The polyglutamine protein ATXN2: from its molecular functions to its involvement in disease

Abstract

A CAG repeat sequence in the ATXN2 gene encodes a polyglutamine (polyQ) tract within the ataxin-2 (ATXN2) protein, showcasing a complex landscape of functions that have been progressively unveiled over recent decades. Despite significant progresses in the field, a comprehensive overview of the mechanisms governed by ATXN2 remains elusive. This multifaceted protein emerges as a key player in RNA metabolism, stress granules dynamics, endocytosis, calcium signaling, and the regulation of the circadian rhythm. The CAG overexpansion within the ATXN2 gene produces a protein with an extended poly(Q) tract, inducing consequential alterations in conformational dynamics which confer a toxic gain and/or partial loss of function. Although overexpanded ATXN2 is predominantly linked to spinocerebellar ataxia type 2 (SCA2), intermediate expansions are also implicated in amyotrophic lateral sclerosis (ALS) and parkinsonism. While the molecular intricacies await full elucidation, SCA2 presents ATXN2-associated pathological features, encompassing autophagy impairment, RNA-mediated toxicity, heightened oxidative stress, and disruption of calcium homeostasis. Presently, SCA2 remains incurable, with patients reliant on symptomatic and supportive treatments. In the pursuit of therapeutic solutions, various studies have explored avenues ranging from pharmacological drugs to advanced therapies, including cell or gene-based approaches. These endeavours aim to address the root causes or counteract distinct pathological features of SCA2. This review is intended to provide an updated compendium of ATXN2 functions, delineate the associated pathological mechanisms, and present current perspectives on the development of innovative therapeutic strategies.

Fig. 1. Chronology of key ataxin-2 discoveries and associated diseases.

A schematic representation highlighting key milestones in the research on ataxin-2 and related diseases, including SCA2 and ALS. The timeline spans from the initial characterization of SCA2 to the latest insights into the protein’s proposed functions.

Fig. 2. Schematic representations of the ATXN2 gene and ATXN2 protein.

The ATXN2 gene (A) comprises 25 exons (green) encoding ATXN2. ATXN2 contains a repetitive CAG region of variable length in exon 1, which may lead to the development of SCA2 when this sequence is expanded beyond 31 CAG units, with full penetrance above 35 repeats. B ATXN2 contains 5 protein domains beyond the polyQ tract: SBM1, Lsm, LsmAD, SBM2, and PAM2. The LsmAD domain includes a caspase-3 cleavage site, a clathrin-mediated trans-Golgi signal and an endoplasmic reticulum (ER) exit signal. Expanded ATXN2 alleles originate ATXN2 forms with expanded polyglutamine (polyQ) sequences, which may lead to alterations in intra- and intermolecular dynamics.

Fig. 3. Molecular functions of ATXN2.

ATXN2 is implicated in many distinct cellular processes, including (A) the positive regulation of mRNA translation by directly binding to and stabilizing mRNAs and, conversely, (B) the negative regulation of mRNA translation by binding to PABP-1 and impairing the formation of the translation initiation complex. C ATXN2 is a regulator of metabolism as it can sequester mTORC1 into SGs under nutrient deprivation conditions, thus hampering downstream signalling and, consequently, protein synthesis and cell growth. D ATXN2 was implicated in calcium homeostasis maintenance by negatively regulating mGluR1 through stabilization of the mRNA of its antagonist RGS8. E ATXN2 is also associated with endocytosis by interacting with endophilins, Cbl, and CIN85, and was shown to influence the internalization rate of EGFR.

Fig. 4. Polyglutamine diseases and spinocerebellar ataxias at the intersection of neurodegeneration and repeat expansion mutations.

Polyglutamine diseases are a group of neurodegenerative diseases, a wider collection of neurologic disorders that includes Alzheimer’s disease (AD), Parkinson’s disease (PD), amyotrophic lateral sclerosis (ALS), among many others. PolyQ diseases are hereditary diseases caused by expansion of CAG microsatellites in the coding region of the respective causative genes. The group comprises Huntington’s disease (HD), spinal and bulbar muscular atrophy (SBMA) and seven types of spinocerebellar ataxias (SCA): SCA1, 2, 3 (also known as Machado-Joseph disease, MJD), 6, 7, 17 and dentatorubral-pallidoluysian atrophy (DRPLA). SCAs are autosomal dominant neurodegenerative disorders typically manifested by progressive ataxia, and not all are associated with polyQ expansions. Among the ones for which a causative mutation has been described, some are caused by nucleotide repeat expansion outside coding regions (for example SCA8, 10 and 12) and others by conventional mutations (examples include SCA5 and SCA11). While neurodegenerative disorders other than polyQ diseases and SCAs are also caused by repeat expansion mutations (including progressive myoclonic epilepsy type q, EPM1), not all repeat expansion diseases involve a clear neurodegenerative profile. Examples include myotonic dystrophy (DM), familial cortical myoclonic tremor with epilepsy type 1 (FCMTE), Fuchs endothelial corneal dystrophy (FECD), fragile X syndrome (FXS) and oculopharyngeal muscular dystrophy (OPMD). Only a portion of ALS cases are known to be associated with repeat expansions.

Fig. 5. Molecular mechanisms proposed to be involved in SCA2 pathogenesis.

Both repeat-expanded sense and anti-sense transcripts of ATXN2 can form hairpin structures and cause toxicity, presumably by sequestering RBPs into RNA foci. The translation of the sense transcript results in a polyQ-expanded ATXN2 protein that is prone to adopt a β-sheet-rich structure and form cytoplasmic insoluble aggregates that eventually recruit other proteins, such as ataxin-1, ataxin-3, and TBP. The accumulation of SQSTM1 and LC3-II indicates a dysfunction of the autophagic pathway, suggesting that neurons struggle to clear out damaged and aggregated proteins, leading to an overall loss of proteostasis. Additionally, low levels of PINK1 in SCA2 patients indicate mitochondrial dysfunction and enhanced oxidative stress, which can contribute to neuroinflammation and neuronal death. Mutant ATXN2 aberrantly interacts with the ER InsP₃R1 and increases its sensitivity to IP₃, resulting in high calcium leakage into the cytosol that can trigger excitotoxicity and neurodegeneration. Interneuronal transfer of mutant ATXN2 could constitute a mechanism contributing to SCA2 progression, whereby cells that are chiefly affected disseminate toxic species to the vicinity, a process known as “disease spreading”.