C-terminus of Hsp70 Interacting Protein (CHIP) and Neurodegeneration: Lessons from the Bench and Bedside.

Neurodegenerative diseases are characterized by the increasing dysfunction and death of neurons, resulting in progressive impairment of a person's mobility and/or cognition. Protein misfolding and aggregation are commonly hypothesized to cause neurotoxicity and, eventually, neuronal degeneration that are associated with these diseases. Emerging experimental evidence, as well as recent findings from human studies, reveal that the C-terminus of Hsp70 Interacting Protein (CHIP), or STIP1 Homology and U-box containing Protein 1 (STUB1), is a quality control protein involved in neurodegeneration. Here, we review evidence that CHIP interacts with and plays a role in regulating proteins implicated in the pathogenesis of Parkinson's disease, Alzheimer's disease, amyotrophic lateral sclerosis, and polyglutamine diseases, including Huntington's disease and spinocerebellar ataxias. We also review clinical findings identifying mutations in STUB1 as a cause of both autosomal recessive and autosomal dominant forms of cerebellar ataxia. We propose that CHIP modulation may have therapeutic potential for the treatment of multiple neurodegenerative diseases. Copyright© Bentham Science Publishers; For any queries, please email at epub@benthamscience.net.

INTRODUCTION

Neurodegenerative diseases are chronic neurological conditions that include disorders such as Parkinson’s disease (PD), Alzheimer’s disease (AD), amyotrophic lateral sclerosis (ALS), and inherited cerebellar ataxias. These diseases are all incurable at present, with no available treatments to modify the neurodegenerative process or provide neuroprotection. Age is a major risk factor for most neurodegenerative diseases and thus, without such disease-modifying or neuroprotective treatments, neurodegenerative diseases will represent an increasingly significant medical and public health concern as life expectancy increases and populations age worldwide [1, 2].

All neurodegenerative diseases are defined by progressive neuronal dysfunction and death. They are associated with a gradual decline in a person’s mobility and/or cognition. Specific symptoms of each disease reflect the neuronal populations affected and, while an explanation for the selective vulnerability of discrete neuronal populations in different disorders remains to be elucidated, mounting evidence indicates that certain cellular and molecular mechanisms are shared across these diseases. Importantly, protein misfolding and aggregation are associated with neurotoxicity and eventual neuronal loss [3].

Quality control systems exist within each cell to maintain protein homeostasis, or ‘proteostasis’. Proteostatic systems include the chaperone system, the ubiquitin-proteasome system (UPS), and autophagy-lysosomal pathway (ALP). Together these systems serve to reduce or prevent the accumulation of misfolded proteins and aggregates. Deciphering the role of quality control proteins is especially important for understanding the pathogenesis of neurodegenerative diseases and developing disease-modifying therapies. This review discusses the role of one such quality control protein, the co-chaperone C-terminus of Hsp70 Interacting Protein (CHIP). Emerging experimental evidence, together with clinical findings, suggest that CHIP may play a critical neuroprotective role in multiple neurodegenerative diseases.

C-TERMINUS OF HSP70 INTERACTING PROTEIN (CHIP)

The chaperone system is comprised of a family of proteins, including molecular chaperones and co-chaperones, that work together to facilitate de novo protein folding as well as refolding of misfolded proteins. Chaperones, such as Heat Shock Protein 70 (Hsp70) and Heat Shock Protein 90 (Hsp90), directly interact with newly synthesized peptides to assist with proper folding and to prevent aggregation into potentially cytotoxic assemblies [4]. Co-chaperones also assist with folding by interacting with chaperones to modulate chaperone activity or by linking chaperones to other proteostatic systems [5]. When proteins are irreversibly misfolded or aggregated, they are targeted for degradation via the UPS or ALP, which often occurs in a chaperone-dependent manner [4, 6]. The activity of chaperones and co-chaperones is therefore, critical to multiple cellular pathways due to their central role in maintaining proteostasis.

CHIP, or STIP1 Homology and U-box containing Protein 1 (STUB1), is a 34.5 kDa co-chaperone encoded by the STUB1 gene. STUB1 is located on chromosome 16 in humans, contains 7 exons, and has 2 alternative splicing transcript variants. STUB1 is highly conserved across species and is expressed in most tissues [7]. CHIP contains an N-terminal tetratricopeptide (TPR) domain and a C-terminal U-box domain. The TPR domain interacts with the chaperones Hsp70 and Hsp90 [7], whereas the U-box domain confers E3 ubiquitin ligase activity [7, 8]. The TPR and U-box domains are connected by a central, highly-charged, coiled- coil domain that mediates the homodimerization of CHIP [9, 10]. CHIP homodimers exhibit conformational flexibility, allowing for both symmetrical and asymmetrical dimeric conformations [11-13], with the asymmetric conformation being necessary for the catalytic activity of the U-box domain [12, 9]. The flexible conformation of CHIP homodimers also allows interaction with a diverse range of substrates [11]. Through these domains, CHIP has a range of functions it can exert through chaperone protein networks and, by extension, various cell systems.

Ubiquitin is a small protein that can be covalently linked to proteins to facilitate a number of diverse functions [14-17]. Substrates can be mono- or poly-ubiquitinated, and the location of attachment between conjoined ubiquitin molecules ultimately determines the type of chain as well as the ultimate fate of the substrate [14, 18]. CHIP regulates the ubiquitination of chaperone-bound proteins by directly engaging chaperones via its TPR domain. This interaction results in the formation of a dynamic complex that promotes conformational changes in the CHIP protein, allowing it to interact with and ubiquitinate chaperone-bound proteins. Specifically, the U-box domain mediates monoubiquitination or polyubiquitination of chaperone-bound targets or even chaperones themselves [19]. In addition to ubiquitinating chaperone-bound targets, CHIP can also ubiquitinate itself, in a process referred to as auto-ubiquitination [20], which may represent an important capacity of the protein to self-regulate. Polyubiquitination requires cooperation between CHIP and E2 ubiquitin-conjugating enzymes. The E2 enzymes that co-ordinate with CHIP have an important role in determining whether CHIP will ubiquitinate itself or other targets [20, 21] and the type of ubiquitin chain formed [22]. CHIP can catalyze lysine-27 (Lys-27), Lys-48, and Lys-63 linked polyubiquitin chains [9, 23-26]. Lys-48 polyubiquitin chains are canonically associated with targeting proteins to the UPS for proteasomal degradation [14, 15, 27], although non-proteolytic functions of these chains have also been reported [17]. Lys-63 chains are largely associated with non-proteolytic functions, including DNA damage response, kinase activation, endocytosis, signal transduction, and selective autophagy, but targeting for proteasomal degradation has also been reported [17, 28]. Lys-27 chains have been associated with mitochondrial damage response, signal transduction, and proteasomal degradation [14, 23, 29, 30]. As CHIP can catalyze the formation of different types of polyubiquitin chains, CHIP-mediated ubiquitination of substrates can serve both proteolytic and non-proteolytic functions [23].

In addition to its function in the UPS, CHIP has an important regulatory role in autophagy. CHIP has been demonstrated to promote the degradation of the inactive phosphorylated form of transcription factor EB (TFEB), a pro-autophagy transcription factor [31, 32], increasing TFEB activity and autophagy [33]. CHIP-deficient cells exhibit impaired chaperone-mediated autophagy (CMA) [34]. Specifically, autophagosome accumulation and a lack of autophagosome-lysosome fusion have been observed in these cells. CHIP has been shown to direct the degradation of some proteins that contain a KFERQ-like motif through interactions with the chaperone Hsp70 protein 8 (HSPA8) [34]. These interactions are necessary to degrade proteins in the lysosome because they are required to target substrates to the lysosomal receptor, lysosomal-associated membrane protein 2A (LAMP2A) [34]. Since Lys-63 linked polyubiquitin chains can target substrates for lysosomal degradation via the ALP pathway [17, 35], CHIP-mediated ubiquitination may also be targeting substrates for lysosomal degradation. Indeed, CHIP has been shown to target Hypoxia-Inducible Factor 1 (HIF1) for lysosomal degradation via the ALP pathway through the addition of Lys-63 linked ubiquitin chains [36]. CHIP has also been demonstrated to localize to mitochondria within primary neurons following stress, and has been proposed to enhance stress-induced mitochondrial autophagy (mitophagy) [37]. In CHIP knockout mice, expression of genes involved in the initiation of autophagy and mitophagy were found to be decreased in cardiac muscle following treatment with fenofibrate (a PPARα agonist which stimulates oxidative metabolism), but not in wild-type controls [38], while voluntary exercise increased cardiac autophagy in CHIP null mice compared to controls [39], indicating that CHIP may influence autophagy differently in response to different stressors. With its connections to chaperones, the UPS, and the ALP, CHIP is well positioned to triage misfolded or aggregated proteins for refolding or degradation [40].

The autophagic activity has been found to decline in aging animal models [41-43], and age-related changes in CHIP protein levels have also been observed in vivo. CHIP protein levels, but not mRNA levels, were found to decrease in the brains of male wild-type C57BL mice with age [44]. CHIP knockout mice also exhibit accelerated aging [45] and senescence-accelerated mouse models have decreased CHIP mRNA and protein levels compared to senescence-resistant controls [46, 47]. These findings suggest anti-aging functions for CHIP and are consistent with a role in regulating autophagic pathways.

CHIP has additional functions that appear to be independent of ubiquitination and its E3 ubiquitin ligase activity. For example, CHIP has been found to function as an autonomous chaperone that promotes conformational changes of adenosine monophosphate-activated protein kinase (AMPK), resulting in increased AMPK protein stability and kinase activity [48]. CHIP was also shown to promote the trimerization, activation, and nuclear localization of heat shock factor-1 (HSF1) [49-51], a transcription factor that promotes the transcription of various heat shock proteins in response to misfolded protein accumulation. Whether these functions of CHIP are related to neurodegeneration, remain to be elucidated. Regardless, they demonstrate that the CHIP function is not restricted to its E3 ligase activity.

Post-translational modifications of CHIP have been demonstrated to impact its function (Fig. ). Cyclin-Dependent Kinase 5 (CDK5) was found to phosphorylate murine CHIP at serine 20, which is located N-terminal to the TPR domain. Phosphorylation at this site reduced turnover of truncated Apoptosis-Inducing Factor (tAIF) by CHIP and the UPS via a mechanism that involved an alteration of CHIP’s binding affinity to the protein without affecting its E3 ligase activity. In contrast, aurora kinase A (AURKA) phosphorylates CHIP at serine 273 and appears to regulate its E3 ligase activity, promoting the ubiquitination and proteasomal degradation of androgen receptor (AR) [52]. The Auto-ubiquitination of CHIP has been shown to assist with translocation to the proteasome. This did not cause the degradation of CHIP but helped to facilitate the degradation of the substrate bound to it [53]. CHIP can also monoubiquitinate itself with the help of Ubiquitin-conjugating enzyme E2 W (UBE2W) following cellular stress, promoting its interaction with and ubiquitination of ataxin-3 [54]. The full breadth and functional significance of these and other potentially undescribed post-translational modifications (PTMs) of CHIP remain to be elucidated as they could be important regulators of CHIP activity and substrate preference.

CHIP is expressed in many different tissues, but it is most highly expressed in the brain, skeletal muscle, and cardiac muscle, suggesting its importance in systems with high metabolic activity [40, 55]. While CHIP is widely expressed throughout the brain, there is a particular abundance within the Purkinje layer of the cerebellum [55]. Stub1 knockout mice show partial lethality during the perinatal development stage [45]. Knockout mice were significantly smaller than wild-type mice and showed an accelerated aging phenotype that resulted in a significantly decreased lifespan [45]. Stub1 knockout mice also displayed severe ataxia, cognitive impairment, hypogonadism, and severe impairments in mitochondrial stress response [55, 56]. Brain lysates from Stub1 knockout mice showed a significant increase in misfolded proteins compared to age-matched or older wild-type controls [45]. Examination of the cerebella of these animals demonstrated a specific and drastic loss of Purkinje cells with increased pyknotic nuclei and severe dendritic swelling [57]. This evidence indicates that aberrations in CHIP can cause significant impairments in multiple physiological processes, including development and neurodegeneration.

In this review, we discuss evidence of the involvement of CHIP within various neurodegenerative diseases, each associated with abnormal aggregation of endogenous macromolecules. Given that CHIP is an E3 ligase that promotes the elimination of mutated, misfolded, or aggregated proteins via the UPS and ALP, we hypothesize that CHIP may be involved in each of these neurodegenerative diseases as a common downstream effector, facilitating the removal of a variety of disease-associated proteins. Consequently, CHIP modulation may have therapeutic potential for the treatment of multiple neurodegenerative diseases.

CHIP AND ITS MOLECULAR INTERACTIONS RELEVANT TO NEURODEGENERATIVE DISEASES

Parkinson’s Disease (PD)

PD is the most common neurodegenerative movement disorder and is associated with the classical ‘parkinsonian’ motor features (bradykinesia, rigidity, tremor, and postural instability) as well as multiple non-motor features, including cognitive impairment. PD is characterized by prominent neuronal loss in multiple brain regions, primarily the dopaminergic neurons of the substantia nigra pars compacta (SNpc). Within degenerating areas of the brain, there is a presence of intraneuronal inclusions, called Lewy bodies and Lewy neurites, which are primarily composed of aggregated α-synuclein (α-syn) protein [5]. The normal turnover of α-syn utilizes UPS and CMA-mediated degradation [58]. Aggregation of α-syn, due to excessive and/or misfolded proteins, typically follows a process in which misfolding is followed by dimerization, oligomerization, protofibril formation, fibrilization and finally, tightly packed Lewy body formation [59]. Aggregated α-syn has been shown to bind to and inhibit the action of the proteasome, an early-onset event which contributes to neurodegeneration [5, 60]. As a consequence, the ALP is the predominant pathway that cells utilize to degrade α-syn aggregates [58].

Experiments performed in cell culture indicate that CHIP, Hsp70, and α-syn form a complex [8, 61], and examination of post-mortem PD brain tissue demonstrates that they also co-localize within Lewy bodies [8]. Several lines of evidence suggest that Hsp70 can inhibit α-syn aggregation and thereby prevent the neuronal loss, including findings in cell culture models, transgenic Drosophila melanogaster, and mouse models [8, 59, 62]. Overexpression of CHIP has been shown to increase the rate of α-syn clearance from cells [8, 61] and reduce α-syn aggregation in rat brains in vivo [63]. CHIP appears to preferentially target specific oligomeric confirmations for degradation [64] via both proteasomal and lysosomal mechanisms [8]. Through direct or indirect interactions mediated by Hsp70, CHIP can ubiquitinate α-syn in vitro [8, 61, 64]. The co-chaperone BCL2 Associated Athanogene 5 (BAG5) can interact with CHIP via Hsp70 and inhibit the E3 ubiquitin ligase activity of CHIP. As a result, BAG5 causes a reduction in α-syn ubiquitination and mitigates the ability of CHIP to reduce α-syn oligomers [61]. Taken together, these data suggest that Hsp70 and CHIP may be important regulators of α-syn pathology.

In addition to α-syn, endonuclease G (EndoG) has been identified as a substrate of CHIP. EndoG is a mitochondrial localised DNase that initiates apoptosis following its nuclear relocalization [65]. EndoG levels have been found to be elevated and increasingly localized within the nucleus in post- mortem PD SNpc tissue compared to healthy controls. Furthermore, EndoG has been found to be crucial for α-syn mediated dopaminergic degeneration in vitro and in vivo [66]. CHIP has been shown to ubiquitinate and promote proteasomal degradation of EndoG, resulting in reduced EndoG protein levels, protecting against cell death [67].

While most PD cases are sporadic, some forms are inherited, such as those caused by mutations in parkin, PTEN-induced protein kinase 1 (PINK1), or leucine-rich repeat kinase-2 (LRRK2) [68]. Mutations in parkin cause autosomal recessive PD [69]. CHIP plays a role in positively regulating the activity of parkin, an E3 ligase that mediates ubiquitination via a RING finger domain [70]. Parkin is part of the PINK1-parkin mitochondrial clearance pathway which acts to remove damaged mitochondria from cells in a selective manner via mitophagy [69]. Parkin acts to ubiquitinate proteins on the outer-mitochondrial membrane to facilitate the degradation of depolarized mitochondria through the ALP [70]. In Drosophila, knockout of Parkin leads to loss of dopaminergic neurons, abnormal wing posture, thoracic indentation, muscle degeneration, and locomotor deficits. Parkin-null Drosophila also exhibit mitochondrial abnormalities including excessive fusion, reduced ATP production, and disorganized crista. Overexpression of CHIP in Parkin-null Drosophila suppressed dopaminergic neuron loss, as well as locomotor and mitochondrial defects. Knockout of CHIP in Parkin-null Drosophila exacerbated Parkin-mediated defects and resulted in a significantly reduced Drosophila lifespan [71]. Together, these findings suggest that CHIP has a critical protective role in the context of parkin dysfunction.

Parkin interacts with probable G-protein coupled receptor 37 (Pael-R), a transmembrane protein that can accumulate inside cells and can promote SNpc cell death, independent of Lewy body pathology [72]. Through this interaction, parkin ubiquitinates Pael-R, promoting its proteasomal degradation [72]. Overexpression of parkin reduced both Pael-R protein levels and aggregate formation [72], whereas parkin inactivation enhanced ER stress and dopamine toxicity-mediated dopaminergic neuron death in the SNpc [73]. CHIP has been shown to outperform Hsp70 for binding to parkin in vitro, potentiating the E3 ligase activity of parkin and promoting parkin-mediated ubiquitination of Pael-R, further supporting a role for CHIP in enhancing the protective effects of parkin [72].

Similar to parkin mutations, PINK1 mutations also cause autosomal recessive PD [74-77]. PINK1 is a serine-threonine protein kinase that, together with Parkin, is known to be highly involved in regulating mitophagy [78]. In Drosophila, knockout of Pink1, similar to knockout of Parkin, leads to loss of dopaminergic neurons, abnormal wing posture, thoracic indentation, locomotor deficits, muscle degeneration, and mitochondrial defects. These mitochondrial defects include excessive fusion, reduced ATP content, and crista disorganisation. Overexpression of CHIP in Pink1-null Drosophila suppressed movement and mitochondrial dysfunctions. Unlike in Parkin-null Drosophila, knockout of CHIP in Pink1-null Drosophila did not exacerbate the phenotype, indicating that CHIP may act downstream of PINK1 [71]. CHIP-null Drosophila exhibit impairments in climbing ability as well as reduced thoracic ATP, reduced thoracic mitochondrial DNA, and abnormal mitochondrial morphology [71]. Importantly, CHIP interacts with PINK1, promoting its polyubiquitylation and proteasomal degradation. Changes in CHIP protein levels are inversely correlated with PINK1 protein levels [79]. CHIP-mediated PINK1 degradation increased cytotoxicity in dopaminergic neuroblastoma SH-SY5Y cells following treatment with staurosporine, an apoptosis inducer [79]. Consistent with this finding, CHIP knockout mice have significantly increased levels of PINK1 protein in whole brain lysates [56].

Mutations in leucine-rich repeat kinase 2 (LRRK2) are the most common cause of autosomal dominant PD [80]. LRRK2 is a large multidomain protein believed to play several roles, including maintenance of neuronal polarity [81, 82]. LRRK2 contains a serine/threonine kinase domain and several PD-causing mutations enhance its kinase activity [72]. CHIP has been shown to interact with LRRK2 in vitro in a manner requiring the TPR domain of CHIP, suggesting that Hsp70 or Hsp90 may mediate this binding [82], as well as weakly binding to the charged coiled-coil domain of CHIP [83]. The ubiquitination of wild-type and mutant forms of LRRK2 by CHIP has been shown to cause proteasomal degradation of LRRK2 [81-83]. Furthermore, levels of CHIP and Hsp90 were found to regulate LRRK2-related cytotoxicity [81]. Taken together, work to date demonstrates a beneficial role of CHIP in modulating the molecular pathogenesis of PD, which may occur, in part, through its interactions with α-syn, EndoG, parkin, PINK1, and LRRK2 (Fig. ).

Alzheimer’s Disease (AD)

AD is the most prevalent neurodegenerative disease and is characterized by progressive cognitive decline, primarily affecting memory, which eventually leads to impairment of a person’s ability to function in daily life and the emergence of behavioural symptoms. All forms of AD appear to share two molecular pathological hallmarks: the presence of intra-neuronal neurofibrillary tangles composed of tau and the deposition of extra-neuronal amyloid β (Aβ) plaques [84]. CHIP protein levels have been shown to be increased in AD patients [85].

Tau stabilizes microtubules, a critical component of the cytoskeleton, and is enriched in neuronal axons. Due to this important structural function, impairment in tau can be highly deleterious and hence its tight regulation is critical for proper neuronal function [86]. Hyper-phosphorylation of tau can cause its detachment from the axonal microtubules and induce aggregation [86]. In AD patient samples, CHIP levels were found to be inversely proportional to sarkosyl-insoluble tau accumulation, and CHIP knockout mice were found to have increased levels of insoluble tau accumulation [85] and soluble phosphorylated tau [87], indicating that CHIP may protect against neurofibrillary tangle formation in AD. Indeed, overexpressing chaperones, such as Hsp70 or Hsp90, has also been shown to reduce tau aggregation [88], and CHIP levels in brains of both AD patients and healthy controls have been found to be highly correlated with Hsp90 protein levels [85]. CHIP co-immunoprecipitated with tau and Hsp70, and CHIP has been shown to ubiquitinate tau to target it to the proteasome for degradation [26]. Hsp90 and CHIP have also been shown to work in concert to refold or degrade aberrant tau [89]. Histone deacetylase 6 (HDAC6) is a deacetylase of Hsp90, which reduced Hsp90-mediated refolding activity, promoting degradation of Hsp90 substrate proteins [90, 91]. Overexpression of HDAC6 has been shown to promote tau accumulation in HeLa cells. Conversely, decreased HDAC6 increased Hsp90-mediated clearance of tau in primary mouse neurons [92]. CHIP has been shown to bind, ubiquitinate, and regulate the expression of HDAC6 [92]. Brain homogenates from CHIP knockout mice exhibited both increased tau accumulation and levels of HDAC6 92]. Thus, HDAC appears to be involved in the interplay between Hsp90 and CHIP in regulating tau accumulation.

Tau phosphorylation by microtubule affinity regulating kinase 2 (MARK2) prevents its recognition by CHIP or Hsp90 [93]. Protein Kinase B (Akt), which can be degraded by Hsp90, has been shown to enhance the activity of MARK2 [89]. Akt knockout mice have decreased levels of CHIP, indicating that CHIP expression may be tied to Akt levels [93]. Interestingly, a reduction in Akt levels still enhanced CHIP-mediated degradation of tau, due to Akt having a higher binding affinity for tau than CHIP [93]. The understanding of CHIP’s regulation of tau requires further investigation and presents an interesting avenue to explore how co-chaperones regulate this protein.

Aβ is formed from its precursor protein, amyloid precursor protein (APP), which is a transmembrane protein enriched at neuronal synapses [94, 95]. Functional roles for APP include signalling, gene transduction, neurite growth, and synaptogenesis [96]. APP has multiple cleavage sites, including sites recognized by α-, β-, and γ-secretases. Cleavage of APP by β-secretase 1 (BACE1) leads to the production of the pathogenic 42 amino acid form of Aβ (Aβ42), an insoluble product that can aggregate to form Aβ plaques [96]. Multiple lines of evidence demonstrate that CHIP can reduce Aβ aggregation either by regulating levels of APP or BACE1. Evidence for the effects of CHIP on APP includes a demonstration that CHIP co-immunoprecipitated with APP from human brain lysates, indicating a possible in vivo interaction [97, 98]. This interaction was shown to either be direct or in a complex with Hsp70 [97]. The ubiquitination of APP by CHIP was shown to target it to the proteasome for degradation [97]. Depletion of CHIP or Hsp70 was associated with an increase in cellular APP levels [97], whereas overexpression of CHIP in an inducible Aβ-producing cell line was found to decrease Aβ accumulation. When primary cortical neurons were challenged with Aβ accumulation, CHIP overexpression increased their survival, supporting a role for CHIP in mitigating Aβ-mediated cytotoxicity [97]. Regulation of BACE1 by CHIP is suggested by the finding of increased expression of BACE1 under CHIP knockdown conditions [98] and potentially indicated by the reduced expression of CHIP in neurons in human AD brains [99]. CHIP knockdown and overexpression have been shown to be associated with increased and decreased BACE1 expression, respectively [98]. Deletion constructs of CHIP lacking either the TPR or U-box domain did not affect BACE1 expression [98], indicating that full-length CHIP protein is required for regulation of BACE1 levels. Similarly, CHIP can ubiquitinate BACE1 but requires both the TPR and U-box domain for this action [98]. Furthermore, CHIP is capable of stabilizing p53, a transcription factor that negatively regulates BACE1, in an active conformation and thereby reduces BACE1 expression [98]. Overall, these findings indicate that CHIP plays a protective role in modulating AD through its interactions with tau, HDAC6, APP, BACE1, and Akt (Fig. ).

Amyotrophic Lateral Sclerosis (ALS)

ALS is a rapidly progressive neurodegenerative disease that causes muscle weakness, disability, and, ultimately death due to selective loss of motor neurons. Pathological protein inclusions, frequently containing TAP DNA binding protein 43 (TDP-43), are found in upper and lower motor neurons as well as glial cells [100]. Patients typically present with a combination of upper and lower motor neuron signs and symptoms, which progress to life-threatening respiratory failure and dysphagia [100]. The median survival from the time of diagnosis is 3 to 5 years [101]. While most cases of ALS are sporadic with no identified cause, approximately 5 to 10 percent of cases are familial with an underlying genetic cause, including mutations in genes such as TDP-43, C9orf72, NIMA-related kinase 1(NEK1), Fused in Sarcoma (FUS), and ubiquilin-2 (UBQLN2). At the time of writing this review, there were no reports of an association between CHIP and these ALS-associated genes. However, an interaction between CHIP and superoxide dismutase 1 (SOD1) has been identified. Mutations in SOD1 account for up to 2 percent of ALS cases [102, 103].

SOD1 is a Cu/Zn superoxide dismutase which prevents damage caused by free radical oxygen molecules. Hsp70 has been shown to interact with SOD1 and to favour interaction with mutant forms over wild-type SOD1 [104]. CHIP expression reduced SOD1 levels, and this effect was attenuated by a proteasome inhibitor, suggesting that CHIP may be mediating proteasomal degradation of SOD1 [104]. Indeed, CHIP has been shown to interact with mutant SOD1 indirectly through Hsp70, which promoted UPS-mediated degradation of mutant SOD1 [104]. Aggregated mutant SOD1 may impair the normal function of the proteasome [104, 105]. In this scenario, SOD1 aggregates can be cleared by Heat Shock Protein beta-8 (HSPB8), a chaperone molecule that can be induced in response to proteasomal inhibition, resulting in reduced size and number of aggregates [105]. HSPB8, together with the co-chaperone BCL2 Associated Athanogene 3 (BAG3), was reported to induce the creation of an autophagosome around the SOD1 aggregate, promoting ALP- mediated degradation [105]. When HSPB8-linked autophagosomes were co-immunoprecipitated, Hsp70, BAG3, and CHIP were present within the protein complex [106]. A CHIP-Hsp70-BAG3-HSPB8 complex has been shown to recruit p62, an autophagy scaffolding protein, promoting autophagosome development around misfolded proteins (Fig. ), highlighting a possible role for CHIP in assisting in the clearance of SOD1 aggregates via both proteasomal and lysosomal mechanisms [106].

CHIP’s ability to degrade mutant SOD1 via two independent mechanisms suggests it could have a beneficial role in SOD1-mutant forms of ALS and has led to the investigation of CHIP as a potential ALS therapeutic. Dorfin is an E3 ubiquitin ligase that degrades mutant SOD1 and reduces mutant SOD1-mediated toxicity [107]. Because of its ability to degrade mutant SOD1, Dorfin has been of interest as a potential ALS therapeutic, but in vivo investigations have only shown modest benefit, and Dorfin has been found to have an extremely short half-life. To increase the efficacy of Dorfin as a therapeutic, engineered chimeric proteins have been generated, combining the substrate-binding domain of Dorfin and the U-box domain of CHIP [108]. This chimeric protein has been shown to have an increased half-life, to increase ubiquitination of mutant SOD1, to reduce aggregate formation, and to rescue neurons from mutant SOD1 toxicity. However, in vivo results have yet to be reported and will be required to support further development of this potential therapeutic approach.

Necroptosis is a pro-inflammatory form of programmed cell death which has been implicated in neuroinflammation and the pathogenesis of neurodegenerative diseases, most notably ALS, but also AD and PD [109-111]. Elevated levels of key necroptosis mediator proteins, receptor-interacting protein kinase 1 (RIPK1) and receptor-interacting protein kinase 3 (RIPK3), have been observed in ALS mouse models and patient samples [110, 111], raising the possibility of inhibition of these proteins as a potential treatment strategy for ALS. CHIP ubiquitinates and decreases levels of both RIPK1 and RIPK3 in vitro and in vivo, reducing necroptosis in response to various stressors [112, 113]. These findings indicate that CHIP could also provide therapeutic benefit for ALS treatment, and potentially other neurodegenerative diseases, through its ability to decrease RIPK1 and RIPK3 protein levels and reduce necroptosis.

Lafora Disease (LD)

LD is an autosomal recessive neurodegenerative disorder, leading to defects in the development of cerebral cortical neurons and is most commonly associated with epileptic seizures [114, 115]. LD is caused by loss of function mutations in the genes encoding laforin, a phosphatase, and malin, an E3 ubiquitin ligase that is responsible for regulating laforin protein levels [116]. LD is characterized by the presence of Lafora bodies, which are cytoplasmic polyglucosan aggregates, not protein aggregates. Lafora bodies exist within both neuronal and non-neuronal tissues. Laforin and malin are both critical proteins involved in glycogen production whose loss of function results in hyperphosphorylated, insoluble glycogen [117, 118]. Overexpression of mutant or wild-type malin can lead to polyglucosan aggregation, proteasomal dysfunction, and cell death. CHIP does not associate with mutant or aggregated malin, but does interact with wild-type malin through Hsp70 and increases malin protein levels in a concentration-dependent manner (Fig. ) [119]. The mechanism by which CHIP increases malin protein levels is currently unknown is independent of the U-box domain and thus could be due to the stabilisation of malin [119]. It is also unknown whether CHIP promotes malin aggregation or cellular dysfunction. While, the role of CHIP in LD is not well understood, current findings suggest CHIP could potentiate LD.

Huntington Disease (HD) and other Polyglutamine (PolyQ) Diseases

Polyglutamine (polyQ) diseases are inherited neurodegenerative disorders caused by the expansion of cytosine-adenine-guanine (CAG) in disease-causing genes. Currently, there are nine clinically described polyQ diseases: HD, spinal-bulbar muscular atrophy (SBMA), spinocerebellar ataxia 1 (SCA1), SCA2, SCA3, SCA6, SCA7, SCA17, and dentatorubral-pallidoluysian atrophy (DRPLA). Protein aggregation is present in all of these diseases, and none have therapies for prevention or cure [120, 121].

HD is a neurodegenerative disease characterized by the progressive movement, behavioural, and psychiatric symptoms. HD is caused by polyQ expansion in the gene encoding the huntingtin protein (HTT) [122]. CHIP has been shown to reduce polyQ-HTT protein aggregation and toxicity [123, 124], but did not promote the degradation of wild- type HTT [125]. Knockdown of CHIP in mice overexpressing neuronal polyQ-HTT resulted in reduced lifespan, accelerated neuronal dysfunction, behavioural abnormalities, and increased cerebellar granular cell HTT inclusions [123]. It has also been shown that high levels of Hsp70 binding protein 1 (HSPBP1) expression in neurons inhibits CHIP activity, reducing CHIP-mediated elimination of neuronal polyQ-HTT, while HSPBP1 knockdown has been shown to increase CHIP activity (Fig. ). Using an AAV CRISPR/Cas9 to target and knockdown HSPBP1 in a HD mouse model prevented accumulation and aggregation of HTT protein, and decreased loss of striatal presynaptic neurons [124].

SBMA is an X-linked genetic neurodegenerative disease caused by polyQ expansion within the gene encoding AR. SMBA is only present in males and is characterized by muscle weakness and progressive muscular atrophy [126]. CHIP has been shown to preferentially target polyQ-AR mutant proteins for proteasomal degradation compared to wild-type AR (Fig. ) [127]. Overexpression of CHIP in neuronal cells reduced mutant AR monomers, and overexpression in a SBMA mouse model improved movement defects and reduced polyQ-AR nuclear accumulation [127].

PolyQ expansion of the protein ataxin-3, a deubiquitinating enzyme (DUB), causes SCA3, also known as Machado-Joseph disease [128]. SCA3 is an autosomal dominant form of inherited cerebellar ataxia (see below). CHIP has been shown to preferentially bind and degrade mutant or polyQ-ataxin-3, but not wild-type ataxin-3, in neuronal cell lines. CHIP-mediated protection was improved by Hsc70 overexpression [123]. Using SCA3 transgenic mouse models, knockout and/or knockdown of CHIP was shown to worsen SCA3 motor defects, increase neuronal ataxin-3 microaggregation, and result in early death in a concentration-dependent manner [129]. PolyQ-ataxin-3 interacts with both monoubiquitinated and non-ubiquitinated CHIP, but the monoubiquitylation of CHIP by UBE2W has been shown to stabilize CHIP’s interaction with ataxin-3 and to enhance ataxin-3 ubiquitination (Fig. ) [53]. It was also shown that ataxin-3, as a DUB, can cleave monoubiquitin from CHIP, reducing its ubiquitylation activity. SCA3 transgenic mice have significantly decreased CHIP levels, while ataxin-3 knockout mice have no difference in CHIP levels compared to wild-type mice [53]. These results suggest a complex interplay between CHIP and polyQ-ataxin-3.

PolyQ expansion of ataxin-1 causes SCA1, another autosomal dominant form of inherited cerebellar ataxia [130]. CHIP associates with both wild-type and polyQ-ataxin-1 in a TPR domain-dependent manner [131]. Overexpression of CHIP in SCA1 mutant Drosophila results in decreased wild- type and polyQ-ataxin-1 protein levels (Fig. ) and significantly suppresses polyQ-ataxin-1 mediated toxicity within Drosophila eyes [132]. CHIP overexpression has also been shown to reduce wild-type ataxin-1 solubility, increasing aggregate formation [131].

CHIP has been shown to associate with and promote degradation of polyQ expanded but not wild-type forms of HTT, AR, ataxin-3, and CHIP is also able to promote degradation of polyQ expanded repeat short peptides [123, 133]. These results indicate that CHIP can specifically recognize polyQ expansions and warrants investigation into the role of CHIP in other polyQ diseases, including other SCAs (SCA 2, 6, 7, and 17) and DRPLA. CHIP appears to have a more complicated role in polyQ-ataxin-1 SCA1, as it may promote ataxin-1 aggregation, but has been shown to have a promising protective role in mouse models of HD, SBMA, and SCA3.

CHIP MUTATIONS IMPLICATED IN NEURODEGENERATIVE DISEASE

Inherited Cerebellar Ataxias

Inherited cerebellar ataxias are neurodegenerative diseases characterized by a progressive loss of cerebellar Purkinje neurons. Cerebellar neurodegeneration leads to a variety of symptoms, including decreased voluntary muscle control and incoordination, as well as clinical signs and symptoms that often reflect the anatomic site of damage. For example, damage to midline structures of the cerebellum often results in the development of gait ataxia, imbalance, truncal ataxia (inability to sit unsupported by their arms), vertigo, and nystagmus (rhythmic oscillations of the eyes). In contrast, damage to the cerebellar hemispheres is usually associated with limb ataxia, intention tremor (a tremor that increases in severity as the hand moves near its target), and ataxic dysarthria (difficulty articulating speech with a scanning quality) [134]. Increasingly, cerebellar lesions have also been suggested to have effects on cognitive and emotional processing [135].

Inherited cerebellar ataxias are phenotypically heterogenous and can demonstrate various modes of inheritance: autosomal dominant (such as the SCAs described above), autosomal recessive, X-linked, or even mitochondrial [136]. Autosomal dominant cerebellar ataxias are estimated to have a prevalence of around 1-5 per 100,000 people and autosomal recessive cerebellar ataxias have a prevalence of around 3 per 100,000 people [137]. A subset of autosomal recessive cerebellar ataxia includes Gordon Holmes syndrome (GHS), a rare autosomal recessive disease characterized by progressive cerebellar ataxia and hypogonadotropic hypogonadism.

Genetic testing of cerebellar ataxia patients of unknown cause has identified 132 patients with disease-associated STUB1 mutations. To date, 69 different disease-associated STUB1 mutations have been identified and confirmed. Of the identified disease-associated STUB1 mutations, 6 mutations map to intronic sequences, 21 mutations affect the TPR domain, 18 affect the coiled-coil domain, 23 map to the U-box domain, and 1 maps to the 3’ UTR (Fig. ). Thirty- eight of the identified mutations are associated with an autosomal recessive inheritance pattern of the disease, while 35 are associated with an autosomal dominant inheritance pattern. Our current understanding of autosomal recessive STUB1-associated disease and autosomal dominant STUB1-associated disease from these cases is described below.

Autosomal Recessive STUB1-associated Disease

Autosomal recessive spinocerebellar ataxia 16 (SCAR16) is a type of cerebellar ataxia caused by autosomal recessive mutations in the STUB1 gene. To date, 42 patients with autosomal recessive forms of STUB1-associated disease have been identified from 23 kindreds, 38 of which were diagnosed with cerebellar ataxia (20 male, 17 female, and 1 unspecified case) [55, 57, 138-150]. Four additional STUB1 patients were identified from a screen of patients with nervous system abnormalities [151]. In more than half of the cases, multiple siblings were affected. Out of all 42 patients identified, no generational family disease history was reported, as is typically observed in autosomal recessive conditions. Patients were ethnically diverse; of those kindreds for which ethnicity was specified, four families were of Chinese descent, two were Taiwanese, ten were European, three were Middle Eastern, and one was Sri Lankan. The age of onset for ataxia ranged from infancy to 57 years, with a mean of 22.3 years. All patients exhibited motor dysfunction and, for those who had neuroimaging, all demonstrated cerebellar atrophy Table (); however, there was large variation in the reported clinical phenotypes. For example, patients exhibited large differences in cognitive status ranging from no impairment to severe impairment/dementia. Additional neurological findings included epilepsy, eye movement abnormalities, chorea, dysarthria, urinary incontinence, and pyramidal symptoms, amongst others. Nerve conduction studies were abnormal in 3 out of 12 patients tested, hyperreflexia was observed in 15 of out 24 patients, and cognitive impairment was observed within 24 out of 35 patients. In addition, impaired sexual development/endocrine function was found in 5 out of 27 patients Table (). The diversity of symptoms exhibited by SCAR16 patients suggest involvement of multiple brain regions Table (). Indeed, one SCAR16 patient (Patient #25, Table ) was shown, with diffusion tensor imaging (DTI), to have degeneration throughout the brain, not solely limited to the cerebellum [144]. Additionally, neuropathological analysis of another SCAR16 patient (Patient #21, Table ) displayed abnormalities that were not restricted to the cerebellum. Within this patient, dramatic cerebellar cortical, Purkinje, and granular cell loss with reactive Bergmann gliosis was observed. Within the frontal cortex, swollen axonal processed we observed, and CHIP staining within neurons was reported to be diffuse and cytoplasmic with intense nuclear staining in neurons within deep layers of the frontal cortex. Ubiquitin and p62 positive neuronal intranuclear inclusions (NII) were also observed within frontal cortex neurons [142].

Twenty-two of the patients with autosomal recessive forms of STUB1-associated disease carried homozygous mutations and 20 carried compound heterozygous mutations (Fig. ). Patients with compound heterozygous mutations had either two different missense mutations or one missense and one nonsense or frameshift mutation. In total, 37 different autosomal recessive associated STUB1 mutations have been identified, 4 map to an intronic region, 9 affect the TPR domain, 10 affect the coiled coil domain, 13 affect the U-box domain, and 1 was found within the 3’ UTR region. In silico prediction tools have mostly supported the STUB1 variants associated with autosomal recessive disease to date as being pathogenic (Fig. ). MutationTaster [152] predicted all identified mutations as disease-causing while SIFT [153] and PolyPhen2 [154] predicted that all but 10 and 2 of the reported mutations would be disease causing, respectively. Most SCAR16 patients tested have been reported to have reduced levels of steady-state CHIP protein [139, 147, 155]. Multiple in vitro studies have been conducted to examine 13 different autosomal recessive associated mutations: E28K, N65S, A79T, A79D, L123V, N130I, K145Q, W147C, L165F, M211I, S236T, M240T, and T246M. Multiple mutants were shown to have reduced cellular expression following transfection [156]. These mutations have also been shown to reduce CHIP’s ability to interact with Hsc70 to varying degrees in vitro, with TPR domain mutants showing the most significant defects [156]. U-box mutations have been found to have increased binding capacity, but not affinity, for Hsp70 [157]. All tested mutations, other than N65S, showed decreased thermal stability and increased propensity for oligomerization [155, 156, 158]. U-box mutations were significantly more likely to form dimers than TPR or coiled- coil mutants, which had a higher propensity to form higher order oligomers [155, 156, 158]. In addition, U-box mutants, as well as the N65S TPR mutation, were all shown to retain their ability to recruit E2s but lacked E3 ligase activity [156], while other TPR and coiled-coil mutants retained E3 ligase activity [157]. CHIP mutants that had more E3 ligase activity and normal tertiary structure were associated with worsened disease outcomes than CHIP mutations lacking E3 ligase activity, indicating that the inhibition of mutant CHIP activity may be beneficial to STUB1-associated disease patients [157]. Interestingly, an earlier report demonstrated that within SCAR16 patients, U-box domain mutations were associated with a higher incidence of cognitive impairment than TPR or coiled-coil domain mutations [157]. Since that report, additional SCAR16 cases have been published and the higher incidence of cognitive impairment with U-box domain mutations is still observed: 62.5% (10/16) of TPR mutated alleles, 59% (13/22) of coiled-coiled mutated alleles, and 92% (22/24) of U-box mutant alleles (P = 0.024, Fisher’s exact test).

To date, the most intensively studied CHIP mutation associated with SCAR16 is T246M. This missense mutation occurs in CHIP’s U-box domain and has been identified in two homozygous patients with GHS [57]. T246M-CHIP has been shown in vitro to possess no E3 ubiquitin ligase activity and to have an increased propensity for oligomerization, but it maintains normal chaperone interaction. Rats and mice expressing homozygous T246M-CHIP exhibited Purkinje cell degeneration, decreased survival, cognitive defects, gonadal atrophy, and reduced body weight and brain mass [155]. Mutant mice also exhibited age-dependent progressive motor dysfunction. In contrast, Stub1 null mice display severe motor impairment at the time of weaning with little progression over time. Thus, the progression observed with T246M-CHIP mice more accurately models symptoms observed in STUB1-associated disease patients and also allows for the study of earlier or pre-clinical symptoms. Tissue lysates from the T246M-CHIP mouse cerebellum, whole brain, and testes, as well as embryonic fibroblasts, all displayed decreased CHIP protein levels compared to controls. Interestingly, proteomics analysis of T246M-CHIP rat cerebella and whole brain, compared to controls, showed increased levels of tau and decreased levels of both α-syn and PINK1 [155]. While whole-brain lysates from Stub1 null mice also showed increased levels of tau, there was no difference in α-syn protein levels [87] and increased PINK1 levels [56], compared to controls. The differences between CHIP mutant and null animal phenotypes support the observation that STUB1-associated disease is not solely the result of the loss of E3 ligase function. Further study is still required to confirm that all of the STUB1-disease associated mutations are indeed disease-causing and not just variants of unknown significance, but the rodent model recapitulation of STUB1-associated disease symptoms, as well as in vitro studies and in silico predictions, provide convincing evidence that CHIP mutations can cause multisystemic symptoms including neurodegeneration and hypogonadism.

The specific molecular pathways affected by CHIP mutations in SCAR16 remain an area open for investigation. SCAR16 patients have been shown to exhibit cyclic AMP (cAMP) signaling dysregulation [159]. cAMP is an ATP derivative involved in intracellular signaling [160, 161]. cAMP signaling involves activation of various protein kinases, including protein kinase A (PKA), and can function to regulate ion channels [162, 163]. Dysregulation of cAMP-PKA signaling has been observed in many other neurodegenerative diseases, including PD [164], AD [165], HD [166], and SCA1 [167-169]. CHIP has been shown to attenuate cAMP-PKA signaling through Hsp70-mediated interaction with the catalytic subunit PKAc, leading to its proteasomal degradation [159, 170]. Further research is necessary to determine how mutant CHIP may lead to cAMP-PKA signaling dysregulation observed within SCAR16 patients. Understanding how this pathway could contribute to disease pathogenesis may provide insights into the development of treatments for SCAR16 as well as into the role of CHIP in other neurodegenerative diseases.

Autosomal Dominant STUB1-associated Disease

Spinocerebellar ataxia 48 (SCA48) is an autosomal dominant form of cerebellar ataxia caused by a heterozygous STUB1 mutation. At the time of this review, 90 patients (26 male, 55 female, and 9 unspecified) from 43 kindreds with autosomal dominant STUB-associated mutations have been described in the literature [150, 171-176]. Interestingly, when sex is specified, 67.9% of identified SCA48 patients are female, while a difference in sex-based prevalence of SCAR16 is not apparent. Sex-based differences in disease prevalence, onset, and outcome have been observed in a number of neurodegenerative diseases [177, 178]. It has been reported that cerebellar CHIP protein levels are lower in females than males, which may play a role in the observed sex-based difference in SCA48 prevalence [150].

Out of the 43 reported kindreds, 34 had a positive family history of disease. An absence of family history may be explained by spontaneous mutations, incomplete penetrance, or later onset of disease with some family members with a STUB1 mutation not living long enough to display symptoms. For cases in which ethnicity was indicated, all patients were of European descent. The age of symptom onset among SCA48 patients varied from age 5 to 74 years; the mean age of onset for these patients was 44.6 years, which was older than the average age at onset of SCAR16 patients. Symptom onset typically begins during mid-adulthood and includes cognitive impairment, including cerebellar cognitive affective syndrome (CCAS), and/or motor disturbances. Patients presenting with primarily cognitive impairment may not display motor features for years after the age of initial disease onset, but most patients eventually develop both cognitive and motor impairments. All but five patients were described as having ataxia. Two non-ataxic patients had no symptoms (Patient #41 & 136, Table ). Two other non-ataxic patients (Patient #36 & 38, Table ) presented with CCAS without motor dysfunction. As mentioned previously, cognitive impairment precedes movement abnormalities for many SCA48 patients, so these patients may go on to develop ataxia. The last non-ataxia patient (Patient #81, Table ) presented with parkinsonism without ataxia. Of the patients reported, 15 out of 22 displayed hyperreflexia, 63 out of 86 presented with cognitive impairment, and 2 out of 11 displayed sexual development disturbances. Additional reported symptoms included but were not limited to anxiety, generalized tonic-clonic seizures, chorea, parkinsonism, dystonia, chorea, and urinary incontinence, among others. One patient had been diagnosed with autism spectrum disorder (Patient #71, Table ). Among patient who had neuroimaging, all showed significant cerebellar atrophy, including an asymptomatic patient. Some patients also exhibited hypoperfusion to various brain regions including the cerebellum as well as frontal, parietal, and temporal lobes. One patient displayed severe frontal lobe atrophy and some displayed additional cortical atrophy. Additionally, T2-weighted imaging (T2WI) revealed bilateral dentate nuclei hyperintensity in 9/10 SCA48 patients [179]. Dentate nuclei hyperintensity by T2WI is not typical of other SCAs [180], and this hyperintensity combined with atrophy of the postero-lateral portions of the cerebellar hemisphere has been termed the “crab sign” [179].

Thirty-two different autosomal dominant mutations were observed in these patients and were reported to segregate with the disease: 12 affected the TPR domain, 8 affected the coiled-coil domain, and 10 mapped to the U-box, and 2 were intronic. Twenty mutations were missense, 3 were nonsense, 8 caused frameshift mutations, and 1 was an in-frame deletion. The most common mutation reported to date is a dominant heterozygous frameshift truncation mutations located within the U-box domain (ENST00000219548: c.823_824delCT, ENSP00000219548: p.L275Dfs*16). This mutation has been identified within 3 unrelated kindreds of 3 different ethnicities. In silico prediction tools for identifying disease causing mutations were preformed using all described autosomal dominant STUB1 mutations (Fig. ). MutationTaster predicted all identified mutations would be disease-causing while SIFT and PolyPhen2 predicted that all but 2 of the reported mutations would be disease causing. To date, no further in vitro or in vivo analysis has been performed to further investigate any of the identified autosomal dominant STUB1 mutations.

Neuropathologic examination of 5 SCA48 patients with missense mutations located within the TPR domain exhibited significant loss of cerebellar Purkinje cells with increased astrogliosis and Bergmann gliosis [150, 175]. One patient displayed mild frontal lobe atrophy. Chen et al. found ubiquitinated inclusions within the cerebellum which were negative for p62, TDP-43, and α-syn [175]. Neurofibrillary tangles were present but Braak staging was low and consistent with the age of the patient. No signs of basal ganglia abnormalities were observed in their 4 cases, despite one patient presenting with mild parkinsonism. Roux et al., 2020 reported one SCA48 case with mild neuronal loss within the SNpc [150]. This case demonstrated only infrequent pretangles and neuropil threads with tau immunostaining in the hippocampus, as well as rare p62 nuclear inclusions in the frontal cortex and hippocampus. No inclusions were seen with immunostaining for ubiquitin, TDP-43, α-syn, polyQ, or prion protein (PrP). Confocal microscopy of healthy controls showed CHIP expression to be prominent within Purkinje cells and localised within the cell bodies and proximal dendrites, with little staining in distal dendritic compartments. In contrast, CHIP localised throughout somatodendritic and distal dendritic regions of Purkinje cells in the brains of SCA48 patients. This may suggest that the mislocalisation of CHIP, not the absolute level of CHIP, is relevant to disease pathogenesis. Purkinje cells from SCA3 and SCA5 patients displayed CHIP localisation consistent with healthy controls, while SCA7 patients showed CHIP mislocalisation consistent with SCA48 patients [175], indicating that the CHIP mislocalisation is not solely due to CHIP mutations and may be attributed to other aspects of Purkinje neurodegeneration.

Neuropathological examinations of 3 related SCA48 patients with a U-box frameshift mutation displayed an almost complete loss of cerebellar Purkinje cells in all cases and Bergmann gliosis. Braak staging was low and consistent with the age of the patient [176]. Significant neuronal loss was also observed within the mesencephalon, medulla oblongata, and subthalamic nucleus. NIIs positive for ubiquitin and p62 were present within the cerebellum, SNpc, and deep pontine nucleus. p62 positive inclusions were also observed within the hippocampus. CHIP staining within neurons was observed to be diffuse throughout but not different from controls. Within one patient with atypical parkinsonism (Patient #81, Table ), tau pathology was observed within many brain regions, including the temporal cortex, hippocampus, SNpc, locus coeruleus, cerebellar dentate nucleus, and spinal motor neurons. None of the patients displayed α-syn or TDP-43 positive inclusions. Some faint diffuse nuclear polyglutamine immunoreactivity was observed within the cerebellum of the 2 brains tested [176].

How CHIP mutations cause disease and the multitude of complex variable symptoms observed within SCAR16 and SCA48 patients still require further elucidation. Variability between patients, even those carrying the same mutation, indicates that other genetic or environmental factors may play a role in the pathophysiology of the disease. Sex-based differences in the prevalence of SCA48 but not SCAR16 may be due to differences in cerebellar CHIP expression between males and females but could also point to additional mechanisms underlying disease pathology, such as microglia function, which has been implicated in sex-based differences in the prevalence of many neurodegenerative diseases [177, 178]. In vitro and in vivo analysis of multiple SCAR16 mutations have reported that mutations result in decreased CHIP protein expression [139, 147, 155, 156]. However, differences between the T246M mouse model and Stub1 null mice suggest that the effects of CHIP mutations are the result of more than solely loss of function. Additionally, post- mortem brain tissue from SCA48 patients were absent of tau pathology [175], aside from one patient with atypical parkinsonism, while Stub1 knockout mice and C. elegans showed substantial levels of tau pathology [93], indicating that mutant forms of CHIP may still be able to eliminate tau, further suggesting the cause of the disease is more than merely CHIP loss of function. Furthermore, the association between cognitive impairment and U-box domain mutations suggests that there may be different mechanisms contributing to specific STUB1-associated disease symptoms. Interestingly, other causative mutations identified in patients with GHS affect Ring Finger Protein 216 (RNF216), which is an E3 ubiquitin ligase, Patatin Like Phospholipase Domain Containing 6 (PNPLA6), which is a phospholipase [181, 182], and OTU Deubiquitinase 4 (OTUD4), which is a DUB [135], all of which play important roles in proteostasis, like CHIP, and may point to a converging molecular pathway. Additionally, multiple genes associated with autosomal dominant cerebellar ataxias, including ataxin-1 and ataxin-3, are known interactors of CHIP (described above), which may also point to the converging pathway among autosomal dominant cerebellar ataxias.

THERAPEUTIC POTENTIAL OF CHIP

As discussed above, CHIP has the potential to promote the degradation of proteins associated with multiple neurodegenerative disorders, including PD, AD, ALS, HD, SBMA, SCA1 and SCA3. Thus, increasing CHIP levels or activity could present a promising treatment strategy for these diseases. To date, a limited number of compounds have been found to increase CHIP protein levels in vitro, which include trehalose, anisomycin, and YL-109. Trehalose, a disaccharide of glucose and known autophagy enhancer, has been implicated as a potential STUB1-associated disease therapeutic requiring further investigation. Treatment of SCAR16 patient fibroblasts with trehalose was shown to increase CHIP and Hsp70 protein levels, increase autophagy [183], and reduce oxidative stress [184]. It has also been shown that anisomycin treatment increases CHIP protein levels and protects cells against OGD-induced necroptosis in in vitro models of cerebellar ischemia [112]. The agent 2-(4-Hydroxy-3-methoxyphenyl)-benzothiazole (YL-109) has been shown to suppress triple-negative breast cancer cell growth in vitro and in vivo [185]. YL-109 was also shown to increase CHIP transcription through aryl hydrocarbon receptor (AhR) recruitment upstream of the STUB1 gene. YL-109 treatment of cells with knockdown of AhR or CHIP rescued proliferation [185]. While evidence to date suggests that increasing CHIP expression or activity could be beneficial for various neurodegenerative diseases, there may be undesirable effects that need to be considered. For instance, as discussed above, CHIP can promote degradation of polyQ-ataxin-1, but it can also promote degradation and decreased solubility of wild-type ataxin-1. Additionally, while CHIP may reduce α-syn aggregation, reduce Pael-R levels, and protect against LRRK2 mutations in PD models, one of the substrates of CHIP may be tyrosine hydroxylase [63], an enzyme critical in the production of dopamine, which could make non-specifically increasing CHIP expression or activity unsuitable as a therapy for PD.

For each neurodegenerative disease, taking different approaches to CHIP modulation may be necessary to optimise efficacy and limit side effects. One interesting avenue warranting investigation is the use of protein-protein interaction (PPI) modulators. Certain small molecules have been shown to stabilise specific PPIs [186], including E3 ligase-substrate interactions [187]. PPI modulators, which could stabilise the interaction between CHIP and specific aggregated, overexpressed, or mutated disease-associated proteins, could be employed to enhance the removal of specific proteins without broadly increasing CHIP-mediated protein degradation. Influencing PTMs of CHIP could also provide another interesting therapeutic treatment avenue that may reduce unwanted effects. For instance, increasing CHIP monoubiquitylation via UBE2W increases degradation of polyQ-ataxin-3, and could potentially be useful for SCA3 treatment. To date, PTMs of CHIP are not well understood and warrant further research, as they may provide insight into improving the therapeutic benefit or target specificity of CHIP. High-throughput mutagenesis of E3 ubiquitin ligases has been used to determine activity enhancing mutations, including mutations that enhance U-box domain interaction with specific E2s, allowing for the enhancement of degradation of certain proteins and not others [188]. High-throughput mutagenesis could also be used to identify mutations that could enhance specific interactions between CHIP and disease-associated proteins.

The generation of chimeric proteins utilising either the TPR or U-box domain CHIP could present another method for developing the therapeutic potential of CHIP, potentially increasing disease-associated protein degradation while decreasing undesirable interactions. As described above, CHIP is able to promote proteasomal degradation of mutant SOD1 through an indirect interaction mediated by Hsp70 [104]. To improve on the therapeutic potential, a Dorfin-CHIP chimeric protein [108] described above, was generated utilised the E3 activity of the U-box domain of CHIP linked to the substrate-binding domain of Dorfin, which is able to interact directly with mutant SOD1. This Dorfin-CHIP chimeric protein was shown to increase mutant SOD1 degradation. The TPR domain of CHIP is able to interact with many different substrates, including polyQ expanded proteins, and may also be useful for the generation of chimeric protein therapeutics. The utility of a chimeric protein containing either the TPR or U-box domain would likely depend on the specific disease and protein being targeted.

For the treatment of STUB1-associated diseases, it is currently unknown how CHIP modulation may impact disease symptoms and progression. Increasing expression of CHIP via trehalose treatment in SCAR16 patient fibroblasts showed benefits in vitro [184], but multiple lines of evidence suggest that STUB1-associated disease may be due to more than just loss of CHIP function. Several SCAR16 mutations have been shown to have increased propensity to form higher-order oligomers in vitro, so increasing CHIP expression may exacerbate disease in these patients. Further investigation is required to determine whether upregulating or downregulating CHIP within STUB1-associated diseases may provide therapeutic benefit for patients. As with other genetic diseases, major therapeutic breakthroughs for STUB1-associated diseases will likely require precision or personalised medicine in which the intervention targets the individual patient’s CHIP defect. And, as with all neurodegenerative disorders, the timing of the intervention during the disease course will be an important consideration.

CONCLUSION

The co-chaperone CHIP plays a major role in proteostasis and hence is emerging as an important regulator of neurodegeneration. Direct evidence for its importance in neurodegenerative disease comes from the identification of disease-associated STUB1 mutations in individuals with both autosomal recessive and autosomal dominant forms of cerebellar ataxia. Multiple lines of evidence also demonstrate that CHIP can reduce protein aggregation in neurodegenerative diseases characterised as proteinopathies, including PD, AD, HD, SBMA, SCA1, SCA3 and ALS, and could be a promising therapeutic target for the treatment of these diseases. Future research will need to elucidate how CHIP levels, CHIP functions, and CHIP interactions and networks are affected in these diseases. Understanding the interplay between CHIP and other contributors to disease pathogenesis, such as genetic and environmental factors, will be needed to explain how CHIP dysfunction may manifest as different neurodegenerative diseases in different individuals. This will also be important in informing disease-specific treatment strategies. Such insights will assist in deciphering how CHIP can most effectively be modified in vivo to enhance its specific degradation of disease-associated proteins.

Table 1

Summary of all reported individuals with STUB1-associated disease.

Kindred	Patient	AAO (y)/ Sex	Ethnicity	Clinical diagnosis	STUB1 mutation	CHIP mutation	Ataxia	Cognitive status	Eye movement abnormalities	Upper motor neuron features	Other movement disorders	Sexual development	Additional clinical features	Neuroimaging	Nerve conduction studies
A [55]	1	17/F	Chinese	SCAR16	c.493C>Tc.493C>T	L165FL165F	Severe truncal/limb ataxia, dysarthria	Severe impairment	Ophthalmoplegia	Normal tendon reflexes	No abnormalities reported	Not reported	No abnormalities reported	Not reported	Not reported
	2	17/F		SCAR16	c.493C>Tc.493C>T	L165FL165F	Moderate truncal/limb ataxia, dysarthria	Normal	Nystagmus	Increased tendon reflexes	No abnormalities reported	Not reported	No abnormalities reported	Severe cerebellar atrophy	Slowed NCV (motor and sensory)
	3	14/F		SCAR16	c.493C>Tc.493C>T	L165FL165F	Moderate truncal/limb ataxia, dysarthria	Moderate impairment	Ophthalmoplegia	Increased tendon reflexes	No abnormalities reported	Not reported	No abnormalities reported	Severe cerebellar atrophy	Slowed NCV (motor and sensory)
	4	19/M		SCAR16	c.493C>Tc.493C>T	L165FL165F	Moderate truncal/limb ataxia, dysarthria	Normal	Nystagmus	Increased tendon reflexes	No abnormalities reported	Not reported	No abnormalities reported	Severe cerebellar atrophy	Slowed NCV (motor and sensory)
B [55]	5	20/M	Chinese	SCAR16	c.389A>Tc.441G>T	N130IW147C	Subtle/mild truncal/limb ataxia, dysarthria	Normal	Normal	Normal tendon reflexes	No abnormalities reported	Not reported	No abnormalities reported	Severe cerebellar atrophy	Not reported
C[55]	6	16/F	Chinese	SCAR16	c.621C>Gc.707G>C	Y207*S236T	Moderate /severe truncal/limb ataxia, dysarthria	Normal	Normal	Increased tendon reflexes	No abnormalities reported	Not reported	No abnormalities reported	Severe cerebellar atrophy	Not reported
D [57]	7	19/F	Chinese	SCAR16	c.737C>Tc.737C>T	T246MT246M	Severe ataxia, dysarthria	Cognitive impairment	Nystagmus	Normal tendon reflexes	No abnormalities reported	Hypogonadism	No abnormalities reported	Cerebellar atrophy	Normal
	8	17/F			c.737C>Tc.737C>T	T246MT246M	Severe ataxia	Cognitive impairment	No abnormalities reported	Pyramidal features, increased tendon reflexes	Tremor	Hypogonadism	No abnormalities reported	Cerebellar atrophy	Not reported
E [138]	9	2/M	German	SCAR16	c.367C>Gc.367C>G	L123VL123V	Ataxia	Normal	No abnormalities reported	Pyramidal features	No abnormalities reported	Normal	UI	Cerebellar atrophy	Normal
F [138]	10	16/F	Turkish	SCAR16	c.719T>Cc.719T>C	M240TM240T	Ataxia	Cognitive impairment	No abnormalities reported	Pyramidal features	No abnormalities reported	Normal	UI	Cerebellar atrophy	Normal
G [138]	11	29/M	Saudi Arabian	SCAR16	c.235G>Ac.236C>A	A79TA79D	Ataxia	Normal	No abnormalities reported	Pyramidal features, increased tendon reflexes	No abnormalities reported	Normal	No abnormalities reported	Cerebellar atrophy	Not reported
	12	49/M		SCAR16	c.235G>Ac.236C>A	A79TA79D	Ataxia	Normal	No abnormalities reported	Pyramidal features, increased tendon reflexes	No abnormalities reported	Normal	UI	Cerebellar atrophy	Not reported
H [139]	13	2/M	Saudi Arabian	SCAR16	c.194A>Gc.194A>G	N65SN65S	Ataxia, dysarthria	Cognitive impairment, delayed development, dyspraxia	Nystagmus	No abnormalities reported	No abnormalities reported	Normal	Aged appearance, alopecia, T1D	Cerebellar hypoplasia	Not reported
	14	0/M		SCAR16	c.194A>Gc.194A>G	N65SN65S	Ataxia, dysarthria	Cognitive impairment	Chronic iridocyclitis with secondary glaucoma, oculomotor dyspraxia with saccadic pursuit	No abnormalities reported	Tremor	Normal	Dysphagia, aged appearance, ulcerative colitis	Severe cerebellar atrophy	Not reported
	15	0.67/F		SCAR16	c.194A>Gc.194A>G	N65SN65S	Ataxia	Cognitive impairment, delayed development, dyspraxia	Nystagmus	No abnormalities reported	No abnormalities reported	Normal	Epilepsy (until 2 yrs), aged appearance, facial dysmorphism	Cerebellar hypoplasia	Not reported
I [139]	16	33/F	Sri Lankan	SCAR16	c.82G>Ac.430A>T	E28KK144*	Gait ataxia, dysarthria	No abnormalities reported	Not reported	No abnormalities reported	No abnormalities reported	Oligomenorrhea, secondary infertility	Dysphagia	Cerebellar hypoplasia	Not reported
J [140]	17	23/M	Belgian	SCAR16	c.433A>Cc.687-690delCTAC	K145QI227P	Gait/limb ataxia, dysarthria	Cognitive impairment	No abnormalities reported	Increased tendon reflexes	No abnormalities reported	Normal	Dysphagia	Cerebellar atrophy	Normal
	18	25/M			c.433A>Cc.687-690delCTAC	K145QI227P	Gait/limb ataxia, dysarthria	Cognitive impairment	No abnormalities reported	Increased tendon reflexes	No abnormalities reported	Normal	Dysphagia	Cerebellar atrophy	Not reported
K [141]	19	15/M	Not reported	SCAR16	c.612+1G>Cc.823C>G	IntronicL275V	Truncal/limb ataxia, dysarthria	Cognitive impairment, attention failures, executive dysfunction	Nystagmus, fractionated pursuit, hypometric saccades	Pyramidal features, increased tendon reflexes	Myoclonus	Normal	Dysphagia	Cerebellar atrophy	Normal
L [142]	20	20/F	Spanish	SCAR16	c.633G>Ac.712G>T	M211IE238*	Generalized ataxia, dysarthria	Severe impairment/dementia	Saccadic ocular pursuit	Pyramidal features, increased tendon reflexes	Myoclonus	Normal	No abnormalities reported	Cerebellar atrophy	Normal
	21	22/M			c.633G>Ac.712G>T	M211IE238*	Generalized ataxia, dysarthria	Severe impairment/dementia	No abnormalities reported	Spasticity, increased tendon reflexes	No abnormalities reported	Normal	No abnormalities reported	Cerebellar atrophy	Normal
M [143]	22	15/F	Not reported	Unconfirmed	Unconfirmed	Unconfirmed	Ataxia, dysarthria	Cognitive impairment	Oculomotor dysfunction	Spastic equinovarus deformity	Myoclonic and tremulous movement	Not reported	No abnormalities reported	Not reported	Not reported
	23	19/M		SCAR16	c.724G>Ac.724G>A	E242KE242K	Truncal/limb ataxia, dysarthria	Cognitive impairment	Nystagmus	No abnormalities reported	Choreoathetosis, dystonia, myoclonus	Not reported	No abnormalities reported	Severe cerebellar atrophy	Not reported
	24	24/M			c.724G>Ac.724G>A	E242KE242K	Ataxia, dysarthria	Cognitive impairment	Vertical ocular flutter	No abnormalities reported	Dystonia, myoclonic and tremulous movement	Not reported	No abnormalities reported	Severe cerebellar atrophy	Not reported
N [144]	25	12/M	German	SCAR16	c.355C>Tc.880A>T	R119*I294F	Ataxia	Severe impairment/ dementia	Nystagmus	Spastic tetraparesis, increased tendon reflexes, Babinski sign	Dystonia	Hypogonadism	Generalized tonic-clonic seizures, hypomimia, UI	Cerebellar atrophy, atrophy of mesencephalon and parieto-occipital cortex	Normal
O [144]	26	12/M	Belgian	SCAR16	c.433A>Cc.728C>T	K145QP243L	Ataxia	Severe impairment/ dementia, mutism	No abnormalities reported	Spastic tetraparesis, increased tendon reflexes, Babinski sign	Dystonia	Normal	Generalized tonic-clonic seizures, UI	Cerebellar atrophy	Normal
	27	20/F			c.433A>Cc.728C>T	K145QP243L	Ataxia	Severe impairment/ dementia, mutism	Nystagmus	Spastic tetraparesis, increased tendon reflexes, Babinski sign	No abnormalities reported	Normal	Generalized tonic-clonic seizures, UI	Cerebellar atrophy	Normal
P [151]	28	--	Not reporter	Nervous system abnormality	c.268G>Tc.268G>T	D90YD90Y	Not reported -WES study
Q [151]	29	--	Not reported	Nervous system abnormality	c.358+2T>Gc.358+2T>G	IntronicIntronic	Not reported -WES study
R [151]	30	--	Not reporter	Nervous system abnormality	c.670-11_670-10delCTc.670-11_670-10delCT	IntronicIntronic	Not reported -WES study
S [151]	31	--	Not reported	Nervous system abnormality	c.844C>Gc.844C>G	P282AP282A	Not reported -WES study
T [145]	32	--	European	SCAR16	c.433A>Cc.502C>T	K145QL168F	Ataxia	Not reported -WES study
U [171]	33	53/F	Spanish	SCA48	c.823_824delCT	L275Dfs*16	Ataxia, dysarthria	CCAS	No abnormalities reported	No abnormalities reported	No abnormalities reported	Not reported	Anxiety, dysphagia, cachexia, UI	Cerebellar atrophy	Not reported
	34	52/F			c.823_824delCT	L275Dfs*16	Ataxia, dysarthria	CCAS	No abnormalities reported	No abnormalities reported	No abnormalities reported	Not reported	Anxiety, dysphagia, cachexia, UI	Cerebellar atrophy	Not reported
	35	41/F			c.823_824delCT	L275Dfs*16	Ataxia, dysarthria	CCAS, aphasia	No abnormalities reported	No abnormalities reported	No abnormalities reported	Not reported	Anxiety, dysphagia, stroke	Cerebellar atrophy, cerebellar and left cerebral hemisphere hypoperfusion	Not reported
	36	53/M			c.823_824delCT	L275Dfs*16	No ataxia	CCAS	No abnormalities reported	No abnormalities reported	No abnormalities reported	Not reported	No abnormalities reported	Cerebellar atrophy, preserved cerebral perfusion	Not reported
	37	42/F			c.823_824delCT	L275Dfs*16	Ataxia, dysarthria	CCAS	No abnormalities reported	No abnormalities reported	No abnormalities reported	Not reported	Anxiety, organic personality disorder, dysphagia, UI	Cerebellar atrophy, atrophy of frontal lobe, cerebellar hypoperfusion	Not reported
	38	56/M			c.823_824delCT	L275Dfs*16	No ataxia	CCAS	No abnormalities reported	No abnormalities reported	No abnormalities reported	Not reported	No abnormalities reported	Cerebellar atrophy, preserved cerebral perfusion	Not reported
	39	F			c.823_824delCT	L275Dfs*16	Ataxia, dysarthria	CCAS	No abnormalities reported	No abnormalities reported	No abnormalities reported	Not reported	Anxiety	Cerebellar atrophy	Not reported
	40	42/F			c.823_824delCT	L275Dfs*16	Ataxia, dysarthria	CCAS	No abnormalities reported	No abnormalities reported	No abnormalities reported	Not reported	Organic personality disorder, dysphagia	Cerebellar atrophy	Not reported
	41	33/M			c.823_824delCT	L275Dfs*16	No ataxia	Normal	No abnormalities reported	No abnormalities reported	No abnormalities reported	Not reported	No abnormalities reported	Cerebellar atrophy	Not reported
V [146]	42	54/F	Not reported	SCAR16	c.103C>Ac.678_679del	R35SI227Pfs*11	Ataxia, dysarthria	Severe impairment	Nystagmus, abnormal ocular pursuit	No abnormalities reported	Chorea	Normal blood hormone levels	No abnormalities reported	Cerebellar atrophy, hypoperfusion of fronto-parietal cortex	Not reported
W [147]	43	57/M	Not reported	SCAR16	c.*204T>Cc.*204T>C	3’UTR3’UTR	Ataxia, dysarthria	Normal	Normal	Pyramidal features, increased tendon reflexes	No abnormalities reported	Normal	No abnormalities reported	Cerebellar atrophy	Not reported
	44	37/F			c.*204T>Cc.*204T>C	3’UTR3’UTR	Ataxia, dysarthria	Normal	Normal	Pyramidal features, increased tendon reflexes	No abnormalities reported	Normal	No abnormalities reported	Cerebellar atrophy	Not reported
	45	31/F			c.*204T>Cc.*204T>C	3’UTR3’UTR	Ataxia, dysarthria	Cognitive impairment	Nystagmus	Pyramidal features, increased tendon reflexes	No abnormalities reported	Normal	Depression	Cerebellar atrophy	Not reported
X [172]	46	50/F	Italian	SCA48	c.97G>A	G33S	Ataxia, dysarthria	Mild cognitive impairment, memory impairment, executive dysfunction	No abnormalities reported	No abnormalities reported	Chorea, parkinsonism	Not reported	Psychiatric symptoms, UI	Not reported	Not reported
	47	12/F			c.97G>A	G33S	Ataxia, dysarthria	Moderate cognitive impairment, memory impairment, executive dysfunction	No abnormalities reported	Increased tendon reflexes	Parkinsonism	Not reported	Generalized tonic-clonic seizures, dysphagia, psychiatric symptoms, UI	Cerebellar atrophy	Not reported
	48	5/M			c.97G>A	G33S	Ataxia, dysarthria	Moderate cognitive impairment, memory impairment, executive dysfunction	No abnormalities reported	Increased tendon reflexes	Chorea, dystonia, parkinsonism	Not reported	Generalized tonic-clonic seizures, dysphagia, psychiatric symptoms	Cerebellar atrophy	Not reported
	49	28/F			c.97G>A	G33S	Ataxia, dysarthria	Moderate cognitive impairment, memory impairment, executive dysfunction	No abnormalities reported	Increased tendon reflexes	Chorea, dystonia, parkinsonism, tremor	Not reported	Seizures, dysphagia, psychiatric symptoms	Cerebellar atrophy	Not reported
	50	45/M			c.97G>A	G33S	Ataxia, dysarthria	Moderate cognitive impairment, memory impairment, executive dysfunction	No abnormalities reported	Increased tendon reflexes	Dystonia, parkinsonism, tremor	Not reported	Generalized tonic-clonic seizures, dysphagia, psychiatric symptoms	Cerebellar atrophy	Normal
	51	44/M			c.97G>A	G33S	Ataxia, dysarthria	Mild cognitive impairment, memory impairment, executive dysfunction	No abnormalities reported	Increased tendon reflexes	No abnormalities reported	Not reported	Seizures, dysphagia, hearing loss, psychiatric symptoms, UI	Cerebellar atrophy	Normal
Y [172]	52	55/M	Italian	SCA48	c.682C>T	P228S	Ataxia, dysarthria	Moderate cognitive impairment	No abnormalities reported	No abnormalities reported	Blepharospasm, chorea, tongue tremor	Not reported	Dysphagia, psychiatric symptoms	Cerebellar atrophy	Normal
	53	35/M			c.682C>T	P228S	Ataxia, dysarthria	Mild cognitive impairment	No abnormalities reported	Increased tendon reflexes	Tremor	Not reported	Dysphagia, hearing loss	Cerebellar atrophy	Not reported
Z [173]	54	31/M	Italian	SCA48	c.689_692delACCT	Y230Cfs*9	Gait ataxia, dysarthria	Normal	Nystagmus, broken smooth pursuit	Increased tendon reflexes	Dystonia	Low testosterone	No abnormalities reported	Cerebellar atrophy	Normal
	55	48/F			c.689_692delACCT	Y230Cfs*9	Gait ataxia, dysarthria	Moderate cognitive impairment	Broken smooth pursuit	Normal	Chorea, dystonia, parkinsonism	Normal	Depression, UI	Cerebellar atrophy	Not reported
AA [173]	56	42/F	Italian	SCA48	c.818_819dupGC	P274Afs*3	Gait ataxia, dysarthria	Moderate cognitive impairment	Normal	Normal	Anxiety, chorea, dystonia	Normal	Dysphagia, diabetes mellitus, UI	Cerebellar atrophy	Not reported
	57	34/F			c.818_819dupGC	P274Afs*3	Gait ataxia, dysarthria	Moderate cognitive impairment	Broken smooth pursuit	Increased tendon reflexes	Chorea	Normal	No abnormalities reported	Cerebellar atrophy	Not reported
AB [173]	58	56/M	Italian	SCA48	c.791_792delTG	V264Gfs*4	Gait ataxia, dysarthria	Moderate cognitive impairment	Broken smooth pursuit, impersistence	Normal	Parkinsonism	Normal	Dysphagia	Cerebellar atrophy	Normal
	59	40/F			c.791_792delTG	V264Gfs*4	Gait ataxia, dysarthria	Moderate cognitive impairment	Broken smooth pursuit, slow saccades	Normal	Chorea	Normal	Frontal meningioma	Not reported	Not reported
AC [173]	60	37/F	Italian	SCA48	c.199G>A	A67T	Gait ataxia, dysarthria	Normal	Impersistence	Increased tendon reflexes	Chorea, dystonia	Normal	Thyroid cancer, UI	Cerebellar atrophy	Normal
AD [173]	61	50/F	Italian	SCA48	c.673C>T	R225*	Gait ataxia, dysarthria	Normal	Impersistence	Normal	No abnormalities reported	Normal	Dysphagia	Cerebellar atrophy	Normal
AE [173]	62	46/F	Italian	SCA48	c.721C>T	R241W	Mild ataxia, dysarthria	Cognitive impairment	Abnormalities	Normal	No abnormalities reported	Hypogonadism	Hashimoto’s thyroiditis	Cerebellar atrophy	Normal
AF [173]	63	30/F	Italian	SCA48	c.823_824delCT	L275Dfs*16	Gait ataxia, dysarthria	Normal	Nystagmus	Increased tendon reflexes	Chorea, parkinsonism	Normal	Anxiety, depression, dysphagia, UI	Cerebellar atrophy	Not reported
AG [173]	64	43/M	Italian	SCA48	c.170C>T	P57L	Gait ataxia, dysarthria	No abnormalities reported	Broken smooth pursuit, hypometric saccades	Increased tendon reflexes	No abnormalities reported	Normal	Self-harm, dysphagia, tongue atrophy	Cerebellar atrophy	Normal
AH [174]	65	51/F	Turkish	SCA48	c.823_24delCT	L275Dfs*16frameshift	Gait ataxia, dysarthria	CCAS, apraxia	No abnormalities reported	Pyramidal features, Babinski sign	Chorea, dystonia,parkinsonism	Not reported	Anxiety, palilalia, dysphagia, cachexia, UI	Cerebellar atrophy,hypoperfusion in frontal, parietal and temporal lobes	Not reported
	66	60/F		Unconfirmed	Unconfirmed	Unconfirmed	Ataxia	CCAS, apraxia	No abnormalities reported	No abnormalities reported	Chorea, dystonia, parkinsonism	Not reported	Palilalia, cachexia, UI	Not reported	Not reported
	67	60/F		Unconfirmed	Unconfirmed	Unconfirmed	Ataxia	CCAS, apraxia	No abnormalities reported	No abnormalities reported	Chorea, dystonia, parkinsonism	Not reported	Palilalia, cachexia, UI	Not reported	Not reported
AI [175]	68	41/M	Not reported	SCA48	c.158T>C	I53T	Gait ataxia, dysarthria	Cognitive impairment/dementia	Saccadic eye movements	Increased tendon reflexes	Parkinsonism	Not reported	Diabetes mellitus, sleep apnea	Cerebellar atrophy, mild cortical atrophy	Not reported
	69	63/M			c.158T>C	I53T	Gait ataxia, dysarthria	Cognitive impairment	Saccadic eye movements	Increased tendon reflexes	No abnormalities reported	Not reported	No abnormalities reported	Cerebellar atrophy at superior vermis and dorsal region	Not reported
	70	26/M			c.158T>C	I53T	Gait ataxia, dysarthria	Dementia	Saccadic smooth pursuit, hypermetric saccades	Increased tendon reflexes	No abnormalities reported	Not reported	Drug and alcohol abuse	Marked cerebellar atrophy	Not reported
	71	-/F			c.158T>C	I53T	Gait ataxia	Severe cognitive impairment, childlike affect	Occasional saccadic interruptions	Increased tendon reflexes	No abnormalities reported	Not reported	Autism spectrum disorder, UI	Severe cerebellar atrophy, moderate cortical atrophy	Not reported
AJ [175]	72	41/F	Not reported	SCA48	c.111C>G	F37L	Gait ataxia, dysarthria	Cognitive impairment, confusion, personality change	Nystagmus	Normal	Unusual wing-beating tremor	Not reported	No abnormalities reported	Marked cerebellar atrophy, basis pontis flattening	Not reported
	73	-/M		Unconfirmed	Unconfirmed	Unconfirmed	Ataxia	No abnormalities reported	No abnormalities reported	No abnormalities reported	No abnormalities reported	Not reported	No abnormalities reported	Not reported	Not reported
	74	-/F		Unconfirmed	Unconfirmed	Unconfirmed	Ataxia	Cognitive impairment	No abnormalities reported	No abnormalities reported	No abnormalities reported	Not reported	Psychosis	Not reported	Not reported
AK [148]	75	45/F	European	SCAR16	c.358+1G>Ac.566A>C	--D189A	Ataxia, dysarthria	Normal	Mild pursuit abnormalities	Increased tendon reflexes	Dystonia	Normal	No abnormalities reported	Cerebellar hemisphere and vermis atrophy, normal basal ganglia	Normal
	76	46/F		Unconfirmed	Unconfirmed	Unconfirmed	Not reported	Memory impairment	No abnormalities reported	No abnormalities reported	No abnormalities reported	Not reported	Dysarthria	Not reported	Not reported
AL [176]	77	71/-	Dutch	SCA48	c.731_732delGC	C244Yfs*24	Gait ataxia	Cognitive impairment	No abnormalities reported	No abnormalities reported	Chorea	Not reported	Social withdrawal	Not reported	Not reported
	78	66/-		SCA48	c.731_732delGC	C244Yfs*24	Gait/limb ataxia, dysarthria	Cognitive impairment	No abnormalities reported	No abnormalities reported	No abnormalities reported	Not reported	No abnormalities reported	Cerebellar atrophy	Not reported
	79	72/-		SCA48	c.731_732delGC	C244Yfs*24	Gait ataxia, dysarthria	Cognitive impairment, memory defects	Saccadic pursuit	No abnormalities reported	No abnormalities reported	Not reported	No abnormalities reported	Cerebellar atrophy, generalized atrophy	Not reported
	80	69/-		SCA48	c.731_732delGC	C244Yfs*24	Gait ataxia, dysarthria	Cognitive impairment, memory defects, apraxia	Saccadic pursuit	No abnormalities reported	No abnormalities reported	Not reported	No abnormalities reported	Cerebellar atrophy	Not reported
	81	65/-		SCA48	c.731_732delGC	C244Yfs*24	No ataxia	Cognitive impairment, memory defects, impulsivity	Saccadic pursuit, gaze palsy	No abnormalities reported	Parkinsonism	Not reported	Dysarthria	Cerebellar atrophy, generalized atrophy	Not reported
	82	67/-		SCA48	c.731_732delGC	C244Yfs*24	Gait/limb ataxia, dysarthria	Cognitive impairment, executive dysfunction, apraxia	Normal	No abnormalities reported	No abnormalities reported	Not reported	No abnormalities reported	Generalized atrophy	Not reported
	83	61/-		SCA48	c.731_732delGC	C244Yfs*24	Gait ataxia, dysarthria	Cognitive impairment, memory defects	Gaze palsy	No abnormalities reported	Chorea, parkinsonism	Not reported	No abnormalities reported	Cerebellar atrophy	Not reported
	84	50/-		SCA48	c.731_732delGC	C244Yfs*24	Gait ataxia, dysarthria	Cognitive impairment, aggressiveness, childish behavior	Gaze palsy	No abnormalities reported	Chorea	Not reported	No abnormalities reported	Cerebellar atrophy	Not reported
	85	67/-		SCA48	c.731_732delGC	C244Yfs*24	Gait/limb ataxia, dysarthria	Cognitive impairment, executive dysfunction, impulsivity, memory defects	Saccadic pursuit, gaze palsy	No abnormalities reported	Chorea, parkinsonism	Not reported	No abnormalities reported	Generalized atrophy	Not reported
	86	--		Unconfirmed	Unconfirmed	Unconfirmed	Motor abnormalities	Cognitive abnormalities	Not reported
	87
	88
	89
	90
	91
AM [149]	92	29/M	Taiwanese	SCAR16	c.433A>CC.721C>T	K145QR241W	Gait ataxia, dysarthria	Normal	Cogwheel pursuit	No abnormalities reported	No abnormalities reported	Normal	No abnormalities reported	Cerebellar atrophy	Not reported
AN [149]	93	22/M	Taiwanese	SCAR16	C.433A>CC.694T>G	K145QC232G	Gait ataxia, dysarthria	Mild cognitive impairment	Cogwheel pursuit	No abnormalities reported	No abnormalities reported	Normal	No abnormalities reported	Cerebellar atrophy	Normal
	94	39/M		SCAR16	C.433A>CC.694T>G	K145QC232G	Gait unsteadiness, dysarthria	No abnormalities reported	No abnormalities reported	No abnormalities reported	No abnormalities reported	Normal	No abnormalities reported	Cerebellar atrophy	Not reported
	95	-/M		Unconfirmed	Unconfirmed	Unconfirmed	Gait difficulties	No abnormalities reported	No abnormalities reported	No abnormalities reported	No abnormalities reported	Not reported	No abnormalities reported	Not reported	Not reported
AO [150]	96	44/M	European	SCA48	c.338 C>A	A113D	Ataxia	Normal	Saccadic pursuit	No abnormalities reported	Myoclonus	Not reported	Weight loss	Not reported	Not reported
	97	40/F		SCA48	c.338 C>A	A113D	Ataxia	Cognitive impairment	Not reported	Pyramidal features	Chorea	Not reported	Lower limb wasting	Not reported	Not reported
AP [150]	98	30/M	European	SCA48	c.728C>T	P243L	Ataxia	Normal	Nystagmus, saccadic pursuit	No abnormalities reported	No abnormalities reported	Not reported	Lower limb wasting	Not reported	Not reported
AQ [150]	99	27/F	European	SCA48	c.787-1G>C	Splice acceptor	Ataxia, dysarthria	Cognitive impairment, apathy	No abnormalities reported	Babinski sign	Myoclonus	Not reported	No abnormalities reported	Not reported	Not reported
AR [150]	100	51/F	European	SCA48	c.194A>G	N65S	Ataxia	Normal	Nystagmus, ophthalmoplegia	No abnormalities reported	No abnormalities reported	Not reported	Hearing loss	Not reported	Not reported
	101	55/F		SCA48	c.194A>G	N65S	Ataxia	Cognitive impairment	No abnormalities reported	No abnormalities reported	No abnormalities reported	Not reported	No abnormalities reported	Not reported	Not reported
	102	30/F		SCA48	c.194A>G	N65S	Ataxia	Normal	Nystagmus, saccadic pursuit	Babinski sign	Dystonia, retrocollis	Not reported	Scoliosis	Not reported	Not reported
AS [150]	103	29/F	European	SCA48	c.596G>T*c.635A>G*	C199FD212G	Ataxia, dysarthria	Cognitive impairment	Nystagmus, saccadic pursuit	Pyramidal features, Babinski sign	No abnormalities reported	Not reported	No abnormalities reported	Not reported	Not reported
	104	23/F		SCA48	c.596G>T*c.635A>G*	C199FD212G	Ataxia, dysarthria	Cognitive impairment	Nystagmus, saccadic pursuit	Pyramidal features, Babinski sign	No abnormalities reported	Not reported	Decreased vibration sense	Not reported	Not reported
	105	29/M		SCA48	c.596G>T*c.635A>G*	C199FD212G	Ataxia	Normal	Saccadic pursuit	No abnormalities reported	Tremor	Not reported	No abnormalities reported	Not reported	Not reported
AT [150]	106	60/F	European	SCA48	c.136G>C	A46P	Ataxia	Normal	Nystagmus	No abnormalities reported	No abnormalities reported	Not reported	No abnormalities reported	Not reported	Not reported
	107	38/F		SCA48	c.136G>C	A46P	Ataxia, dysarthria	Cognitive impairment	Saccadic pursuit	No abnormalities reported	Hypokinesia, facial chorea	Not reported	Scoliosis	Cerebellar, cortical, and pons atrophy	Not reported
	108	47/F		SCA48	c.136G>C	A46P	Ataxia, dysarthria	Cognitive impairment, aggressive	No abnormalities reported	No abnormalities reported	Chorea, facial dystonia	Not reported	No abnormalities reported	Not reported	Not reported
AU [150]	109	30/F	European	SCA48	c.460C>T	R154C	Ataxia	Cognitive impairment, frontal syndrome, pseudobulbar affect	Saccadic pursuit	No abnormalities reported	Hypokinesia	Not reported	No abnormalities reported	Not reported	Not reported
AV [150]	110	30/M	European	SCA48	c.358+2T>C	Splice donor	Ataxia, spastic gait	Cognitive impairment, frontal syndrome	No abnormalities reported	No abnormalities reported	Chorea	Not reported	No abnormalities reported	Not reported	Not reported
AW [150]	111	52/F	European	SCA48	c.426_428del	K143del	Ataxia	CCAS, FTD- like	Nystagmus	No abnormalities reported	Hypokinesia, myokymia	Not reported	No abnormalities reported	Not reported	Not reported
AX [150]	112	27/F	European	SCA48	c.194A>G	N65S	Ataxia	No abnormalities reported	No abnormalities reported	No abnormalities reported	No abnormalities reported	Not reported	No abnormalities reported	Not reported	Not reported
AY [150]	113	53/F	European	SCA48	c.440G>A	W147*	Ataxia, dysarthria	Cognitive impairment	No abnormalities reported	Babinski sign	Axial hypertonus	Not reported	No abnormalities reported	Not reported	Not reported
AZ [150]	114	40/M	European	SCA48	c.136G>C	Y49C	Ataxia, dysarthria, dysmetria	Cognitive impairment, FTD-like, apraxia	Saccadic pursuit, nystagmus	No abnormalities reported	Dystonia	Not reported	Weight loss, lower limb wasting	Cerebellar atrophy	Not reported
	115	51/F		SCA48	c.136G>C	Y49C	Ataxia, dysarthria	CCAS, pseudobulbar affect	No abnormalities reported	No abnormalities reported	No abnormalities reported	Not reported	Dysphonia	Not reported	Not reported
	116	38/F		SCA48	c.136G>C	Y49C	Ataxia, dysarthria	Cognitive impairment, frontal syndrome	Diplopia, ptosis	No abnormalities reported	No abnormalities reported	Not reported	No abnormalities reported	Cerebellar atrophy	Not reported
BA [150]	117	41/F	European	SCA48	c.194A>G	N65S	Ataxia, dysarthria	Cognitive impairment, frontal syndrome	Oculomotor apraxia, slow saccades	No abnormalities reported	No abnormalities reported	Not reported	No abnormalities reported	Not reported	Not reported
BB [150]	118	-/F	European	SCA48	c.301G>C	G101R	Ataxia	Normal	Slow saccades	No abnormalities reported	Neck and shoulder dystonia	Not reported	Fasciculations	Not reported	Not reported
	119	25/M		SCA48	c.301G>C	G101R	Ataxia	Normal	Diplopia, nystagmus, saccadic pursuit	Babinski sign	No abnormalities reported	Not reported	No abnormalities reported	Not reported	Not reported
	120	30/M		SCA48	c.301G>C	G101R	Ataxia, dysarthria	Normal	Nystagmus, saccadic pursuits, ophthalmoplegia	Pyramidal features	Head tremor	Not reported	No abnormalities reported	Not reported	Not reported
	121	50/F		SCA48	c.301G>C	G101R	Ataxia	Cognitive impairment, apathy	Slow saccades, saccadic pursuit	Babinski sign	Tremor, hypokinesia	Not reported	Epilepsy	Not reported	Not reported
	122	-/F		SCA48	c.301G>C	G101R	Ataxia	Normal	Saccadic pursuit, ophthalmoplegia	Babinski sign	No abnormalities reported	Not reported	Decreased vibration sense	Not reported	Not reported
BC [150]	123	-/F	European	SCA48	c.824delT	L275Rfs*11	Ataxia	Normal	Slow saccades, oculomotor apraxia	No abnormalities reported	Tremor	Not reported	No abnormalities reported	Not reported	Not reported
BD [150]	124	40/F	European	SCA48	c.235G>C	A79P	Ataxia, dysarthria	Cognitive impairment	No abnormalities reported	Babinski sign	No abnormalities reported	Not reported	Epilepsy	Not reported	Not reported
BE [150]	125	33/F	European	SCA48	c.728C>T	P243L	Ataxia, dysarthria	Normal	Nystagmus	Pyramidal features	Facial dystonia	Not reported	No abnormalities reported	Not reported	Not reported
BF [150]	126	50/M	European	SCA48	c.665G>A	R222K	Ataxia, dysarthria	Normal	Slow saccades, saccadic pursuit	Babinski sign	Tremor	Not reported	Orthostatic hypotension	Not reported	Not reported
BG [150]	127	48/F	European	SCA48	c.849C>G	N283K	Ataxia, dysarthria	Normal	No abnormalities reported	No abnormalities reported	No abnormalities reported	Not reported	No abnormalities reported	Not reported	Not reported
BH [150]	128	65/F	European	SCA48	c.544C>T	R182*	Ataxia	CCAS, frontal syndrome	No abnormalities reported	No abnormalities reported	No abnormalities reported	Not reported	CJD-like	Not reported	Not reported
BI [150]	129	65/F	European	SCA48	c.221A>C	Q74P	Ataxia, dysarthria	Cognitive impairment	Oculomotor apraxia	No abnormalities reported	No abnormalities reported	Not reported	Facial dysmorphism	Not reported	Not reported
BJ [150]	130	44/M	European	SCA48	c.206G>A	C69Y	Ataxia	Cognitive impairment	Saccadic pursuit	No abnormalities reported	No abnormalities reported	Not reported	Cataract	Not reported	Not reported
BK [150]	131	67/F	European	SCA48	c.794G>A	G265D	Ataxia, dysarthria	Cognitive impairment	Slow saccades, saccadic pursuit	No abnormalities reported	FXTAS-like	Not reported	Cataract	Not reported	Not reported
BL [150]	132	60/M	European	SCA48	c.170C>T	P57L	Ataxia	Cognitive impairment, impulsive, aggressive	Diplopia, ptosis	Babinski sign	Tremor, rigidity	Not reported	No abnormalities reported	Not reported	Not reported
	133	40/F		SCA48	c.170C>T	P57L	Ataxia, dysarthria	Normal	Diplopia, ptosis	No abnormalities reported	Hypokinesia	Not reported	No abnormalities reported	Not reported	Not reported
	134	40/F		SCA48	c.170C>T	P57L	Ataxia, dysarthria	Normal	No abnormalities reported	No abnormalities reported	No abnormalities reported	Not reported	No abnormalities reported	Not reported	Not reported
BM [150]	135	27/F	European	SCA48	c.426_428del	K143del	Ataxia	Normal	Nystagmus, saccadic pursuit	No abnormalities reported	Cervical dystonia	Not reported	No abnormalities reported	Not reported	Not reported
	136	M		SCA48	c.426_428del	K143del	No ataxia	Normal	Saccadic pursuit	No abnormalities reported	No abnormalities reported	Not reported	No abnormalities reported	Not reported	Not reported
BN [150]	137	51/W	European	SCA48	c.146A>G	Y49C	Ataxia, dysarthria	Cognitive impairment, frontal syndrome	No abnormalities reported	Increased tendon reflexes	Chorea, dystonia, rigidity	Not reported	No abnormalities reported	Not reported	Not reported
	138	37/M		SCA48	c.146A>G	Y49C	Ataxia, dysarthria	Cognitive impairment, behavior change	Slow saccades, saccadic pursuit, oculomotor apraxia	No abnormalities reported	Chorea, myoclonus	Not reported	No abnormalities reported	Not reported	Not reported
	139	37/F		SCA48	c.146A>G	Y49C	Ataxia, dysarthria	Cognitive impairment, behavior change	Slow saccades, saccadic pursuit, oculomotor apraxia	No abnormalities reported	Chorea, myoclonus, rigidity	Not reported	No abnormalities reported	Not reported	Not reported
	140	-/W		SCA48	c.146A>G	Y49C	Ataxia	No abnormalities reported	No abnormalities reported	No abnormalities reported	No abnormalities reported	Not reported	No abnormalities reported	Not reported	Not reported
BO [150]	141	74/W	European	SCA48	c.206G>A	C69Y	Ataxia	Cognitive impairment	Nystagmus, slow saccades, ophthalmoplegia	No abnormalities reported	Tremor, hypokinesia	Not reported	No abnormalities reported	Not reported	Not reported
BP [150]	142	-/W	European	SCA48	c.502C>T	L168F	Ataxia	Mild cognitive impairment	No abnormalities reported	No abnormalities reported	No abnormalities reported	Not reported	No abnormalities reported	Not reported	Not reported
BQ [150]	143	20/M	European	SCAR16	c.359C>Tc.632T>C	A120VM211T	Ataxia	Cognitive impairment	Nystagmus, saccadic pursuit	No abnormalities reported	No abnormalities reported	Not reported	No abnormalities reported	Not reported	Not reported
	144	21/F		SCAR16	c.359C>Tc.632T>C	A120VM211T	Ataxia	Cognitive impairment	No abnormalities reported	Babinski sign	Voice and neck dystonia	Not reported	No abnormalities reported	Not reported	Not reported
BR [150]	145	17/M	European	SCAR16	c.691C>Gc.691C>G	L231VL231V	Ataxia	Cognitive impairment	No abnormalities reported	Babinski sign	No abnormalities reported	Hypogonadism	Dysphagia, UI	Not reported	Not reported

Grey boxes indicate patients with disease symptoms who did not have genetic testing. These patients without a confirmed genotype were not included in patient numbers indicated within the manuscript text. Where only one STUB1 mutation/CHIP mutation is indicated, the patient is heterozygous with one mutant allele and one wild-type allele. AAO: age at disease onset; CCAS: cerebellar cognitive affective syndrome; CJD: Creutzfeldt-Jakob disease; FTD: frontotemporal dementia; FXTAS: fragile X-associated tremor/ataxia syndrome; NCV: nerve conduction velocity; SCA48: spinocerebellar ataxia autosomal dominant type 48; SCAR16: spinocerebellar ataxia autosomal recessive type 16; T1D: type 1 diabetes; UI: urinary incontinence; UTR: untranslated region; WES: whole exome sequencing. *Variants in cis.