to form the central and peripheral nervous systems. As cells differentiate, they undergo vast proteome remodeling to facilitate cell state transitions [13]. Given their complex cellular morphology with long axons and branching dendrites, neurons require a high degree of functional compartmentalization. The development of axons, dendrites, and synapses is coordinated by local proteome turnover, which is regulated in large part by the ubiquitin-conjugation machinery [14]. With such a strong reliance on the UPS for regulating cell state transitions, loss of function mutations in ubiquitin-related genes have the potential for vast developmental outcomes. As such, there are numerous monogenic neurodevelopmental disorders where key cellular regulators of the ubiquitin system are disrupted, often leading to incomplete neurogenesis or disrupted synaptic plasticity (Supplemental Table 1).

As the ubiquitin system regulates nearly every aspect of cellular biology, a few components have been selected for a more in-depth review due to recent advancements towards understanding their role in neurophysiology and the mutations that lead to neurodevelopmental disease (Figure 2).

Figure 2. Components of the ubiquitin system mutated in neurodevelopmental disease.

(A) UBE3A is an E3 ligase whose substrates include synaptic regulators, enzymes in retinoic acid biosynthesis, and AMPAR regulators. (B) CUL3 is a Cullin scaffold protein and part of CRL3 complexes. Its substrates include regulators of cytoskeletal organization, histone methylation, and cap-dependent translation. (C) USP9X is a deubiquitylating enzyme that removes ubiquitin from substrates such as centrosome-associated proteins and cytoskeletal regulators. Abbreviations: XLID99, intellectual developmental disorder, X-linked 99; MRXS99F, female-restricted X-linked syndromic intellectual developmental disorder; NEDAUS, neurodevelopmental disorder with or without autism or seizures.

Although the greatest diversity within the ubiquitin system exists at the level of the E3 enzymes and DUBs, there are a handful of mutations identified among the E1 and E2 enzymes linked to neurological disease, as well as frameshift mutants in a ubiquitin gene (Box 1). These classes of genes are likely more infrequently associated with diseases because the number of E3 and DUB enzymes far outweighs the number of E1 and E2s.

Box 1: Ubiquitin, E1, and E2 in neuronal disease.

UBB

There are 4 sources of ubiquitin in the human genome. The UBB and UBC genes encode polyubiquitin precursors, while the other two genes, RPS27A and UBA52, encode ribosomal proteins fused to a single copy of ubiquitin. Transcriptional variation in the UBB gene can result in a frameshifted form of ubiquitin arising from molecular misreading [117]. The resultant UBB+1 protein has the canonical N-terminus, but frameshifting causes an atypical C-terminus that lacks the glycine residue for polyubiquitylation [118]. The presence of UBB+1 proteins has been associated with Alzheimer’s Disease (AD) in post-mortem tissue, and the use of 3D human neural cell culture models indicates that UBB+1 is sufficient to induce amyloid-Beta and tau aggregation [119]. Future work will need to establish the link between UBB+1 frameshifting, protein aggregation, and AD pathogenesis.

UBA1

The E1 activating enzymes are responsible for priming ubiquitin in an ATP-dependent reaction to form a thioester intermediate and transfer the ubiquitin to the catalytic cysteine of a cognate E2 enzyme. There are two ubiquitin E1 enzymes in the human genome, UBA1 and UBA6. Somatic mutations in UBA1 are associated with VEXAS (vacuoles, E1, X-linked, autoinflammation, somatic, MIM: 301054) syndrome. These UBA1 mutations are predominantly found in myeloid lineages at Met41, leading to the formation of a truncated variant of UBA1 that is catalytically deficient [120].

Rare germline UBA1 mutations have also been identified in X-linked infantile spinal muscular atrophy (XL-SMA, MIM: 301830) [121]. The SMA mutations are clustered in the active adenylation domain (AAD) of UBA1, suggestive of a dissimilar mechanism of pathogenesis compared to VEXAS syndrome [122]. The exact molecular mechanisms linking UBA1 mutations to VEXAS syndrome and SMA remain unclear.

UBE2A

There are roughly 40 E2 ubiquitin conjugation enzymes in humans that accept ubiquitin from the activating E1 enzymes and engage with E3 ligases to facilitate the transfer of ubiquitin to substrates [123]. A nonsense mutation in UBE2A leads to premature termination (Q128X) and is associated with Nascimento type of X-linked syndromic intellectual developmental disorder (MIM: 300860) [124]. Using a Drosophila model of Ube2A mutation and in vitro ubiquitylation, one study linked the E2 enzyme to Parkin-mediated clearance of dysfunctional mitochondria [125]. The existence of E2 mutations in neurological disease highlights the importance of priming E3 ligases for proper ubiquitylation of substrates in neuronal integrity.

UBE3A

The HECT-type E3 ligase UBE3A/E6AP is one of the most characterized components of the ubiquitin system implicated in human neurodevelopmental disorders (Figure 2A) [15, 16]. The gene is imprinted in neurons, where the paternally inherited copy is silenced by an antisense RNA transcript (UBE3A-ATS), leaving only the maternal copy to be expressed [17]. The importance of this E3 ligase during neural development is underscored by two genetic disorders. Loss of function mutations in UBE3A, chromosomal deletions, or uniparental disomy leads to Angelman syndrome (MIM: 105830), a neurodevelopmental disorder characterized by frequent laughter, intellectual disability, motor delay, seizures, and speech impairment [18, 19]. Conversely, increased UBE3A activity through chromosomal duplications or hyperactivating mutations is associated with ASD [20].

As a monogenic neurodevelopmental disorder, Angelman syndrome has attracted attention for gene therapy. To restore UBE3A levels in neurons, many strategies have attempted to unsilence the paternal allele. Targeting the antisense RNA transcript UBE3A-ATS, topoisomerase inhibitors were first used to prevent long non-coding RNA transcription. Though the strategy restored the expression of UBE3A in mouse models, the off-target effects of topoisomerase inhibitors such as topetecan encouraged the development of more targeted therapies [21]. To directly degrade UBE3A-ATS, antisense oligonucleotide (ASO) approaches were devised to form RNA/DNA hybrid duplexes that would be degraded by RNase H1 [22]. In recent years, reactivation of the paternal allele of UBE3A with ASOs has shown efficacy in increasing UBE3A activity in animal models [23–25]. Other approaches using CRISPR-Cas9 editing to reactivate the paternal allele have also shown potential for ameliorating AS-associated phenotypes [26]. Site-specific delivery and the resolution of AS symptoms will be the next challenges for genetic therapeutics.

As an E3 ligase, several substrates for UBE3A in neurons have been reported, many of which are important for synaptic transmission. Ephexin5 negatively regulates synapse development and is marked for degradation by UBE3A [27]. Deletion of Ephexin5 in a mouse model with a maternal deletion of UBE3A rescued hippocampal dysfunction phenotypes [28]. UBE3A is shown to promote presynapse elimination by suppressing BMP signaling, and Angelman syndrome-associate mutations disrupt this process. On the other hand, increasing the dosage of UBE3A leads to precocious synapse elimination and diminished synaptic transmission mirroring ASDs [29]. Increased UBE3A dosage was also found to disrupt cellular retinoic acid signaling through ALDH1A2, an enzyme in the retinoic acid synthesis pathway [30]. The Arc protein is a regulator of synaptic plasticity that controls the surface expression of AMPA receptors. Though it was originally identified as a direct substrate of UBE3A, the physical interaction and ubiquitylation of Arc has seen varied results [31–34]. The mRNA nuclear export protein SARNP is downregulated in UBE3A-overexpressing neurons, preventing the translation of AMPA receptors [35]. With such wide-reaching effects in the cell, future work will need to delineate the direct substrates of UBE3A from proteins that are secondarily altered in their function or expression.

CUL3

Cullin-3 (CUL3) is a Cullin-RING E3 ligase that requires binding to BTB domain-containing substrate receptors to target proteins for ubiquitylation (Figure 2B) [36]. CUL3 is a high-risk gene associated with neurodevelopmental disorder with or without autism or seizures (NEDAUS), with a number of de novo variants identified (MIM: 619239) [37]. Numerous animal and stem cell models have been developed in recent years to mimic the likely pathogenic variants in the gene to identify the underlying molecular mechanisms connecting CUL3 mutations to neurodevelopmental defects.

A variety of heterozygous Cul3 knockout mouse models have employed proteomic approaches to understand the substrates of Cul3 that underlie the associated neurodevelopmental defects. Cytoskeletal proteins Plastin 3 (Pls3), RhoA, Tropomyosin, and Transgelin-2 have all been identified as potential targets for Cul3-mediated ubiquitylation [38–40]. Consistently, previous reports have linked CUL3-mediated degradation of RhoA to actin cytoskeletal dynamics in HeLa cells and psychiatric disease [41, 42]. Additionally, the BTB domain-containing substrate receptor KCTD13 has been implicated in RhoA degradation, where deletion of KCTD13 in mice decreased synaptic transmission and increased RhoA protein levels [43]. In contrast, heterozygous CUL3 knockout in iPSC-derived cortical glutamatergic neurons did not observe differences in RhoA protein level by immunoblot compared to isogenic controls but instead saw alterations in FGF signaling [44]. Although many studies have linked CUL3 loss to cytoskeletal dysregulation, they differ in the mechanism and reported substrates of the E3 ligase.

Beyond the cytoskeleton, the histone methyltransferase Smyd3 was identified as a substrate of Cul3 in mice, where pharmacological Smyd3 inhibition reversed heterozygous Cul3 knockout-associated social behaviors and electrophysiological changes [45]. Another approach using a GFAP-Cre to target neural progenitor cells for heterozygous Cul3 knockout described high levels of eIF4G1 and linked cap-dependent translation as a potential pathophysiological mechanism [46]. As one of the 7 Cullin scaffolds found in human cells, CUL3 has many reported substrates reaching a variety of processes essential for cellular viability [47].

In addition to mammalian systems, recent advancements in Drosophila melanogaster have shown versatility for studying ubiquitin-mediated neurodevelopmental disorders with powerful genetic tools. RNAi-mediated Cul3 neuronal-specific knockdown in Drosophila revealed sleep disturbances and behavioral phenotypes paralleling known ASD symptoms and exhibited novel metabolic dysregulation [48]. Consistently, previous neuron-specific work on Cul3 in Drosophila has identified the BTB substrate receptor Insomniac, which is hypothesized to govern sleep and wakefulness [49, 50]. Fly models may prove to be valuable tools for studying the mechanistic connection between CUL3 and neurological disease.

Recent advancements in understanding the mechanism behind CUL3 mutations and neurodevelopmental defects have revealed a confusing number of potential substrates and molecular pathways in neurons. As a scaffold for many different Cullin-RING ligase complexes, CUL3 loss of function mutations have the potential for wide-reaching consequences in neurodevelopment and beyond. Another disease highlights the pleiotropic consequences of CUL3 mutations, pseudohypoaldosteronism type IIE (MIM: 614496), where mutations in CUL3 or the substrate receptor KLHL3 lead to dysregulation of blood pressure and electrolyte balance [51]. Future work to systematically identify the ubiquitylation targets of CUL3 will help narrow down the link between CUL3 mutations and neurodevelopmental phenotypes.

USP9X

USP9X is a member of the ubiquitin-specific protease (USP) family of DUBs that use a catalytic cysteine to remove ubiquitin from substrates (Figure 2C). Located on the X-chromosome, USP9X escapes X-inactivation, and mutations or truncations in the gene are associated with intellectual developmental disorder, X-linked 99 (XLID99, MIM: 300919), and female-restricted X-linked syndromic intellectual developmental disorder (MRXS99F, MIM: 300968) [52, 53]. A number of key substrates for USP9X have been identified, including apoptosis regulators MCL1 and XIAP, mTOR protein RAPTOR, β-catenin, and the Hippo pathway transcription factor YAP1 [54–58]. Despite the numerous developmental and signaling pathways regulated by USP9X, the mechanistic connections between USP9X mutations, substrate stability, and neurological disease remain elusive.

Major strides have been made in recent years to characterize USP9X-substrate interactions that relate to neurophysiology. As cells divide during mitosis, the centrosomes act as the microtubule organizing centers to separate chromosomes into daughter cells. As such, centrosome duplication is tightly controlled to occur exactly once per cell cycle for faithful cellular division [59]. USP9X has been identified as a critical factor regulating centrosome stability through its deubiquitylating activity on centrosomal proteins PCM1, CEP55, and CEP131 [60, 61]. Additionally, USP9X was found to be recruited to centrosomes through SFI1 to deubiquitylate and stabilize centrosomal duplication protein STIL [62]. Mutations in STIL have been associated with primary microcephaly (MIM: 612703), highlighting the importance of centrosome stability in neurophysiology [63]. Fibroblasts from MRXS99F patients with USP9X mutations were shown to have reduced levels of STIL, providing a possible explanation for the neurodevelopmental phenotype associated with USP9X mutations through centrosome biogenesis defects [62]. However, the effect of USP9X on centriole stability and duplication in primary neurons or neural progenitor cells remains to be shown.

Beyond centrosomes, USP9X has been implicated in stabilizing neuronal scaffolding proteins. Using a yeast 2-hybrid screen, USP9X was identified as an upstream regulator of the scaffold protein ankyrin-G. [64]. Variants in the ANK3 gene encoding Ankyrin-G are also linked with intellectual disability and neuropsychiatric disorders, pointing to the importance of regulating Ankyrin-G in synaptic development [65, 66]. A follow-up paper showed that TGF-β signaling stabilized the Usp9X-Ankyrin-G interaction to promote dendritic spine growth [67].

The ubiquitin system in neurodegeneration

Although neurodegenerative diseases represent a vast and inhomogeneous class of neuronal disorders, they frequently share the aggregation of misfolded proteins that result in declining cellular function and viability. These so-called proteinopathies are particularly detrimental in neurons since they are post-mitotic and cannot mitigate the cytotoxic effect of protein aggregates through cell division to dilute the pool of aggregates. Instead, neurons rely on protein degradation and quality control mechanisms to maintain proteostasis [4]. Most neurodegenerative diseases are idiopathic and increase in frequency with age. However, some causative mutations have been identified that link neurodegenerative diseases to mutations in the ubiquitin-proteasome system (Supplemental Table 2).

A selection of genes in the ubiquitin-proteasome system with mutations linked to neurodegenerative disease and recent advancements in understanding the mechanistic connection between mutation and neurodegeneration have been chosen for further discussion (Figure 3).

Figure 3. Components of the ubiquitin system mutated in neurodegenerative disease.

(A) Normal Parkin ubiquitylates outer mitochondrial membrane proteins to facilitate mitochondrial turnover. Dysfunctional Parkin cannot ubiquitylate outer mitochondrial membrane proteins leading to the accumulation of damaged mitochondrial. (B) Mutations in gigaxonin prevent substrate recognition, leading to the accumulation of substrates such as intermediate filaments which cause axonal distension. (C) Mechanism schematic is shown for CAG repeat expansion in ATXN3 leading to spinocerebellar ataxia 3.

Parkin

Parkin, an E3 ubiquitin ligase encoded by the PRKN gene, is fundamental to mitochondrial quality control and is implicated in Parkinson’s disease (PD, MIM: 600116). These genetic alterations lead to an inactive form of Parkin by disrupting its domain interfaces, underscoring the protein’s significance in cellular health [68]. In recent years, the scientific community has made substantial progress in elucidating Parkin’s role in PD, particularly regarding its activation mechanisms and pivotal function in maintaining mitochondrial integrity in neurons (Figure 3A). The activation of Parkin is a complex process requiring significant conformational changes [69–71], and its phosphorylation [72] is a critical step highlighted by the discovery of Parkin S65N mutations in PD patients [73]. These activation steps are crucial for its role in targeting various substrates [74, 75] and degrading dysfunctional mitochondria [76].

Research efforts spanning the last five years have taken various approaches to investigate Parkin, focusing on genetic mutations, cellular pathways, and potential therapeutic targets. These studies deepened our understanding of PD, highlighting Parkin’s critical functions in mitochondrial dynamics, autophagy, and neuron survival. Investigations have revealed how specific Parkin mutations affect its enzymatic activity and impact mitochondrial turnover, shedding light on the genetic underpinnings of PD’s progression [77]. Studies also demonstrate Parkin’s role in neuron activity, which is activated through CaMK2, recruited to synaptic vesicles, aiding in vesicle endocytosis [78]. Mutations in Parkin result in defective vesicle recycling and the accumulation of toxic oxidized dopamine, implicating it in dopaminergic neuron dysfunction.

Recent research has aimed to better understand the mechanism behind Parkin’s loss of function in dopaminergic (DA) neurons using induced pluripotent stem cell (iPSC) lines [79–82]. These studies have shown disruptions in the lysosome, mitochondria, mitochondria-lysosome contact sites, and oxidative stress pathways in DA neurons. Additionally, a study has found that Parkin mutation can lead to changes in astrocytes, suggesting a non-autonomous cell death pathway for dopaminergic neurons in patients with Parkin mutation [83].

Moreover, research into Parkin’s interaction with the ubiquitin-like modifier FAT10 has unveiled a novel regulatory layer, where FAT10ylation leads to Parkin degradation and impaired mitochondrial maintenance in SH-S5Y5 cells [84]. However, whether these mechanisms are relevant to PD remains to be determined. The advancements in our understanding of Parkin’s intricate role in PD highlight the protein’s complex functionality, particularly in mitochondrial quality control, autophagy, and neuron health. These findings contribute to a more nuanced understanding of PD’s etiology and underscore the disease’s multifactorial nature, involving a confluence of genetic mutations, mitochondrial dysfunction, and cellular process disruptions. As research continues to unravel the complexities of Parkin’s function and its implications for PD, it becomes increasingly clear that tackling this disease will require a comprehensive understanding of these interrelated cellular mechanisms. Parkin activation through phosphorylation by PINK1 provided an important example for how post-translational modification on the ubiquitin-like domain regulates enzyme function. How other modifications may alter Parkin’s activity in the cell and the role of phospho-ubiquitin for other E3 ligases remains to be shown.

Gigaxonin

Gigaxonin, a CUL3 adaptor protein also known as KLHL16, is crucial for targeting specific substrates for ubiquitylation and degradation in neurons (Figure 3B). Mutations in the KLHL16 gene cause Giant Axonal Neuropathy 1 (GAN) (MIM: 256850) [85], a pediatric neurodegenerative disease hallmarked by abnormal accumulation of intermediate filaments (IF) in axons, leading to axonal distension. This condition is part of a broader spectrum of neurodegenerative disorders, including ALS and Charcot-Marie-Tooth disease, characterized by disorganized neurofilament architecture. Gigaxonin is a BTB-Kelch protein that acts as a substrate receptor for the CUL3-RBX1-GAN E3 ligase complex and is best known for regulating intermediate filament stability [86]. Post-translational modification on gigaxonin regulates substrate stability, with evidence indicating that O-GlcNAcylation is required for intermediate filament recognition and turnover [87]. Beyond IF substrates, gigaxonin has been shown to regulate autophagy by controlling ATG16L1 turnover [88] and Sonic Hedgehog signaling through the Patched receptor [89].

Missense and homozygous truncating mutations are found throughout the gene in GAN patients that disrupt the normal function of gigaxonin in the CRL3 E3 ligase complex [90], leading to the disease’s hallmark features. Recent work has also contributed to the report of various novel pathologic variants [91–94]. Additionally, a recent study showed that the loss of GAN dramatically inhibited the transport of intermediate filaments along microtubules by the microtubule motor in a kinesin-1-specific manner [95].

To better study the function of gigaxonin, GAN patient-derived iPSCs were generated and differentiated into neural progenitor cells, astrocytes, and brain organoids [96]. These lines showed increased perinuclear vimentin expression and abnormal nuclear morphology, providing evidence that the intermediate filament vimentin is a likely substrate of GAN. Although recent work has showcased that our current understanding of gigaxonin is limited, further research is necessary to better understand its role and function in diseases. Additionally, methods to reduce the accumulation of intermediate filaments or other substrates of gigaxonin would prove useful for GAN treatment.

Ataxin-3

Ataxin-3 is a deubiquitylating enzyme mutated in the neurodegenerative disorder Spinocerebellar Ataxia 3 (SCA3, MIM: 109150). SCA3, also known as Machado-Joseph Disease (MJD), results from an expansion of CAG repeats within the ATXN3 gene, leading to an abnormal ataxin-3 protein with an extended polyglutamine (polyQ) tract (Figure 3C) [97–99]. The polyQ tract leads to aberrant translation products that are prone to aggregation and neurotoxicity. Such CAG repeats are common to neurodegenerative disorders including the HTT gene in Huntington’s disease and the androgen receptor gene AR in spinal-bulbar muscular atrophy [100]. Accumulation of intracellular ubiquitinated aggregates has been observed in SCA3 patients, and mouse models have shown the expression of mutant ataxin-3 leads to differential abundance of ubiquitylated species in an age and region-specific manner [101]. Understanding the mechanism behind CAG expansion and the pathophysiological consequence of polyQ ataxin-3 will be paramount for treating SCA3.

The function of ataxin-3 outside of an SCA3 context links it to endoplasmic reticulum-associated degradation (ERAD) and VCP response to misfolded proteins [102, 103]. The polyQ domain of ataxin-3 was found to be important for stabilizing beclin 1, and extended polyQ tracts inhibit the interaction and impair beclin 1-dependent autophagy [104]. VCP was later shown to have a stabilizing role in the ataxin-3/beclin-1 interaction and the PI3K complex critical for autophagosome formation [105]. Additionally, ataxin-3 has been reported as a transcriptional regulator through stabilizing histone deacetylase HDAC3 [106].

Wild-type ataxin-3 also plays a multifaceted role in genome integrity. In complex with p97, ataxin-3 regulates the abundance of the DNA repair E3 ligase RNF8 [107]. Additionally, it has been shown to stabilize double strand break repair protein MDC1 and checkpoint protein Chk1 [108, 109]. The polynucleotide kinase 3′-phosphatase (PNKP) critical for DNA end processing was found to be impaired by mutant ataxin-3, and complementation of PNKP rescued SCA3 phenotypes in a Drosophila model [110]. The importance of wild-type ataxin-3 in genome maintenance suggests another potential role for ataxin-3 in SCA3 pathophysiology beyond polyQ protein aggregation. Future work to uncouple the loss of DUB activity from the aggregation of polyQ protein would assist in the development of SCA3 therapeutics.

Preclinical studies on nucleic acid-based therapeutics have been conducted for the treatment of SCA3, often using viral vectors to deliver RNA-based mutant ATXN3 silencers [111, 112]. More recently, non-viral strategies to silence polyQ ATXN3 mRNA have been developed to prevent off-target mutagenesis and immunogenicity associated with viral vector delivery. Extracellular vesicles packaged with miRNA-based silencing sequences and neural targeting peptides were administered intranasally to an MJD mouse model, resulting in a decrease in mutant ATXN3 mRNA in the cerebellum [113]. To directly target CAG repeats, CRIPSR/Cas9 has shown promise in SCA3-iPSCs to repair the ATXN3 gene with a homologous recombination strategy. This approach not only silences the expression of the mutant ataxin-3 protein but also ameliorates several abnormal phenotypes associated with SCA3, such as oxidative stress and mitochondrial dysfunction [114]. These findings mark a significant step towards a potential cure, as well as highlighting the potential for SCA3-iPSCs for disease modeling.

Concluding Remarks

The ubiquitin system is critical for developing neurons through embryogenesis and maintaining proteostasis during adulthood. Neurons undergo vast proteome turnover during differentiation, moderated in part through ubiquitin-mediated proteasomal degradation. In terminally differentiated neurons, misfolded proteins and protein aggregates, hallmarks of neurodegenerative disease, are partially cleared out of the cell due to ubiquitin signaling. De novo mutations in E3s or DUBs lead to pleiotropic effects on neuronal physiology due to their vast array of potential substrates. Identifying ubiquitylated substrates for E3s and deubiquitylated substrates for DUBs is paramount for understanding their function in neuronal homeostasis and dysfunction (see Outstanding Questions).

Outstanding Questions Box.

• How can we systematically identify the substrates of ubiquitin ligases and deubiquitylating enzymes?

• Are there specific activation or deactivation signals that temporally restrict the function of E3s and DUBs during neurogenesis, and do mutations alter this regulation?

• How does the ubiquitin system regulate the clearance of aggregated proteins in neurodegenerative disease?

• How do the known somatic and germline mutations in UBA1 differ, and how do mutations in the E1 enzyme impact the transfer of ubiquitin and stability of the cellular proteome?

• How can the effect of de novo mutations in the ubiquitin system be measured on the ubiquitylome of neurons?

• How can stem cell-derived neurons better recapitulate the cellular environment for neurological disorders?

Numerous methods have been developed to profile ubiquitin enzyme specificities, including differential ubiquitylation proteomics, global protein stability profiling, and in vitro ubiquitylation/deubiquitylation assays, all of which come with their own sets of challenges and drawbacks [115, 116]. Redundancy and multiplicity within the ubiquitin system obfuscates the regulation of specific proteins. Given the human genome contains approximately ~600 E3 ligases, ~100 DUBs, and ~20,000 canonical proteins acting as potential substrates, a single E3/DUB likely has multiple substrates. E3s/DUBs may also exhibit “redundant” function, where the same substrate may be targeted by different E3 ligases or DUBs under different cellular conditions. Furthermore, the levels of ubiquitin enzymes’ expression can differ among various cell types and tissues, with certain enzymes being exclusively expressed in specific tissues. To comprehend this variation, it would be critical to integrate different stem cell and animal models.

Despite being a challenging task, understanding the relationship between E3/DUB substrates has led to significant clinical advancements in other fields. The discovery of the von Hippel-Lindau (VHL) E3 ligase and its substrate hypoxia-inducible factor (HIF) led to significant progress in the study of the hypoxic response in cells and the eventual development of HIF-2α inhibitors like Belzutifan for VHL disease-associated renal cell carcinoma.

Further efforts in systematically identifying E3 ligase and DUB substrates in specific cell types could substantially enhance our comprehension of protein homeostasis and its dysregulation in human diseases, including neurological disorders, and potentially offer fresh therapeutic insights.

Highlights.

• The ubiquitin system regulates protein homeostasis and is critical for neurophysiology.

• Protein-coding mutations in ubiquitin ligases and deubiquitylating enzymes underlie numerous neurological diseases.

• Substrate profiling for ubiquitin ligases and deubiquitylating enzymes in neurons has begun to help identify the mechanism behind some neurodevelopmental and neurodegenerative diseases.

• Targeting the ubiquitin system may offer therapeutic potential for neurological disease.

Glossary

Antisense oligonucleotide (ASO)

a short, single-stranded nucleic acid which binds a given RNA species to alter its function

Axonal distension

Swelling in the axon caused by trauma or the accumulation of aggregated proteins

Deubiquitylating enzyme (DUBs)

a family of enzymes known as proteases that cleave chemical bonds within or between ubiquitin molecules and other proteins

Ubiquitin ligase (E3s)

enzymes that catalyze the transfer of ubiquitin to substrates

Cullin-RING ligases (CRLs)

a superfamily of E3 ubiquitin ligases working as modular multi-subunit complexes

Mendelian Inheritance in Man (MIM)

a database of human genes and associated genetic disorders

Neurogenesis

the process of developing a neuron from a neural stem cell progenitor

Synaptic plasticity

activity-dependent modification of the strength or efficacy of synaptic transmission at preexisting synapses

Ubiquitin-conjugation machinery

encompasses the enzymes required for catalyzing the attachment of ubiquitin to substrates

Ubiquitin-proteasome system (UPS)

intracellular protein degradation pathway that plays a crucial role in maintaining protein quality control and homeostasis. In this system, proteins are tagged with ubiquitin for subsequent degradation by the proteasome

X-inactivation

the process of transcriptionally silencing one of the two copies of the X-chromosome to prevent increased dosage of genes in the X-chromosome