Protein aggregation and degradation mechanisms in neurodegenerative diseases

Abstract

Neurodegenerative diseases are characterized by selective neuronal vulnerability and neurodegeneration in specific brain regions. The pathogenesis of these disorders centrally involves abnormal accumulation and aggregation of specific proteins, which are deposited in intracellular inclusions or extracellular aggregates that are characteristic for each disease. Increasing evidence suggests that genetic mutations or environmental factors can instigate protein misfolding and aggregation in these diseases. Consequently, neurodegenerative diseases are often considered as conformational diseases. This idea is further supported by studies implicating that impairment of the protein quality control (PQC) and clearance systems, such as the ubiquitin-proteasome system and autophagosome-lysosome pathway, may lead to the abnormal accumulation of disease-specific proteins. This suggests that similar pathological mechanisms may underlie the pathogenesis of the different neurodegenerative disorders. Interestingly, several proteins that are known to associate with neurodegenerative diseases have been identified as important regulators of PQC and clearance systems. In this review, we summarize the central features of abnormal protein accumulation in different common neurodegenerative diseases and discuss some aspects of specific disease-associated proteins regulating the PQC and clearance mechanisms, such as ubiquilin-1.

Introduction

The pathogenesis of different neurodegenerative diseases, such as Alzheimer’s (AD), Parkinson’s (PD), and Huntington’s disease (HD), shares several common features. One of these is the abnormal accumulation and aggregation of disease-specific proteins, which is suggested to lead to neurodegeneration [1,2]. The accumulated proteins typically form intracellular inclusions or extracellular aggregates in specific brain areas. These are considered specific pathological hallmarks for the diseases. The proteins that accumulate in neurodegenerative diseases are typically misfolded and yield a β-sheet structure that promotes aggregation and fibril formation, suggesting that these diseases are conformational diseases [1,2]. Genetic mutations or different environmental factors, such as oxidative or metabolic stress, can induce protein misfolding and aggregation, but the exact underlying mechanisms of protein aggregation in different neurodegenerative disorders are still not completely understood.

It is estimated that approximately 30% of newly synthesized proteins are incorrectly folded and degraded [3]. Under normal conditions, the cells are able to efficiently utilize their protein quality control (PQC) system to handle the misfolded proteins and maintain the protein homeostasis. The molecular chaperones involved in the cellular PQC systems, such as heat shock proteins (Hsp), recognize misfolded proteins, assist in their refolding, prevent their aggregation, and help to repair the damaged proteins [4]. The molecular chaperones may also interact with the ubiquitination machinery and target the misfolded proteins to degradation by the ubiquitin-proteasome system (UPS) [5,6] or the autophagosome-lysosome pathway (ALP) [7-9]. Accumulating evidence suggests that deficiencies in PQC and clearance mechanisms may lead to the abnormal accumulation of proteins in neurodegenerative diseases. Moreover, the excessive accumulation of misfolded and aggregated proteins may overwhelm the PQC and clearance systems, which leads to further protein accumulation, cellular stress, and finally to neurodegeneration [7,10,11]. Interestingly, several proteins that are known to associate with the pathogenesis of specific neurodegenerative diseases have also been identified as central regulators of PQC and protein clearance systems [12]. In this review, we will discuss common aspects of PQC systems and specific proteins, such as ubiquilin-1, regulating the protein levels, accumulation, and targeting in the context of common neurodegenerative diseases.

Protein accumulation and pathological inclusions in neurodegenerative diseases

Abnormal intracellular or extracellular protein accumulation in the affected brain regions is a typical pathological hallmark of neurodegenerative diseases and thought to lead to neurotoxicity, neurodegeneration, and finally clinical manifestation of the disease. The intracellular inclusions detected in the brain of patients with a neurodegenerative disease are often ubiquitin-positive and contain misfolded disease-specific proteins that have acquired an amyloidogenic conformation containing β-sheet structures [1,2]. This conformational change results in the exposure of the hydrophobic regions of the protein that are normally buried within the protein structure when it is in its natively folded conformation. The exposed hydrophobic structures promote oligomerization and aggregation of the protein [1,13-15].

In AD, the pathological hallmarks in specific cortical areas of the brain include extracellular amyloid plaques consisting of aggregated β-amyloid peptide, and intracellular neurofibrillary tangles (NFTs) containing aggregated, hyperphosphorylated tau protein [16,17]. In a subset of patients with frontotemporal dementia (FTD), TAR DNA-binding protein 43 (TDP-43)- or tau-positive inclusions are detected [18]. In PD, the intracellular Lewy bodies containing aggregated α-synuclein [19], and in HD, the intranuclear inclusions of aggregated huntingtin protein containing polyglutamine (polyQ) expansion [20], are typical hallmarks. In amyotrophic lateral sclerosis (ALS), aggregates of superoxide dismutase (SOD) in motor neurons can be detected [21]. It has been suggested that genetic mutations, environmental factors, or different stress conditions induce protein misfolding and aggregation in these diseases [22], implicating that similar pathological mechanisms may underlie their pathogenesis. Recent evidence also indicates that the aggregated proteins may spread from one cell or brain area to another and function as seeds to instigate protein misfolding and aggregation in these previously unaffected cells or areas [23]. This may explain the gradual progression of the disease pathology in the brain over time in the case of many neurodegenerative disorders.

Underlying factors of protein misfolding and aggregation

Several genetic and environmental factors have been suggested to promote protein misfolding and aggregation in different neurodegenerative diseases. These include gene mutations, gene dose, and promoter polymorphisms, which may affect protein levels and conformation. Also, inefficient protein biogenesis, excess unassembled units of oligomeric protein complexes, and inefficient translocation of secretory or mitochondrial protein precursors may result in the accumulation of misfolded proteins [22,24]. Different conditions, such as metabolic or environmental stress or aging, further increase the production of misfolded proteins and thus challenge the capacity of the PQC system [25]. It is also suggested that during aging, the cells lose their ability to efficiently deal with misfolded proteins as the capacity of the PQC system declines. Reduced activity of UPS and ALP are known to associate with neurodegenerative diseases and aging [26-31]. In neurodegenerative diseases, the deficiencies in the PQC system together with mutations in the disease-associated proteins and inflammation and oxidative stress, which are intimately involved in the pathogenesis of these diseases, further enhance the accumulation and aggregation of proteins and may lead to aberrant protein modifications. These mechanisms together are thought to underlie the excessive accumulation and aggregation of proteins, which cause neuronal dysfunction and neurotoxicity and ultimately lead to widespread neurodegeneration [32].

Many neurodegenerative diseases have a strong genetic component that affects the disease onset and progression. The patients with Down syndrome are a good example of the effects of gene dose on disease pathogenesis. These individuals develop AD-like pathology and typically suffer from the symptoms of AD already early in their life. The trisomy in the chromosome 21 leads to the triplication of the APP gene, resulting in increased levels of the APP protein and subsequently augmented early deposition of β-amyloid [33,34]. Also, PD may be caused by α-synuclein gene locus triplication, in addition to mutations in the α-synuclein gene [35,36]. Indeed, many neurodegenerative diseases are associated with the inheritance of pathological gene mutations that lead to the disease onset. These forms of the disease are typically rare, but cause an earlier onset of the disease and often lead to a more aggressive disease progression as compared to the more common sporadic cases with a late-onset and slow progression. For example, the P301L mutation in the MAPT gene encoding tau protein or mutations in other genes, such as C9ORF72 or GRN, are known to cause FTD [18]. Rare early-onset familial forms of ALS (fALS) are caused by mutations in the SOD1, TARDBP (encoding TDP-43), or FUS genes, but the pathogenesis can differ between these different types of fALS [37]. Early-onset familial forms of AD are known to result from mutations in genes encoding APP or presenilins (PSEN) [38]. A subset of early-onset PD cases has mutations in the genes encoding parkin, DJ-1 or PINK1 [39-41]. On the other hand, apart from gene mutations, polymorphisms in the promoter areas of disease-associated genes may lead to increased transcription or alternative splicing of the gene, resulting in a general increase in protein levels or in the levels of a more aggregation-prone transcript variant [42].

In addition to the genetic factors, the pathogenesis of many neurodegenerative diseases is often associated with inflammation and increased oxidative stress [32]. These conditions may induce alterations in the covalent post-transcriptional protein modifications, which consequently may affect the protein function, interaction, and levels. These modifications include changes in protein oxidation, nitration, phosphorylation, ubiquitination, SUMOylation, and proteolytic cleavage [25]. The levels of oxidized and nitrated proteins are known to increase in the brains of AD patients, implying that inflammation and oxidative stress are central phenomena that associate with the disease process [43-45]. In AD, hyperphosphorylation of the tau protein is known to result in increased tau aggregation and destabilization of the microtubules. Tau aggregates are harmful and they further accumulate as NFTs within neurons, eventually causing neurodegeneration [46]. In addition, missorting of tau from the axons to the somatodendritic compartment in AD brain is an early sign of neurodegeneration and is at least partially caused by oligomeric β-amyloid peptides [47,48]. Furthermore, increased proteolytic processing of APP by β- and γ-secretases leads to an augmented generation of β-amyloid, and consequently increased deposition of amyloid plaques [49]. The familial causative mutations in APP or PSEN genes also result in the enhanced cleavage of APP by β- and γ-secretase and subsequently increased β-amyloid production [38,49,50], confirming that enhanced amyloidogenic processing of APP is a central mechanism underlying the pathogenesis of AD. Interestingly, many disease-specific inclusions, such as NFTs or Lewy bodies, typically contain proteins that are ubiquitinated [2]. This suggests that the accumulated proteins with abnormal conformation have been detected and tagged by the PQC system, but due to inefficient or impaired clearance, they remain within the cells and are deposited in the inclusions.

Protein quality control systems

In the case of protein misfolding and aggregation in cells, the PQC system uses three main parallel strategies to maintain protein homeostasis (Figure 1). The misfolded protein may be refolded to recover the protein’s normal conformation. Different molecular chaperones, such as Hsps, play an essential role in protein refolding. Alternatively, if the protein cannot be refolded, it is targeted to the UPS or ALP for degradation. In the case when the misfolded or aggregated proteins cannot be directed for refolding or degradation, they may be sequestered as specific protein inclusions within the cells. Also these steps involve the function of molecular chaperone proteins [9].

Figure 1.

The main parallel strategies to maintain protein homeostasis. The misfolded proteins may be refolded, degraded, or sequestered within cells. All these mechanisms centrally involve the function of different molecular chaperones, such as heat shock proteins (Hsps). Abbreviations: ALIS, aggresome-like inducible structure; ALP, autophagosome-lysosome system; IPOD, insoluble protein deposit; JUNQ, juxtanuclear quality control; UPS, ubiquitin-proteasome system.

The signal for targeting a protein for UPS- or ALP-mediated degradation is polyubiquitination. When the protein is polyubiquitinated, it is covalently tagged with four or more ubiquitin molecules in its lysine residues [51]. Ubiquitin itself has seven lysines (e.g. K48 and K63) and canonical binding of other ubiquitin molecules to these internal lysines then forms the polyubiquitin chain. Ubiquitination of proteins requires the coordinated function of different ubiquitin ligases E1, E2 and E3. Differently linked polyubiquitin chains have been shown to mediate differential targeting of the polyubiquitinated proteins. K48-linked polyubiquitin chains are the classical signal for proteasomal degradation. The ubiquitinated proteins are then recognized and degraded by the 26S proteasome, which is a complex structure comprising two regulatory 19S subunits and the 20S catalytic core subunit [52]. The 26S proteasome is a barrel-shaped structure containing a channel through which the protein travels and is enzymatically degraded on the way. During degradation, the ubiquitin moieties are removed from the proteins by deubiquitinating enzymes and recycled [52]. The K63-linked polyubiquitination, on the other hand, may target the protein for autophagy [53]. In the ALP, the proteins or protein aggregates are engulfed within a double-membrane, which forms the autophagosome. Different autophagy receptors, such as p62/SQSTM1, are essential in the recruitment of K63-ubiquitin-linked proteins to autophagic degradation [54,55]. The Atg family proteins on the membranes of the autophagosomes are important for the formation of the autophagosomal vesicles. Finally, the mature autophagosomes fuse with lysosomes, which results in the degradation of the contents of the autophagosome [56].

If the protein cannot be refolded and is not targeted for degradation, it may be sequestered to a specific cellular site to generate an intracellular inclusion body, such as an aggresome [57-59]. The presence of the inclusions typically reflects a pathological state and can be used as a disease marker. However, according to current understanding, the formation of the inclusion bodies likely functions as a cytoprotective mechanism rather than a pathogenic one [57]. Active sequestration of misfolded proteins or protein aggregates in intracellular inclusions may reduce the accumulation of potentially toxic protein oligomers and aggregates and prevent abnormal interactions of these with other proteins, cell organelles, or the PQC machinery [9]. Protein refolding, targeting to UPS- or ALP-mediated degradation, or sequestration in inclusions centrally involve the function of different molecular chaperones. Interestingly, a number of these proteins have been associated with the pathogenesis of neurodegenerative diseases.

Protein sequestration in intracellular compartments

Increasing evidence suggests that impairment in the protein clearance systems, such as UPS, ALP or chaperone-mediated autophagy (CMA), takes place in the diseased brain [26,27,29-31]. The abnormal protein accumulation may further overwhelm these systems and, as the result, even more proteins start accumulating within the cells. Furthermore, the aggregated proteins typically cannot be degraded by the UPS, shifting the burden in their clearance to the ALP [60]. When the level of protein accumulation and aggregation exceeds the capacity of the UPS or ALP disposal pathways, the misfolded or accumulated proteins may be actively compartmentalized as different kinds of inclusions at specific cellular sites to minimize their toxic effects. Depending on their solubility and other properties, the proteins may be targeted to different kinds of compartments or inclusions (Figure 2).

Figure 2.

Protein sequestration into different compartments. The misfolded or aggregated proteins may be targeted to different intracellular compartments. After misfolding, most proteins are recognized and ubiquitinated, which directs them to the JUNQ, a region that contains chaperones and 26S proteasomes. JUNQ concentrates soluble misfolded proteins, which may be proteasomally degraded or refolded by the chaperones. The insoluble aggregated proteins which may not be ubiquitinated, such as HD-associated huntingtin or prions, can be targeted to IPOD. It does not contain proteasomes, but colocalizes with autophagy-associated proteins, such as Atg8. Aggregated proteins can also be targeted to the aggresomes, which localize at the MTOC, are surrounded by a vimentin envelope, and cause an indentation of the nucleus. Aggresomes also contain chaperones and components of the ubiquitin-proteasome system (UPS). Targeting to the aggresomes, IPOD, or JUNQ involves active retrograde transport of the cargo by the motor proteins on the microtubules. Immune activation or stress conditions may induce the formation of transient ALIS inclusions. These colocalize with ubiquitin and p62/SQSTM1 and concentrate soluble proteins targeted to clearance by the UPS or autophagy. Abbreviations: ALIS, aggresome-like induced structure; IPOD, insoluble protein deposit; JUNQ, juxtanuclear quality control; MT, microtubule; MTOC, microtubule-organizing center; N, nucleus.

Insoluble misfolded proteins are often targeted to juxtanuclear structures termed aggresomes [57-59]. The misfolded proteins or protein aggregates are ubiquitinated and actively transported via the microtubules on retrograde motor proteins to the microtubule-organizing center (MTOC) next to the nucleus. There they typically colocalize with γ-tubulin, an MTOC marker. The aggresome core contains components of the proteasome, Hsps, and mitochondria in addition to the ubiquitinated misfolded proteins. The core is enveloped within a cage formed by vimentin or other filament proteins. The aggresomes appear to be relatively stable structures within cells, but there is increasing evidence that they are eventually cleared from the cells by autophagy [57-59]. Soluble misfolded proteins that are targeted to the UPS for degradation or alternatively to refolding by cytoplasmic chaperones are typically concentrated to a structure called juxtanuclear quality control (JUNQ), which contains chaperones and proteasomal subunits [61]. Here, the proteins may be refolded or degraded by the proteasome. The insoluble aggregated proteins, such as disease-associated huntingtin or prion proteins, can be directed to a compartment termed insoluble protein deposit (IPOD) [61]. These colocalize with autophagy-associated proteins, suggesting that the IPODs or their constituents may be disposed of by autophagy. Enhanced ubiquitination of the protein at IPOD may redirect it to JUNQ. In addition, immune activation or stress conditions may induce the formation of transient inclusions termed aggresome-like inducible structure (ALIS) that colocalize with ubiquitin and p62/SQSTM1 and concentrate soluble proteins targeted to clearance by the UPS or autophagy [62].

Accumulating evidence implies that sequestration of potentially harmful misfolded or aggregated proteins into specific compartments and formation of intracellular inclusions is a cytoprotective response, which aims to prevent unspecific interactions of the harmful proteins or protein aggregates with other proteins, cell organelles, or components of the PQC system and thus diminish their toxicity [9,63]. It has been suggested that the soluble misfolded oligomers or aggregates are especially toxic and reduce the capacity of protein folding systems by sequestering chaperones and other factors. This leads to impaired protein homeostasis. In contrast, large insoluble aggregates can be protective and promote cell survival [64]. Therefore, the compartmentalization of the harmful misfolded proteins and protein aggregates may enhance their clearance and prevent them from blocking the UPS or ALP and occupying the cellular chaperones.

Disease-associated proteins involved in protein quality control systems

Several key proteins, which are involved in the function of UPS and ALP at different stages, are known to contain mutations that lead to neurodegenerative disease. Many of these proteins typically contain domains that mediate interaction with polyubiquitinated proteins or the proteasome, such as ubiquitin-like domain (UBL), ubiquitin-associated domain (UBA), or ubiquitin-interacting motifs (UIM) [65,66]. The presence of such domains in these proteins suggests that they may function as shuttles targeting polyubiquitinated proteins to the UPS or ALP for degradation or to intracellular inclusions. A number of genes, which encode proteins that are involved in the function of the UPS or are degraded by the UPS, are mutated in inherited forms of PD. For example, mutations in the PARKIN gene, which encodes the E3 ubiquitin ligase parkin containing an N-terminal UBL-domain, are known to cause autosomal recessive juvenile parkinsonism [41]. Furthermore, wild-type parkin protein is found in Lewy bodies in sporadic PD [67]. UCHL1, a ubiquitin C-terminal hydrolase L1, is a deubiquitinating enzyme that generates free monomeric ubiquitin from polyubiquitin chains, and is also associated with PD [68]. In addition, α-synuclein, parkin, synphilin (polyubiquitinated by parkin), and mutated DJ-1 proteins, all implicated in PD pathogenesis, are substrates for UPS-mediated degradation [69]. Interestingly, the PD-linked mutation in DJ-1 disrupts the correct folding of DJ-1 protein. As a result, a misfolded, aggregation-prone protein is generated [70].

Ataxin-3 is a deubiquitinating enzyme harboring UIM domains that bind polyubiquitinated proteins. Wild-type ataxin-3 is found in the intranuclear inclusions in spinocerebellar ataxia (SCA) [71]. Furthermore, ataxin-3 gene has been reported to contain mutations, which associate with a specific type of SCA [72].

HDAC6 is a histone deacetylase that links polyubiquitinated proteins to the dynein motor complex for transport along the microtubules. HDAC6 has been shown to localize in aggresomes in vitro and the Lewy bodies in PD brain [73,74]. Furthermore, HDAC6 has been reported to control autophagosome-lysosome fusion, suggesting that it is involved in the regulation of the ALP [75] Interestingly, parkin-mediated K63-linked polyubiquitination targets misfolded DJ-1 to aggresomes via binding to HDAC6 [76], indicating that specific disease-associated proteins may interact with each other and are able to affect each other’s functions. This further suggests that defect in the function of one of these proteins may disrupt the function of its interacting partners and thus potentially amplify neurodegeneration.

p62/SQSTM1 is a major cargo receptor for autophagic degradation of ubiquitinated targets. p62/SQSTM1 binds to Atg8 on the autophagosomes and is itself a substrate for ALP-mediated degradation [77,78]. In addition to many protein-protein interaction domains, it contains an UBA domain in its C-terminus through which it binds polyubiquitinated proteins and targets them for sequestration or degradation [79,80]. p62/SQSTM1 is found present in neuronal and glial inclusions in AD and inclusions in Pick’s disease and in synucleinopathies, e.g. PD, dementia with Lewy bodies and multiple system atrophy [81]. The early accumulation of p62/SQSTM1 in NFTs in AD brain suggests that p62/SQSTM1 may be involved in the formation of NFTs [82]. Also, p62/SQSTM1 colocalizes with TDP-43 and with or without ubiquitin in neuronal and glial inclusions in frontotemporal lobar degeneration (FTLD) and in polyQ-containing inclusions in SCAs [83,84]. These facts together with the large number of proteins that p62/SQSTM1 interacts with suggests that it is a multifunctional protein that likely regulates a variety of physiological and pathophysiological functions [85]. Another disease-associated protein that is structurally related to p62/SQSTM1 and has also been reported to regulate the levels and targeting of many proteins is ubiquilin-1 (see below).

Ubiquilin-1

Ubiquilin-1 (also known as PLIC-1) is an AD-associated protein, which belongs to the highly conserved group of ubiquitin-like proteins that deliver polyubiquitinated proteins to UPS for degradation [86-88]. We have previously demonstrated an association between sporadic AD and genetic variation of UBQLN1, the gene encoding ubiquilin-1 [89] and shown that ubiquilin-1 protein regulates APP processing and β-amyloid production [90,91], suggesting that ubiquilin-1 may influence the AD pathogenesis at both genetic and mechanistic levels. Interestingly, mutations in a homologous gene, UBQLN2, were recently found to cause fALS and ALS/dementia [92], further implying a role for ubiquilin family proteins in neurodegenerative diseases. The full-length ubiquilin-1 is encoded by eleven exons. These give rise to specific domains in the ubiquilin-1 protein, such as the characteristic UBL domain in the N-terminus and the UBA domain in the C-terminus (Figure 3) [86,87,93]. The central region of ubiquilin-1 consists of conserved asparagine- and proline-rich repeats that mediate the interaction of ubiquilin-1 with specific domains of other proteins [87]. We have identified four alternatively spliced ubiquilin-1 transcript variants (TVs) in human brain [89,94] that encode different forms of ubiquilin-1 protein having different domains. In contrast to the full-length ubiquilin-1 TV1, TV2 variant lacks the exon 8, whereas TV3 lacks exons 2, 3 and 4 and therefore has an incomplete UBL domain. The smallest isoform, TV4 has only exons 1, 2 and 3, and thus is devoid of the UBA-domain. TV4 also contains a unique 32-amino acid insertion in its C-terminus (Figure 3). It is possible that the different ubiquilin-1 transcript variants have different functions.

Figure 3.

Ubiquilin-1 transcript variants (TV). The full-length ubiquilin-1 TV1 is encoded by 11 exons. Ubiquilin-1 has two signature domains of ubiquitin-like proteins: The N-terminal UBL (ubiquitin-like) domain (orange) and the C-terminal UBA (ubiquitin-associated) domain (green). UBA binds to polyubiquitinated proteins and UBL mediates interaction with the proteasome. The central region consists of conserved asparagine- and proline-rich repeats that mediate ubiquilin-1 interaction with other proteins. TV2 lacks exon 8 (light blue). TV3 lacks exons 2, 3, and 4 and therefore has an incomplete UBL domain. TV4 contains the first 3 exons and a unique short C-terminus (grey) due to a frame shift leading to a 32-amino acid insertion after the exon 3/5 junction.

Ubiquilin-1 is ubiquitously expressed in different tissues, such as brain, liver, kidney, heart and skeletal muscle. In cells, ubiquilin-1 localizes in the cytoplasm and to a lesser extent, in the nucleus and periphery of the cells [87] Staining of human brain has revealed the presence of ubiquilin-1 in neurons [87]. Ubiquilin-1 interacts specifically with a variety of cytosolic and transmembrane proteins, including γ-aminobutyric acid (GABA) and nicotinic acetylcholine receptors, G-proteins, and CD47, via its functional domains [95]. Moreover, many ubiquilin-1-interacting proteins have been implicated in the pathogenesis of neurodegenerative disorders, suggesting that ubiquilin-1 may regulate many physiological and pathophysiological events.

Ubiquilin-1 and UPS and ALP

Ubiquilin-1 appears to play a versatile role in regulating protein levels and subcellular targeting under different stress conditions that are centrally associated with the pathogenesis of neurodegenerative diseases. Ubiquilin-1 mediates the proteasomal targeting of misfolded or accumulated proteins by binding to their polyubiquitin chains with its UBA domain and by directly interacting with the S5a-domain of the 19S proteasomal subunit through its UBL domain [86-88] (Figure 3). These observations suggest that ubiquilin-1 functions as a shuttle protein between the proteins targeted for degradation and the proteasome. Our previous studies suggest that under excessive protein accumulation, specific ubiquilin-1 TVs may promote targeting of the accumulated proteins to both UPS and aggresomes [96,97] (see below). Furthermore, in the same study, we observed that ubiquilin-1 TVs are present in autophagosomes, suggesting that ubiquilin-1 may regulate the ALP [96]. Other studies have also reported that ubiquilin-1 colocalizes with the autophagosome marker LC3, associates with autophagosomes most probably through its UBA domain, and regulates ALP-mediated degradation of cellular cargo [98,99]. Additionally, ubiquilin-1 regulates recycling of nutrients and protects cells from apoptosis by enhancing the maturation of autophagosomes during nutrient starvation [98]. In contrast, depletion of ubiquilin-1 has been shown to inhibit autophagosome formation [99]. Ubiquilin-1 has been shown to bind the protein kinase mammalian target of rapamycin (mTOR), which is a major inhibitor of autophagy, but the effects of this interaction on autophagy are currently not known [100]. Ubiquilin-1 itself has been suggested to be a substrate for CMA [99].

Several studies indicate that ubiquilin-1 localizes in the ER. ER is an organelle, which is essential for protein folding and for redirecting undesired proteins to degradation in the UPS or APL pathways [94,101,102]. Disturbances in the ER-associated degradation system (ERAD) cause ER-stress and activate the unfolded protein response (UPR) in cells. These mechanisms have been suggested to be involved in the pathogenesis of many neurodegenerative diseases [103]. We and others have shown that ubiquilin-1 levels are up-regulated in cells by ER-stress and that ubiquilin-1 alleviates ER-stress and subsequently increases cell viability [94,101]. In accordance with this, loss of ubiquilin-1 in Caenorhabditis elegans increases the accumulation of misfolded proteins in the ER, activates ER-stress, and shortens the lifespan of the worms [102].

Ubiquilin-1 and accumulation and targeting of disease-associated proteins

Ubiquilin-1 has been shown to colocalize with NFTs in AD brain and Lewy bodies in PD brain [87]. Moreover, ubiquilin-1 interacts with and regulates many proteins involved in the pathogenesis of AD and other neurodegenerative diseases. Ubiquilin-1 was first identified as a presenilin (PS)-interacting protein [87,91]. PS1 and PS2 are essential catalytic components of the γ-secretase complex, which generates β-amyloid by proteolytic cleavage of APP. Ubiquilin-1 was shown to specifically increase the accumulation of full-length PS to form ubiquitinated high-molecular-weight (HMW) complexes [87]. Since that, the role of ubiquilin-1 as a PS1- and PS2-stabilizing protein has been confirmed by us and others [96,104,105]. Massey et al. [104] reported that ubiquilin-1 reduces the degradation of HMW-PS2 and colocalizes with PS2 in the aggresomes under proteasomal inhibition. We have recently shown that specific ubiquilin-1 variants, TV1 and TV3, regulate HMW-PS1 formation and targeting to aggresomes [96]. The increased aggresome formation was not associated with UPS impairment in our study, suggesting that ubiquilin-1 does not globally inhibit proteasomal activity [96]. The potential functional consequences of ubiquilin-1-induced accumulation and aggresomal targeting of PS1 remain to be resolved in future studies.

In addition to AD-related proteins, ubiquilin-1 is implicated in the regulation of other neurodegenerative disease-associated proteins. Heir et al. reported that the UBL domain of ubiquilin-1 is required for aggresomal targeting of aggregated proteins containing polyQ expansions [106]. Furthermore, ubiquilin-1 has been shown to regulate the aggregation and suppress the toxicity of polyQ proteins involved in HD in several studies [106-109]. Recently, ubiquilin-1 was reported to colocalize in intracellular inclusions with TDP-43, a protein which is a major component of ubiquitin-positive cytosolic inclusions in patients with ALS and ubiquitin-positive frontotemporal lobular dementia (FTLD-U). In these studies, the UBA domain of ubiquilin-1 was shown to mediate the stability and toxicity of the TDP-43 aggregates [110,111].

Taken together, the current data show that ubiquilin-1 interacts through its functional domains with a number of proteins and therefore regulates a variety of physiological and pathophysiological functions. Ubiquilin-1 also may play a crucial role in dictating the pathway to which specific proteins are targeted for degradation, especially under different stress conditions. The lack of UBL or UBA domains from specific ubiquilin-1 TVs suggests that different ubiquilin-1 TVs may have differential effects on the regulation of protein degradation pathways. As ubiquilin-1 participates in the aggregation, deposition, and degradation of several abnormally accumulated proteins in neurodegenerative diseases, it might represent a common mechanistic link between distinct neurodegenerative diseases.

Concluding remarks

Mounting evidence suggests that the pathogenesis of different neurodegenerative diseases centrally involves deficits in the PQC systems, which lead to the pathogenic accumulation and aggregation of proteins. The deficits in the PQC together with aging and other factors involved in the pathogenic mechanisms underlying neurodegenerative disorders, such as inflammation and oxidative or metabolic stress and pathogenic disease-associated mutations, play an important role in determining the onset and progression of the disease and finally causing widespread neurodegeneration in specific brain regions. Many of the disease-associated proteins also interact with each other and a number of other binding partners that are involved in important physiological functions, which may further aggravate the pathogenic events during disease pathogenesis. Therefore, characterization of the specific interactions of the disease-associated proteins and identification of factors regulating the PQC systems may help to recognize common molecular mechanisms between different neurodegenerative diseases. This may provide novel opportunities to better understand the disease pathogenesis and subsequently to identify new disease biomarkers and therapeutic targets for an earlier diagnosis and treatment of patients suffering from different neurodegenerative disorders.

Declaration of conflicts of interest

The authors do not have any conflicts of interest to declare.