UBE3A: Bridging the gap between neurodevelopment,neural function,and neurodegenerative woes

Introduction

Post-translational modifications (PTMs) play an essential role in normal protein processing. It is also shown to be instrumental in disease pathogenesis. Ubiquitination of proteins is critical for signaling a protein for degradation with the ubiquitin-proteasome system (UPS). There has been much interest in UPS in the study of neurodegenerative diseases due to its role in the removal of misfolded, defective, and aggregation-prone proteins; accumulation of misfolded, defective, and aggregate-prone proteins has been found to accumulate in neurodegenerative diseases (reviewed in ¹ ). The degradation of misfolded or aggregated proteins by the UPS is a multistep ATP-dependent proteolytic mechanism. It involves a pathway of ubiquitin (Ub) transfer steps from E1 to E2 to E3 ligases. The E3 ligase transfers Ub to a target protein which directs that protein to the proteasome for degradation. There are an estimated 600–700 E3 ligases that enable ubiquitination of specific substrates.

It has been proposed that the large amounts of aggregation-prone proteins in neurodegenerative diseases may overwhelm the UPS and compromise other essential functions that UPS mediates for maintaining cellular homeostasis. However, the scenario is most likely more complicated due to other functions of ubiquitination such as orchestrating protein localization, activity modulation, endocytosis, transcriptional control, and autophagy. ¹ Ubiquitin chain addition and substrate specificity are determined by a large spectrum of ubiquitin-ligating and -modifying enzymes, E3 ligases, whose expression levels and activities are tightly regulated in a cell-specific manner. While most ubiquitin chains can target proteins for proteasomal degradation, ubiquitination can contribute to other functions within the cell, including protein localization, protein activity, endocytosis, transcription, and autophagy. ¹ The fate of a given substrate protein can be determined by the type of ubiquitin assembly to which it is connected. This is possible because ubiquitin contains seven Lys residues in its amino acid sequence at positions 6, 11, 27, 29, 33, 48, and 63, which can serve as acceptors for additional ubiquitin monomers in the construction of polyubiquitin chains. One E3 ligase, UBE3A, has garnered much interest of late because of its involvement in synaptic function, UPS activity, and possibly neurodegenerative diseases. This review will present the current knowledge of UBE3A and its involvement in neurological disorders.

UBE3A

UBE3A, also known as E6-associated protein (E6AP), is an E3 Ub-ligase and the founding member of the HECT (homologous to E6-AP C-terminus) domain family. As the name suggests UBE3A was found because of its association with the human papillomavirus (HPV) E6 oncoprotein. It was identified that the E6 protein of HPV can highjack UBE3A to induce ubiquitination and degradation of tumor suppressor p53, mechanistically contributing to cervical cancer.^2,3 UBE3A is a 100 kDa protein (852 amino acids) that structurally possesses a Zn²⁺-binding N-terminal (Amino-terminal Zn-finger of Ube3a Ligase or AZUL) domain and a catalytic HECT domain at the C-terminus (Figure 1A). There are other regions in UBE3A that are required for binding to the HPV E6 protein as well as binding to the HERC2 protein, which appears to be involved in the regulation of UBE3A activity. In humans, there are three UBE3A isoforms that are distinct only in their N-terminus, with isoforms 2 and 3 having 23 and 20 additional amino acids, respectively, compared to isoform 1 (Figure 1B). Isoform 1 appears to be the predominant isoform. ⁴ It is currently unknown if these isoforms have distinct functions in vivo or not, but they may have different localization pattern within cells. ⁵

OPEN IN VIEWER

Figure 1.

Diagrammatic representation of UBE3A structure. (A) UBE3A domains; (B) Sequence alignment for human variants of UBE3A. Variants 2 and 3 contain the completely identical sequence of variant 1 but have additional sequences at their N-terminal region as shown above in underline.

UBE3A is expressed in all cell types, but its cellular substrates and the pathways in which E6AP participates have not been fully elucidated. Several cellular proteins that interact with E6AP have been identified, but the physiological relevance of any of these proteins has not been established. Some of the proteins identified as UBE3A substrates include Small Conductance Potassium Channel (SK2), X-linked inhibitor of apoptosis (XIAP), Ephexin V, Brain and Muscle ARNT-Like 1 (BMAL1), 26S subunit, non-ATPase 4 (PSMD4), and phosphotyrosyl phosphatase activator (PTPA) (reviewed in ⁶ ). UBE3A may also have activity outside of protein degradation. For example, it has been shown that UBE3A can regulate the Arc protein at the transcriptional level and does not directly target the Arc protein for ubiquitination. ⁷ This suggests a role for UBE3A in the nucleus to regulate the transcription of important proteins in neurons.

UBE3A is involved in the regulation of the UPS machinery, with loss of UBE3A activity leading to decreased proteasome function.^8–12 This in part may be due to the interaction between UBE3A and PMSD4 (also known as Rpn10/S5a) which is part of the regulatory lid complex and known to bind to polyubiquitin chains resulting in their recruitment to the proteasome. Interestingly, point mutations in the AZUL domain inhibit this interaction and can lead to Angelman syndrome. ¹³ Ramirez et al. identified eight proteins of the integral or regulatory subunits of proteasome. ¹² Investigation of one of these proteins, DDI1, revealed that its levels were not altered with increased UBE3A expression, suggesting that DDI1 is not being targeted for degradation, but its ubiquitination may be involved in protein regulation. UBE3A may also participate in protein folding quality control. In particular, UBE3A interacts with the chaperone protein HSP70 and promotes the degradation of HSP70-bound substrates. ¹⁴

Autism spectrum disorders and UBE3A

UBE3A has garnered much attention because of its involvement in learning and memory as well as its association with neurodevelopmental autism spectrum disorders (ASDs). The genetics of ASDs is highly heterogeneous, but identified genes show strong enrichment for defects in cell adhesion and mobility, cytoskeleton regulation, ubiquitination pathway, synapse formation, and kinase signaling.^15–17 One major chromosomal region implicated in ASDs is 15q11-q13, which encompasses the UBE3A gene. ¹⁷ Whole-genome copy number variations (CNV) studies in ASDs have identified that this region contributes significantly to the etiology of ASD.^17–20 Deletions in the 15q region is associated with two distinct neurodevelopmental disorders: Prader-Willi syndrome (PWS) and Angelman syndrome (AS). AS and PWS diseases have distinct phenotypes due to maternal or parental inheritance respectively.^21,22 Increased UBE3A is known to cause a related neurodevelopmental disorder Dup15q, which results from the duplication of a portion of the 15q11.2–13.1 chromosome.^23,24

AS is possibly the most studied disease in relation to UBE3A. AS is characterized by severe cognitive and motor deficits, seizures, abnormal EEGs, speech impairments, sleep disturbances, and a generally happy demeanor. ²⁵ The majority of AS patients (∼70%) have a de novo deletion within 15q11-q13 on the maternal chromosome but AS can be caused via other mechanisms such as imprinting defects of the maternal copy (∼6%), paternal uniparental disomy (∼3%), and mutations within the maternal UBE3A (∼13%).^26,27 Despite the profound and penetrant symptoms in AS, there are no gross anatomical aberrations noted in either the AS human brain or in the currently studied animal models.^25,28–31 UBE3A undergoes neuron-specific imprinting, which leads to transcriptional silencing of the paternal allele, and predominant maternal UBE3A expression within the brain. ³² Thus, a maternal disruption of UBE3A results in a > 95% reduction of neuronal UBE3A within the central nervous system (CNS) which leads to the manifestation of AS symptoms. This makes AS-modelled animals ideal for studying the contribution of UBE3A to cognition and synaptic plasticity.

Using a rat model of AS, UBE3A protein was identified in the CSF and in the interstitial fluid of the brain. ³³ The levels of extracellular UBE3A appeared to be regulated in an activity-dependent manner with increases in interstitial UBE3A during the fear conditioning task. ³³ Interestingly, recombinant UBE3A protein incubated with hippocampal slices from AS rats for 30 min before initiating long-term potentiation (LTP) recordings improved the LTP deficit. ³⁴ Furthermore, this improvement translated in vivo with a rescue of the fear conditioning deficit in the AS rat after direct hippocampal injection of recombinant UBE3A protein. ³⁴ One potential target for extracellular UBE3A could be SK2 receptors (small conductance Ca2+-activated K + channels) which are regulators of NMDA receptor function and have been reported to be direct targets of E6AP ubiquitination. Deficits in Ube3a lead to increased SK2 levels, directly impacting NMDAR activation and consequently, impairing LTP. ³⁵ UBE3A could potentially affect other synaptic receptors or membrane proteins such as APP via ubiquitination, which may regulate their function and/or interaction with binding partners. However, currently, nothing is known about this potentially exciting functionality of UBE3A in synaptic activity. Further studies are required to understand the mechanistic function involved.

Increased levels of UBE3A are known to cause a related neurodevelopmental disorder called Dup15q, stemming from the duplication of a portion of the 15q11.2–13.1 chromosome. The clinical symptoms of Dup15q are similar to those observed in AS but typically lack the severe ataxia seen in AS.^23,24 While Dup15q results in the duplication of the UBE3A gene, the pathogenic role of increased UBE3A levels in Dup15q syndrome has not been definitively proven, and other genes in this chromosomal region could be contributing. It has been demonstrated in animal models that increased gene dosage of Ube3a results in glutamatergic synaptic transmission suppression from reduced presynaptic activity and postsynaptic action potential coupling. ³⁶ We have also observed a decrease in the presynaptic response in wild-type slices incubated with the exogenous recombinant UBE3A protein. ³⁴ Yi et al. ³⁷ identified a missense mutation in UBE3A linked to autism. This mutation blocked the phosphorylation of UBE3A at T485 which is normally involved in reducing UBE3A activity, thus enhancing its activity and increasing protein turnover leading to a greater dendritic spine development in the brain.

There is an overlap of symptoms between ASDs and neurodegenerative diseases such as Alzheimer's disease (AD). Dementia, insomnia, and weak neuromuscular interaction, as well as communicative and cognitive impairments, are shared symptoms. ³⁸ This has garnered much interest in potential associations of dysregulated pathways between these diseases. Further work is essential in this direction to mechanistically understand this overlap.

Polyglutamine disease

UBE3A protein declines across the lifespan in human, monkey, and cat cortex with substantial losses of 50–80%. ³⁹ Similar findings were also observed in the aged mouse. ⁴⁰ Loss of UBE3A in the human cortex highlights a specific vulnerability in human brain aging that may signify a dramatic shift in the balance of protein ubiquitination and degradation. This could potentially have profound effects on protein aggregation diseases such as polyglutamine diseases. Polyglutamine diseases are a group of inherited neurodegenerative disorders caused by an aberrant expansion of CAG repeats in the causative genes. Disorders caused by such expansions include Huntington's disease (HD), spinal and bulbar muscular atrophy (SBMA), dentatorubral pallidoluysian atrophy, and several spinocerebellar ataxias (SCA1, SCA2, SCA3, SCA6, SCA7, and SCA17). ⁴¹

Using an in vitro system, Mishra et al. showed that decreasing UBE3A enhances the rate of polyglutamine aggregate formation and cell death mediated by the polyglutamine proteins. ⁴² Conversely, over-expression of UBE3A results in reduced polyglutamine protein aggregation and protection from protein aggregate-induced cell death. ⁴² Similarly, Bhat et al. showed that increasing UBE3A can reduce mutant Huntingtin (mHtt) protein and that this effect was blocked with the UPS inhibitor MG132. ⁴⁰ UBE3A was further shown to colocalize with Htt aggregates with a greater preference for longer polyQ repeats.^14,40,42,43 Protection from protein aggregation was more prominent when UBE3A was co-expressed with Hsp70. ⁴² It appears that UBE3A interacts with the substrate binding domain of Hsp70/Hsc70 chaperones and promotes the degradation of chaperone-bound substrates. UBE3A is upregulated in vitro under a variety of stresses, including in expanded polyglutamine protein-expressing cells and cells exposed to proteasomal stress,^14,42 which may suggest an attempt to compensate for the increased protein aggregation. This cellular stress leads to the redistribution of UBE3A around the microtubule-organizing center, a subcellular structure for the degradation of the cytoplasmic misfolded proteins. ¹⁴

HD is an autosomal dominant mutation in chromosome 4 (4p16.3) which leads to an increased CAG trinucleotide repeat expansion in the huntingtin gene (Htt). The number of repeats correlates to the risk of HD development. HD symptomology generally appears in middle age with onset typically ranging from 30–50 years of age. Symptoms include behavioral abnormalities, motor dysfunction, psychiatric imbalances, cognitive loss, or a combination of these. ⁴⁴ In vivo, the removal of Ube3a selectively from the HD mice brain results in accelerated disease pathology and motor phenotype with a shorter lifespan in comparison to HD mice with normal UBE3A levels. ⁴³ There was an overall increase in global aggregate load, and these aggregates were less ubiquitinated when compared with age-matched HD mice. ⁴³ Similarly, Purkinje cell loss is highly accelerated in spinocerebellar ataxia type 1 (SCA1) mice with reduced UBE3A. ⁴⁵ UBE3A was found to be reduced in HD mice with age and there was a decrease in association of UBE3A and Htt proteins with age, which may account for the increased K63-ubiquitination and subsequent aggregation of Htt protein. ⁴⁰ Similarly, Maheshwari et al. found UBE3A within aggregates, reducing soluble active UBE3A. ⁴⁶ This led to increased levels of Arc and decreased GluR1 puncta at synapses, ⁴⁶ which is consistent with the AS phenotype.^7,47

Overall, these published studies suggest that age-dependent UBE3A levels result in differential ubiquitination and degradation of Htt fragments, thereby contributing to the age-related neurotoxicity of mHtt. Interestingly in HD mice, rAAV-mediated expression of UBE3A reduced K63-ubiquitination and subsequent aggregation of mHtt protein. ⁴⁰ In HD mice treated with Azadiradione (a bioactive limonoid), there was improved disease pathology, which was associated with increased HSF1 and UBE3A levels. ⁴⁸ Similarly, treatment of HD mice with Topotecan (a topoisomerase inhibitor) resulted in improved motor and pathology phenotypes. ⁴⁹ Topotecan treatment increased the expression of UBE3A in these HD mice, ⁴⁹ which is consistent with previous studies in AS mice which showed that Topotecan can induce paternal UBE3A gene expression. ⁵⁰ These data suggest that regulating UBE3A activity may be a viable approach to reduce HD pathology.

Alzheimer's disease

AD is the most common form of dementia, with a progressive loss of memory and cognitive function due to neurodegeneration. Pathologically, there is a development of amyloid plaques and hyperphosphorylated tau tangles, and neurodegeneration. There are significant links between UBE3A and AD. AS patients (lacking neuronal UBE3A) have increased Aβ compared to controls, ⁵¹ and Ube3a duplication models show decreased APP levels. ⁵² Thus, UBE3A may be involved in regulating APP, including its processing. In line with this, it has been reported that ASD patients are 2.6 times more likely to develop early-onset AD. ¹⁸ Transcriptomic analysis of asymptomatic-AD and AD patient brains showed 30–40% decreases in UBE3A levels in the temporal lobe and entorhinal cortex. ⁵³ A study using genome-wide association neural networks (a novel approach that uses neural networks to account for nonlinear and SNP-SNP interaction effects) of patients with a family history of AD in the UK Biobank identified known AD gene associations but also identified 67 novel gene associations including genes linked with ASDs (UBE3A, BDNF, ASIC2, NTRK2, CTNNA3). ⁵⁴

In the amyloid mouse models, Tg2576 and APPswe/PS1, there were reduced UBE3A protein levels, and oligomers of Aβ can reduce UBE3A expression in rat primary neurons.^55,56 In the APP mouse models, reduced UBE3A protein was observed as an early event that occurred before cognitive impairments and is likely mediated via c-Abl phosphorylation which leads to enzyme inactivation and degradation. It has been proposed that loss of UBE3A activity leads to an increase in its downstream UBE3A targets, namely Ephexin-5 and Arc, which could account for the loss of dendritic spine density and synaptic dysfunction observed in the Tg2576 mice. ⁵⁵ It is well established in the Angelman field that deletion of Ube3a in mice reduces dendritic spine density and length in hippocampal CA1, Layer III-V cortical neurons, basal dendrites of Layer II/III cortical neurons, and Layer V neurons in the visual cortex.^57,58

It is known that soluble PHF-tau is ubiquitinated at residues K-254, K-311, and K-353, in its microtubule-binding domain, suggesting that ubiquitination of PHF-tau may be an early pathological event and that this ubiquitination could play a regulatory role in modulating the stability of microtubules during the course of AD. ⁵⁹ Oligomeric-tau (oTau) with K63 but not K48 ubiquitination was found to accumulate in AD brains. ⁶⁰ The K63 ubiquitination of tau was also associated with enhanced seeding and propagation of oTau in primary neurons. ⁶⁰ Examination of the conformational antibodies Alz-50 (which detects an early conformational change in oTau) and Tau-66 (which detects a late conformational change) found that ubiquitin was associated with the early conformational change of tau but significantly less with the late conformational change. ⁶¹ This indicates that a more intact conformation of the N-terminus of tau may facilitate tau ubiquitination, and conversely that ubiquitination may be reduced when tau is truncated or when its N-terminus is conformationally masked due to aggregation. Some studies have suggested that E3 ubiquitination of tau can decrease rates of elongation and secondary nucleation, thus interfering with fibril formation. ⁶² Thus, it appears that selective oTau ubiquitination could influence tau aggregation, accumulation, and subsequent pathological propagation. K48 ubiquitination likely protects neurons from the toxicity of oTau by promoting proteasome degradation and or insoluble neurofibrillary tangles formation. Promoting tau ubiquitination with elevated proteasome activity might provide an effective avenue for AD treatment by reducing tau protein levels. Direct evidence for this comes from studies in Drosophila where overexpression of UBE4B, an E4 ubiquitin elongation enzyme, promoted tau ubiquitination and degradation and alleviated eye neurodegeneration. ⁶³

Further work is needed to identify which E3 ligases are capable of ubiquitinating tau in vivo, and in particular, the ones that may contribute to K48 targeted UPS degradation versus K63 ubiquitination, which appears to increase the more pathological oTau. ⁶⁰ It should be noted that UBE3A was identified as a K48 ligase; ⁶⁴ thus, its decrease in AD could have profound effects on tau pathology. We propose that UBE3A offers a novel therapeutic target to enhance the degradation of pathogenic tau potentially through UBE3A's ubiquitination of tau but also through its ability to activate the UPS.^8,10,11

Concluding remarks

Current research indicates that UBE3A functions as a cellular quality control ubiquitin ligase, playing a crucial role in maintaining protein homeostasis. Consequently, its dysfunction is implicated not only in the pathogenesis of neurodevelopmental disorders such as AS but also in the biology of other disorders (Figure 2). In neurodegenerative diseases with increased protein aggregation, UBE3A may offer a novel pathway to target to improve neuronal cell homeostasis and reduce protein deposition. However, more studies are required to establish UBE3A's molecular role in these neurodegenerative diseases. Interestingly, HPV may have evolved to interact with UBE3A because of its strong link to the UPS; if so, this relationship may offer an ideal target for approaches such as proteolysis targeting chimera (PROTAC). PROTACs are designed to harness the cell's own degradation machinery to selectively degrade pathogenic proteins. This may be enhanced through association with UBE3A.

OPEN IN VIEWER

Figure 2.

Altered UBE3A activity and its relationship with neurological diseases. Both loss of function and altered activity of UBE3A contributes to neurodegenerative and neurodevelopmental diseases.

Finally, understanding the age-dependent decline of UBE3A remains a complex puzzle. Several potential mechanisms can be hypothesized. For example, it may reflect a decline in the overall number of neurons and their synapses with aging. However, proteomics from the human cortex showed that the loss of UBE3A was substantially greater than that of other synaptic proteins, suggestive of a selective loss of UBE3A. ³⁹ Alternatively, there could be a general slowdown of ubiquitination with age. Epigenetic changes with age could also play a role. Moreover, it is known that DNA methylation and histone modifications change with age^65–68 and since UBE3A is highly regulated with only the maternal gene is transcribed in neurons, changes in methylation could lead to altered UBE3A gene expression. Examination for any age-dependent changes in maternal and paternal epigenetic changes of UBE3A is needed to substantiate this possibility. Lastly, disruption of normal synaptic activity might contribute to decreased UBE3A expression. Studies have shown that experience-driven neuronal activity in adult mice increases UBE3A expression in the hippocampus.^69,70 We have also shown that UBE3A is released into the interstitial fluid and that this release appears to be regulated in an activity-dependent manner. ³³ This could lead to potential negative feedback as reduced neuronal activity leads to reduced UBE3A expression which leads to reduced synaptic function and plasticity. Further work is needed to determine the full scope of experience-dependent changes in UBE3A expression in the brain and especially with aging. Additionally, transient overexpression of UBE3A in the brain in young and old mice, for example, might help define the impact of an age-dependent reduction of UBE3A in the brain.