Compromised autophagy and neurodegenerative diseases.

REVIEWS

Compromised autophagy and neurodegenerative diseases Fiona M. Menzies1, Angeleen Fleming1,2 and David C. Rubinsztein1

Abstract | Most neurodegenerative diseases that afflict humans are associated with the intracytoplasmic deposition of aggregate-prone proteins in neurons and with mitochondrial dysfunction. Autophagy is a powerful process for removing such proteins and for maintaining mitochondrial homeostasis. Over recent years, evidence has accumulated to demonstrate that upregulation of autophagy may protect against neurodegeneration. However, autophagy dysfunction has also been implicated in the pathogenesis of various diseases. This Review summarizes the progress that has been made in our understanding of how perturbations in autophagy are linked with neurodegenerative diseases and the potential therapeutic strategies resulting from the modulation of this process. Tauopathies A group of neurodegenerative diseases that are characterized by the prominent accumulation of tau protein in the CNS.

Ubiquitin–proteasome system A cellular protein degradation system in which proteins tagged with ubiquitin, a small regulatory protein that is covalently attached to a target protein, are broken down by an enzyme complex called the proteasome.

Department of Medical Genetics, University of Cambridge, Cambridge Institute for Medical Research, Addenbrooke’s Hospital, Hills Road, Cambridge CB2 0XY, UK. 2 Department of Physiology, Development and Neuroscience, University of Cambridge, Downing Street, Cambridge CB2 3EG, UK. Correspondence to D.C.R. e‑mail: dcr1000@hermes. cam.ac.uk doi:10.1038/nrn3961 1

Intracytoplasmic protein misfolding and aggregation are features of many late-onset neurodegenerative diseases called proteinopathies. These proteinopathies include Alzheimer disease (AD), Parkinson disease (PD), tauopathies and the neurodegenerative diseases caused by (CAG)n trinucleotide tract expansions that encode abnormally long polyglutamine (polyQ) tracts, such as Huntington disease (HD). In many of these diseases, the proteins that accumulate are thought to be toxic. This concept is supported by overexpression mouse models of HD, the occurrence of autosomal dominant tauopathies that are caused by mutations in the gene encoding tau, and PD that is caused by triplication of the α‑synuclein (SNCA) locus (reviewed in REFS 1,2). Accordingly, reductions in the steady-state levels of these proteins may be beneficial in targeting the root cause of these diseases. Recent data suggest that the lifetime of mutant huntingtin (mHTT), the protein that accumulates in and causes HD, may be a superior predictor of neuronal toxicity compared with its steady-state levels3, supporting strategies that aim to enhance the clearance of such proteins. Although inducing clearance of intracytoplasmic aggregate-prone proteins may represent a therapeutic strategy, compromised clearance may exacerbate or contribute to disease by increasing levels of key substrates such as aggregate-prone proteins4,5 and dysfunctional mitochondria6; enhancing susceptibility to cell death7,8; and perturbing flux through the ubiquitin–proteasome system 9. The genomics era has revealed many genes associated with neurodegenerative diseases, either as causative mutations or as risk factors for disease.

Investigations into the function of these genes have identified multiple intersections with the autophagy pathway. In this Review, we examine the role of autophagy — in particular, macroautophagy — in neurodegeneration. We begin by briefly reviewing the core aspects of autophagy biology, before focusing on studies from the past 2 years that have revealed the diversity of autophagy components that can be hijacked by disease-causing lesions and how this process may be manipulated to protect against neurodegeneration.

The autophagy machinery Autophagy, literally meaning ‘self-eating’, is an intracellular degradation pathway that is responsible for the digestion and recycling of nutrients via autophagosomes. These intracellular double-membraned structures engulf cytoplasmic proteins and organelles and deliver them to the lysosome for degradation (FIG. 1). The process is essential for cell survival during nutrient starvation, as it provides cellular energy 10. Autophagy is also a major pathway for the degradation of intracellular organelles and aggregate-prone proteins11. Through both the recycling of nutrients and the degradation of protein oligomers and damaged organelles, autophagy serves a vital role in maintaining homeostasis within the cell. Initiation of autophagy. The initial step in the formation of autophagosomes is the fusion of vesicles that have been proposed to arise from a variety of membrane sources, including plasma membrane-derived endosomal intermediates12, the endoplasmic reticulum (ER)13,14,

NATURE REVIEWS | NEUROSCIENCE

VOLUME 16 | JUNE 2015 | 345 © 2015 Macmillan Publishers Limited. All rights reserved

REVIEWS H+ Traﬃcking

Degradation Cystatins (LSDs) and glucocerebrosidase (LSDs and PD)

ATG5–ATG12–ATG16L LC3-II Receptor Recycling Spastizin and spatacsin (HSP)

Autophagic cargo

Initiation Alsin (ALS), ATG7 (HD) and spastizin (HSP)

Precursor vesicle

pH ATP13A2 (PD) and PS1 (AD)

Pre-autophagosomal structures

Biogenesis TFEB (multiple)

Autophagosome

Lysosome

Autolysosome

VPS34 Beclin 1 Precursor formation VPS35 and α-synuclein (PD), PICALM (AD) and WIPI4 (BPAN)

Receptors p62 and optineurin (ALS) and HTT (HD)

Maturation of autophagosomes and lysosomes PICALM (AD), CHMP2B (CMT2) and SIGMAR1 (ALS)

Figure 1 | Overview of the autophagy pathway and the sites of action of disease-associated proteins. This diagram Nature Reviews | Neuroscience shows a simplified version of autophagy. Initiation of autophagy is signalled via the activity of the vacuolar protein sorting 34 (VPS34) complex. Precursor vesicles fuse to form pre-autophagosomal structures that grow to eventually become double-membraned autophagosomes. Substrates for degradation by autophagy are engulfed by these growing membranes or may be sequestered into the forming structure by receptor proteins. The completed autophagosomes are then trafficked to fuse with lysosomes. The acidic environment inside the lysosomes is maintained by ATPases. This low pH is required for the correct function of the lysosomal degradative enzymes (depicted as scissors) and, therefore, the breakdown of the autophagy substrates. Perturbations throughout the pathway, from initiation of autophagosome formation to degradation in the autolysosomes, have been suggested to be involved in neurodegenerative diseases; some disease-associated proteins function at multiple points in the process. The key points in the pathway and the selected disease-associated proteins that are thought to act at these points are highlighted in boxes; the relevant disease is indicated in parentheses. AD, Alzheimer disease; ALS, amyotrophic lateral sclerosis; ATG, autophagy protein; ATP13A2, ATPase type 13A2; BPAN, β‑propeller protein-associated neurodegeneration; CHMP2B, charged multivesicular body protein 2B; CMT2, Charcot–Marie–Tooth disease type 2; HD, Huntington disease; HSP, hereditary spastic paraplegia; HTT, huntingtin; LC3, microtubule-associated protein 1 light chain 3; LSD, lysosomal storage disease; PD, Parkinson disease; PICALM, phosphatidylinositol-binding clathrin assembly protein; PS1, presenilin 1; SIGMAR1, sigma non-opioid intracellular receptor 1; TFEB, transcription factor EB; WIPI4, WD repeat domain phosphoinositide-interacting protein 4.

Risk factors Factors (in this article, alterations in DNA sequences) that may increase the chance of developing a disease.

Lysosome An intracellular membrane-bound organelle containing hydrolytic enzymes that are capable of breaking down proteins and other cellular components.

the Golgi15,16 and mitochondria17. These vesicles coalesce to form a flattened membrane sac called a phagophore. The fusion of additional vesicles results in the formation of a cup-shaped double membrane that surrounds and eventually engulfs portions of cytoplasm as the two edges of the structure come together and fuse (FIG. 1). Various autophagy (ATG) proteins are essential for the formation of autophagosomes. The vesicles that give rise to the phagophore contain a complex of ATG5–ATG12–ATG16L1, the formation of which requires the catalytic activities of ATG7 and ATG10. In addition, the phagophore membrane contains ATG9, the only multipass transmembrane protein in the ATG family. Recently, it was shown that ATG9- and ATG16L1‑containing vesicles arise independently via clathrin-mediated endocytosis but subsequently fuse in the recycling endosome18. A second enzyme cleavage and conjugation pathway that is essential to autophagosome biogenesis involves the processing of microtubule-associated protein 1 light chain 3 (LC3). LC3 is a protein that has been proposed to recruit membrane to the developing phagophore. LC3

is first cleaved by ATG4B to form LC3‑I, which is then conjugated to phosphatidylethanolamine by ATG7 and ATG3 to form LC3‑II19. As the phagophore enlarges and approaches closure, the ATG5–ATG12–ATG16L1 complex dissociates from the outer membrane, whereas LC3‑II remains associated with the completed autophagosome. Following closure, autophagosomes are trafficked by dynein motors along microtubules20 to the perinuclear region where they fuse with the lysosome and their contents are degraded. SNAREs (soluble NSF attachment protein receptors) have a critical role in autophagosome formation and degradation. The primary function of these proteins is to facilitate vesicle fusion, and different members of the SNARE family are critical for various steps of the process. For example, vesicle-associated membrane protein 3 (VAMP3)18 is important for ATG9–ATG16L1 vesicle fusion; VAMP7, syntaxin 7 and syntaxin 8 (REF. 21) are important for phagophore elongation; and VAMP7, VAMP8 and VTI1B (vesicle transport through interaction with t‑SNAREs homologue 1B)22,23 are involved in autophagosome–lysosome fusion.

346 | JUNE 2015 | VOLUME 16

www.nature.com/reviews/neuro © 2015 Macmillan Publishers Limited. All rights reserved

REVIEWS Control over the initiation of autophagy is mediated through a protein complex comprising ULK1 or ULK2, ATG13, ATG101 and the focal adhesion kinase family interacting protein of 200 kDa (FIP200)24. This ULK1/ ULK2–ATG13–FIP200 complex serves as both an initiator of autophagy and a sensor for upstream signalling through three major pathways: the mTOR complex 1 (mTORC1), adenosine monophosphate (AMP)-activated protein kinase (AMPK) and p53 pathways24,25. Phosphorylation of ULK1 or ULK2 and ATG13 by mTORC1 inactivates this complex, whereas AMPK activates autophagy through the direct phosphorylation of ULK1 at sites that are distinct from those targeted by mTORC1. p53 can also control autophagy through transcriptional regulation. Activated p53 transcriptionally upregulates pro-autophagic proteins such as ULK1, ULK2, AMPK and tuberous sclerosis 2 protein (TSC2)25, whereas transcriptionally inactive, cytoplasmic p53 negatively regulates autophagy through direct binding with FIP200 (REF. 26), which may block formation of the ULK complex. The ULK1/ULK2–ATG13–FIP200 complex controls the initiation of autophagy through the PI3K complex, which comprises the vacuolar protein sorting 34 (VPS34; also known as PIK3C3) and its accessory partners VPS15 (also known as PIK3R4), beclin 1 and ATG14L (also known as barkor). Activation of ULK1 results in the phosphorylation of beclin 1 and a subsequent increase in the activity of the VPS34 complex 27. This complex is required for the formation of phosphatidylinositol 3‑phosphate (PtdIns3P) on nascent autophagosomes, which facilitates the recruitment of PtdIns3P‑binding proteins such as WD repeat domain phosphoinositide-interacting protein 1 (WIPI1) and WIPI2 (both of which are mammalian orthologues of yeast Atg18) to autophagosome membranes. The recruitment of ATG16L1 to these structures appears to occur through binding to WIPI proteins28. ULK1 also regulates the recruitment of the transmembrane protein ATG9 to the phagophore29,30. Recent data suggest that PtdIns5P can substitute for the functions of PtdIns3P in autophagosome biogenesis31.

Endocytic pathway A general term to describe the vesicle trafficking routes by which cells internalize molecules from the plasma membrane.

Autophagy receptors. Until recently, autophagy was thought to be a non-selective degradation pathway, but an emerging field of research has revealed a series of receptor proteins (which have sometimes been referred to as adaptor proteins; for clarification on terminology, see REF. 32) that confer substrate specificity on this process. These receptor proteins bind to a diverse range of cargoes including bacteria and viruses; however, more pertinently, they have also been shown to bind to aggregated proteins33–35. Receptor proteins — such as p62 (also known as sequestosome 1), next to BRCA1 gene 1 protein (NBR1), nuclear domain 10 protein 52 (NDP52; also known as CALCOCO2) and optineurin — are able to bind to their cargo selectively. For example, such proteins may recognize ubiquitylated proteins (the cargoes) via ubiquitin-binding domains and bind to LC3 family members on autophagosomes through their LC3‑interacting region motifs36 (FIG. 1). Selective targeting of autophagic cargoes can also be mediated by autophagy receptors that form a bridge between the cargo–autophagy

receptor complex and components of the autophagosome membrane (for example, ATG5 and PtdIns3P36) (FIG. 1). Autophagy-linked FYVE protein (ALFY) may function in this way. ALFY is localized to the nuclear envelope under basal conditions but redistributes to colocalize with ubiquitin-positive cytoplasmic structures upon proteasome inhibition, and with ATG5 or LC3‑positive cytoplasmic puncta under amino acid starvation conditions37. The FYVE domain of ALFY is able to bind to PtdIns3P, but it also contains additional domains, notably five WD40 repeats, which are required for its interaction with Atg5 (REF. 38), and a pleckstrin homology‑like domain and a BEACH domain, which together interact with p62 (REF. 39). These interactions suggest that ALFY has a scaffolding role in the formation of p62‑labelled aggregates, which can then be targeted to autophagosomes. Additional autophagy proteins have been implicated in the selective degradation of mitochondria (for example, BNIP3L (BCL‑2/adenovirus E1B 1 19 kDa protein-interacting protein 3‑like), Atg32 and p62) and peroxisomes (for example, Atg30 and p62). Although Atg32 and Atg30 have only been characterized in yeast and it is unclear whether mammalian homologues of these proteins exist, BNIP3L, a mitochondrial membrane protein containing a LC3‑interacting region motif, has been shown in mammalian cells to play a part in the targeted clearance of damaged mitochondria40,41 (see BOX 1 and FIG. 2 for further details on mitophagy and neurodegeneration). Autophagosome trafficking and degradation. The maturation of autophagosomes and the final steps of the autophagic pathway have been the subject of a recent expert review 42. Once formed, autophagosomes are transported along microtubules to the lysosome-rich perinuclear region. In neurons, autophagosomes formed at the distal tip of axons are transported towards the cell body by dynein-mediated retrograde transport 43. Autophagosomes may fuse with vesicles from the endocytic pathway (early or late endosomes) to form amphisomes, which subsequently fuse with lysosomes, or they may fuse directly with lysosomes. This final step of the autophagy process is mediated by VAMP7, VAMP8 and VTI1B, members of the SNARE family of proteins (reviewed in REF. 44). Efficient degradation of autophagy substrates can only occur if the acidic lysosomal pH is correctly maintained and the degrading enzymes are functioning correctly. Sufficient lysosomes must also be available for fusion with autophagosomes because this process consumes the lysosomes. They can be regenerated through lysosomal reformation, in which tubules formed on auto lysosomes mature into new lysosomes45. Transcription factor EB (TFEB) has been identified as a master regulator for lysosomal biogenesis46; as such, this protein has an important role in the control of autophagy 47.

Autophagy failure in neurodegeneration The vast majority of neurodegenerative diseases, including sporadic forms of such diseases, have a genetic component, and the assessment of the cellular functions of these genes has implicated autophagic dysfunction in their pathogenesis. These findings have



REVIEWS Box 1 | Mitophagy in neurodegenerative disease Mitochondria have long been studied as key organelles in the pathogenesis of many neurodegenerative diseases, particularly Parkinson disease (PD). Recently, much of this research has focused on the potential contribution of altered clearance of dysfunctional mitochondria by autophagy (termed mitophagy). This research has centred on two genes that are the major causes of autosomal recessive PD, namely parkin (PARK2) and phosphatase and tensin homologue-induced putative kinase 1 (PINK1; also known as PARK6), which have been shown in Drosophila melanogaster to act via the same pathway146–148. In functional mitochondria, which are able to maintain their mitochondrial membrane potential, PINK1 is imported via the translocase of the outer membrane (TOM) complex and the translocase of the inner membrane (TIM23) complex (FIG. 2). It is then cleaved by mitochondrial processing peptide (MPP) and presenilins-associated rhomboid-like protein (PARL) before being translocated back to the cytoplasm for degradation by the proteasome149–151. However, in the absence of a mitochondrial membrane potential, PINK1 accumulates on the mitochondrial outer membrane where it recruits parkin152–155, resulting in the targeting of mitochondria for degradation by autophagy. The exact mechanism by which mitophagy is promoted downstream of parkin recruitment has yet to be established (reviewed in REF. 156). A major hypothesis for the mechanism is that parkin ubiquitylates a number of mitochondrial outer membrane proteins, such as voltage-dependent anion-selective channel protein 1 (VDAC1), TOM20 and mitochondrial RHO GTPase 1 (MIRO1)152,157, which then recruit p62 to the mitochondrial outer membrane where autophagsomes subsequently form, although this role for VDAC1 has been disputed158. PINK1 and parkin are also reported to have roles in mitochondrial fission and fusion, as well as in mitochondrial trafficking (reviewed in REF. 159). These roles could represent either mechanisms by which PINK1 and parkin act to ultimately control mitophagy or independent processes that affect mitochondrial health and function. Investigations into the role for the PINK1–parkin pathway in mitophagy have mostly been carried out in non-neuronal cultured cells overexpressing parkin and treated with carbonyl cyanide m‑chlorophenylhydrazone (CCCP) to induce mitochondrial

Polymorphisms The existence of multiple variants of a DNA sequence that occur within different individuals in the population, with no sequence being regarded as the standard sequence.

depolarization, and it is not clear how relevant this process is to disease progression. However, recent work has shown that endogenous parkin enables mitochondrial clearance157. This is further supported by work using mt‑KR, a photoactivatable protein that produces local increases in reactive oxygen species. Using this construct to damage mitochondria in the distal axons of cultured neurons, it was observed that the PINK1–parkin pathway induced autophagy of these damaged mitochondria in the axons160. An alternative mechanism controlling mitophagy is the externalization of cardiolipin, a phospholipid that is normally only present on the mitochondrial inner membrane. Cardiolipin is able to bind to microtubuleassociated protein 1 light chain 3 (LC3)161, thereby promoting autophagy. However, this effect has only been observed in cells treated with mitochondrial toxins such as rotenone and 6‑hydroxydopamine, which did not cause the same increase in PINK1 levels that are seen after CCCP treatment. This suggests that different causes of mitochondrial damage may result in mitophagy via different mechanisms. When considering the role of mitophagy in PD, it must be asked how mutations in PINK1 or PARK2, which account for only a small proportion of PD cases, relate to other forms of the disease. A recent study has addressed this, suggesting that sterol regulatory element-binding transcription factor 1 (SREBF1), a gene identified by a genome-wide association study as a risk factor for PD162, has a role in parkin recruitment to damaged mitochondria and mitophagy163 and could thus represent a link between mitophagy and sporadic PD. Mitophagy might have further disease relevance because xeroderma pigmentosum group A, a DNA repair disorder associated with neurodegeneration, has also been linked to defective mitophagy that is associated with increased cleavage of PINK1 (REF. 164). Importantly, impaired autophagy flux will also affect mitophagy and perturb the clearance of dysfunctional mitochondria, which may increase the cellular levels of reactive oxygen species and the propensity to caspase activation. Although mitophagy is not the focus of this Review (and has been the subject of a recent excellent review156), these consequences are important to consider as pathogenic mechanisms in the diseases in which autophagy is compromised.

been complemented by studies showing that autophagy markers are perturbed in post-mortem tissue in many neurodegenerative diseases. However, in many cases, the molecular mechanisms have yet to be determined, and it remains to be ascertained whether this truly represents dysfunctional autophagy or whether compromised autophagy contributes to disease progression. Here, we do not discuss the evidence for altered autophagy disease-by‑disease as we have done previously48 but, rather, consider how recent research has revealed a number of key points in the autophagic pathway that can be affected by disease-causing mutations. To date, there has only been one disease — β‑propeller protein-associated neurodegeneration (BPAN; also known as neuro degeneration with brain iron accumulation 5 (see below)) — in which the causal mutation occurs in a gene that is proposed to function solely in autophagy (although other functions of this gene may emerge with further research). Indeed, most of the proteins discussed in this Review have functions beyond a role in autophagy. Many of these functions may be pertinent to the pathogenesis of neurodegenerative disease but are, for the most part, not discussed here. We do not suggest that autophagy dysfunction is the single mechanism underlying the pathogenesis of

neurodegenerative diseases. Instead, we aim to highlight it as a common aspect of these complex diseases and discuss the new information coming to light about the intersection of autophagy and neurodegeneration. Mutations in core autophagy genes. Loss-of-function mutations have been identified in the gene encoding the β‑propeller scaffold protein WIPI4 (WDR45) in BPAN49. WDR45 is a core autophagy gene, encoding one of four homologues of yeast Atg18, involved in autophagosome formation. Parents or siblings of patients with WDR45 mutations do not share the mutation, suggesting that such mutations occur de novo50. Lymphoblastoid cell lines derived from these patients have reduced autophagic activity 50, which may result from disrupted ATG9A vesicle translocation at the autophagosome formation site owing to altered WIPI4 activity 51. Given the importance of autophagy in cellular function, mutations in core autophagy genes are likely to have severe consequences. Thus, it is perhaps unlikely that late-onset neurodegenerative diseases will be caused by recessive mutations in autophagy genes, and it is more likely that polymorphisms in these genes contribute to the age of onset and disease progression of such disorders.

348 | JUNE 2015 | VOLUME 16


REVIEWS PINK1 Proteasome TOM

Mitochondrion

TIM23 MPP

PARL Cardiolipin

H+

MIRO1 VDAC1 Parkin

Phagopore

Ubiquitin LC3-II ATG5–ATG12–ATG16L

Figure 2 | Initiation signals for mitophagy. In mitochondria that are able to maintain Nature Reviews | Neuroscience their membrane potential (as shown in the upper half of the diagram), phosphatase and tensin homologue-induced putative kinase 1 (PINK1) is imported across the mitochondrial membrane by the translocase of the outer membrane (TOM) complex and the translocase of the inner membrane (TIM23) complex before being cleaved by mitochondrial processing peptidase (MPP) and presenilins-associated rhomboid-like protein (PARL) and then targeted for degradation by the proteasome. In the absence of a membrane potential (lower half of the diagram), PINK1 cannot be transported across the membrane and instead recruits parkin to the membrane, resulting in the ubiquitylation of other mitochondrial outer membrane proteins (such as voltage-dependent anion-selective channel protein 1 (VDAC1) and mitochondrial RHO GTPase 1 (MIRO1)), phagophore formation and subsequent degradation. Alternatively, externalization of cardiolipin onto the outer membrane following mitochondrial depolarization has also been reported to result in phagophore formation. ATG, autophagy protein; LC3, microtubule-associated protein 1 light chain 3.

Support for this assertion can be found in HD. The age of onset of HD is closely correlated with the polyQ repeat length, which accounts for 70% of the variance; however, other genetic factors have been proposed to contribute to age of onset. One such gene is ATG7: the V471A polymorphism has been associated with an earlier onset of disease by 4–6 years in certain populations52,53. However, it has yet to be demonstrated whether this polymorphism has functional importance in terms of autophagic activity. Beclin 1 interactions. As noted above, beclin 1 is req uired for the early steps of autophagosome formation. Neurodegeneration-associated proteins are now coming to light that may affect the interactions of beclin 1 with its protein partners and influence its ability to mediate autophagy. Overexpression of the E46K variant of α‑synuclein, which causes a rare autosomal dominant form of PD, leads to decreased levels of phosphorylated BCL‑2, which then enhances the interaction between beclin 1 and BCL‑2 and therefore results in autophagy inhibition54. Regulation of the beclin 1–BCL‑2 interaction has also been suggested to be compromised in HD. HTT has been demonstrated to interact with Ras homologue enriched in striatum (RHES), which is selectively expressed in the striatum, where HD pathology

is observed55. In turn, RHES has been demonstrated to enhance autophagy via its interaction with beclin 1, blocking the inhibitory interaction of beclin 1 with BCL‑2. Importantly, mHTT prevents this autophagy activation56. Beclin 1 is also involved in the maturation of autophagosomes through its role in the ultraviolet radiation resistance-associated gene protein (UVRAG)– rubicon–beclin 1 complex. Beclin 1 also interacts with spastizin57 (encoded by ZFYVE26, which is mutated in hereditary spastic paraplegia (HSP) type 15), and disease-associated mutations disrupt this interaction58. Autophagosomes accumulate in cells lacking spastizin or expressing mutant forms of this protein58. This accumulation has been proposed to result from loss of function of the UVRAG–rubicon–beclin 1 complex 58 but may also be due to defects in lysosomal function (see below). Mutations in genes involved in trafficking. Membrane trafficking is a vital element of autophagy, and the identification of proteins involved in such trafficking is currently an area of intensive research. Many neuro degenerative disease-causing mutations have been identified in genes that can be classed as trafficking genes, and an attractive hypothesis for the pathogenic mechanism of late-onset neurodegenerative diseases is that the mutations are likely to impair autophagy efficiency. Phosphatidylinositol-binding clathrin assembly protein (PICALM) was identified through genomewide association studies to be associated with AD59,60. Moreover, PICALM has been reported to be abnormally cleaved in the brains of patients with AD, resulting in a decreased level of the full-length protein61. PICALM is required for the endocytosis of VAMP2 and VAMP3, which are necessary for autophagosome formation, and VAMP8, which is important for autophagosome– lysosome fusion. In the absence of PICALM, decreased fusion of autophagosome precursor vesicles is observed (VAMP2- and VAMP3‑dependent processes), leading to impaired autophagosome biogenesis, along with a decrease in the fusion of autophagosomes with lyso somes (a VAMP8‑dependent process), resulting in an overall reduction in autophagic flux 62. These defects contribute to the accumulation of tau, an autophagy substrate and a probable contributor to AD pathology. The small GTPase RAB protein family comprises well-known regulators of membrane trafficking and fusion events. Mutations in RAB7, which encodes a late endosomal RAB protein known to be involved in autophagy (reviewed in REF. 63), have been reported to cause Charcot–Marie–Tooth type 2B disease 64. Similarly, the gene encoding alsin (ALS2) is mutated in some cases of recessive amyotrophic lateral sclerosis (ALS)65. Alsin is an activator of RAB5 (REF. 66), which is required for the clearance of aggregate-prone proteins by autophagy 67. Missense mutations in ALS2 result in mislocalization of the protein and a decreased fusion of autophagosomes with endosomes68. Hexanucleotiderepeat expansions in a non-coding region of chromosome 9 open reading frame 72 (C9ORF72) have been reported as a major genetic factor in patients with ALS



REVIEWS Multivesicular bodies A type of late endosome containing internal vesicles.

Lewy bodies Protein aggregates found within neurons in patients with certain neurodegenerative diseases, including Parkinson disease and Lewy body dementia.

and in patients with frontotemporal dementia (FTD). Recently, C9ORF72 has been suggested to have a role in endosomal trafficking, and it contains DENN (differentially expressed in normal and neoplastic cells) domains, which are normally associated with RABGEFs, activators of endocytic RAB proteins69. In neuronal cell lines, C9ORF72 colocalizes with several RAB proteins, and small interfering RNA (siRNA)mediated knockdown of C9ORF72 inhibited shiga toxin trafficking from the plasma membrane to the Golgi and endocytosis of tropomyosin-related kinase B (TRKB), as well as increased the levels of LC3‑II70. However, the importance of these mechanisms in the pathogenesis of ALS is unclear. This is because mutations in C9ORF72 are found in non-coding regions, and it is considered that repeat associated non-ATG-mediated translation (RAN) of arginine-rich dipeptides are the toxic entities in the case of this mutation71. Mutations in charged multivesicular body protein 2B (CHMP2B) can cause FTD72. CHMP2B is part of the ESCRT-III (endosomal secretory complex required for transport III) complex, and CHMP2B mutations result in disruption of this complex, causing defects in the endosome–lysosome pathway. Overexpression of mutant CHMP2B has been demonstrated to result in the accumulation of autophagosomes, suggesting that their maturation is blocked73. This finding is also consistent with the role of ESCRT-III in the formation of multivesicular bodies, which can fuse with autophagosomes before the fusion of autophagosomes with lysosomes. VPS35 is a component of retromer, which mediates endosome-to‑Golgi trafficking (reviewed in REF. 74). Retromer is also required for the endosomal recruitment of the actin nucleation-promoting WAS protein family homologue (WASH) complex that is required for the formation of actin patches and efficient protein sorting 75. Recently, the D620N substitution in VPS35 has been identified as causing autosomal dominant PD76, and this mutation impairs binding of the WASH complex to endosomes. As a consequence, ATG9 trafficking

Box 2 | Chaperone-mediated autophagy Chaperone-mediated autophagy (CMA) is a form of lysosomal degradation in which substrates are delivered directly to the lysosome rather than being trafficked via autophagosomes as occurs in macroautophagy. CMA substrates are recognized by a pentapeptide motif based on the charge of the amino acids, allowing a motif to also be generated by post-translational modifications. Proteins to be degraded are bound by the heat shock protein HSC70 and are targeted to the lysosome, where they interact with a multiprotein complex, which includes lysosomal-associated membrane glycoprotein 2 (LAMP2). They are unfolded and translocated directly across the lysosomal membrane. This process is not the focus of this Review and has been reviewed elsewhere165–167; however, it is important to note that CMA dysfunction has also been implicated in the pathogenesis of neurodegenerative diseases. CMA is able to degrade proteins associated with neurodegeneration, such as α-synuclein168, leucine-rich repeat serine/threonine-protein kinase 2 (LRRK2)169 and tau170, and disease-associated mutant versions of these proteins are less efficiently degraded by CMA and in fact block the degradation of other substrates by CMA168,169. It is also important to note that the defects in lysosomal function outlined in this Review are equally likely to affect CMA and macroautophagy and may thus contribute to neurodegeneration by perturbation in CMA.

is disrupted, resulting in the inhibition of the early stages of autophagosome formation and decreased α‑synuclein clearance75. The ATG9 mislocalization that results from VPS35-D620N is reminiscent of that seen to result from increased levels of α-synuclein77, the protein that accumulates in Lewy bodies in PD. Increased levels of α‑synuclein are sufficient to cause PD78, and this protein has been shown to inhibit RAB1A79, leading to an alteration of ATG9 localization and decreased autophagosome formation in cell culture and in vivo77. The identification of the autophagy defect caused by VPS35-D620N could be a mechanistic feature that is shared by other neurodegenerative diseases. For example, mutations in the gene encoding strumpellin (KIAA0196; a WASH complex subunit) have been identified in some cases of HSP80, and mutations in the gene encoding SWIP (also known as KIAA1033; another WASH complex subunit) have been associated with autosomal recessive intellectual disability 81. Moreover, another gene mutated in PD, DNAJC13 (also known as RME8)82, encodes a protein that associates with the WASH complex through binding to the same subunit as does VPS35 (REF. 80). Mutations in the 3ʹ untranslated region and the coding region of sigma non-opioid intracellular receptor 1 (SIGMAR1) can cause forms of ALS and FTD83,84, and levels of SIGMAR1 are decreased in the spinal cord of patients with sporadic ALS85. Loss of Sigmar1 in mice results in motor deficits86, and treatment with a SIGMAR1 agonist improves motor function and survival in a mutant superoxide dismutase 1 (SOD1) ALS mouse model87. SIGMAR1 is a chaperone that is located at the ER–mitochondria interface and has many functions, including a role in trafficking. Knockdown of SIGMAR1 impairs vesicle trafficking from the ER to the Golgi and results in a decrease in the fusion of autophagosomes with lysosomes and therefore a reduction in the degradation of autophagy substrates88. Lysosomal defects in neurodegenerative diseases. Lysosomal storage disorders (LSDs) are the most common neurodegenerative diseases of childhood and comprise more than 50 diseases. The clinical phenotypes of these diseases vary; however, they frequently show progressive CNS defects. Because the final step in the autophagic process is the fusion of autophagic vesicles with lysosomes and the degradation of the contents of these autolysosomes, lysosomal defects are highly likely to affect the autophagic capacity of the cell (reviewed in REF. 89), both in terms of macroautophagy, as we discuss here, and in terms of chaperone-mediated autophagy (BOX 2). Although the defects causing LSDs can vary widely from one type of LSD to another (LSDs may arise as a consequence of defects in lysosomal enzymes, lysosomal membrane-associated proteins or even nonlysosomal-associated enzymes), all cases studied to date show impaired autophagy, suggesting that this is the common pathogenic mechanism in these diseases. A LSD that has been widely studied in terms of its autophagic characteristics is Niemann–Pick disease type C1 (BOX 3).

350 | JUNE 2015 | VOLUME 16


REVIEWS Box 3 | Niemann–Pick disease type C1 — excessive or defective autophagy? The study of the role of autophagy in Niemann–Pick disease type C1 (NPC1) highlights not only the difficulties associated with interpreting the contribution of autophagy function and dysfunction to a disease but also the challenges of understanding the potential of autophagy modulation as a therapeutic approach. NPC1 is caused by mutations in NPC1, which result in the loss of correct trafficking of cholesterol within the cell and the accumulation of unesterified cholesterol and glycosphingolipids in lysosomes and late endosomes (which would normally be trafficked on to the Golgi and endoplasmic reticulum for further processing). This mislocalization disrupts both lysosomes and endosomes, as well as other membrane compartments, which do not receive the cholesterol they require for their correct membrane composition171. A role for autophagy in NPC1 was first suggested after autophagosomes were found to accumulate in degenerating Purkinje cells in mice with mutant NPC1 (REF. 172), and autophagy was initially thought to be upregulated in this disease (reviewed in REF. 171). However, with improved knowledge of the autophagic pathway and how to measure autophagy activity, such as the use of tandem red fluorescent protein–GFP-tagged microtubule-associated protein 1 light chain 3 (LC3)173, it has become clear that autophagy is defective in this disease, leading to the proposal that autophagy was both induced and defective174,175. These studies suggested that treatment with either cholesterol-depleting drugs (such as cyclodextrin) or autophagy inhibitors (such as 3‑methyladenine) may be beneficial. Cyclodextrin was suggested to decrease the accumulation of autophagosomes as it restored autophagic function175. However, more recent studies have argued against the evidence for any upregulation of autophagy in NPC1 (REFS 176–178), demonstrating that the increase in autophagic markers is solely a result of decreased maturation of autophagosomes. Furthermore, it was suggested that cyclodextrin treatment in fact inhibits autophagy. A proposed therapeutic strategy for this disease involved treatment with low-dose cyclodextrin to alleviate the cholesterol build‑up, along with autophagy inducers such as rapamycin177 or carbamazepine176.

The LSD Gaucher disease is caused by homozygous mutations in glucocerebrosidase (GBA). Heterozygosity for GBA mutations (Gaucher disease carrier status) is the most common known genetic risk factor associated with PD90. Loss of GBA results in lysosomal dysfunction owing to the accumulation of its substrate, glucocerebroside. Mice lacking Gba have defective autophagy and accumulation of potential autophagy substrates in the brain, such as p62, monomeric and oligomeric forms of α‑synuclein, and ubiquitylated proteins, along with decreased mitochondrial function and decreased mitophagy 91. Patients with sporadic PD have decreased levels of GBA in affected regions of the brain, along with increased levels of α-synuclein92,93. Lysosomal membrane permeabilization (LMP) may also cause lysosomal dysfunction in LSDs (reviewed in REF. 94). LMP leads to a release of cathepsins into the cytoplasm, which has been associated with calpain activation and apoptotic cell death, and may also lead to a block in autophagy. For example, in Niemann–Pick disease type A, which is caused by mutations in the gene encoding acid sphingomyelinase, increases in sphingomyelin levels lead to LMP and a concomitant decrease in autophagosome degradation95. Although there is clear evidence for a defect in autophagy in LSDs, the specific contribution of autophagy defects versus lysosomal dysfunction has not been parsed out. In addition to classical LSDs, it is now becoming evident that lysosomal dysfunction may play a part in other neurodegenerative diseases. Kufor–Rakeb syndrome and some early-onset forms of PD are associated with mutations in the gene encoding lysosomal ATPase type 13A2 (ATP13A2)96. Fibroblasts

derived from patients with ATP13A2 mutations show increased lysosomal pH and reduced processing of lysosomal proteases from their precursor to mature forms. This results in decreased protease activity 97 and an increase in the number of autophagic vesicles owing to a block in fusion with lysosomes. Mutations in presenilin 1 (PS1) in AD can also result in an increase in lysosomal pH owing to the role of the presenilin holoprotein (rather than the cleaved form that is active in the γ‑secretase complex) in the glycosylation of the lysosomal v‑ATPase subunit V0a1, which results in decreased autophagic turnover 98,99. Although the exact mechanism that leads to this decrease in turnover is unclear 100,101, there does seem to be a consensus regarding the idea that PS1 has a role in autophagic clearance98,99,102,103. TFEB enhances both lysosomal biogenesis and autophagosome formation46,47. Mutant forms of the androgen receptor, which cause the polyQ disorder spinobulbar muscular atrophy (SBMA; also known as Kennedy disease), interact with and inhibit TFEB, with consequent effects on the autophagy–lysosome pathway 104. A role for altered TFEB activity has also been suggested in PD because α‑synuclein has been reported to interact with TFEB; in dopaminergic neurons from patients with PD, TFEB seems to be sequestered in the cytoplasm (away from its target of action in the nucleus)105. In AD, enhanced levels of acid sphingomyelinase, one of a group of sphingomyelin-metabolizing enzymes, have been shown in fibroblasts, brain and plasma of patients106. This increase in acid sphingomyelinase causes a decrease in TFEB levels. The genetic reduction of acid sphingo myelinase in heterozygous knockout mice, or pharmacological inhibition of this enzyme with amitriptyline, protected against disease development in the APP/PS1 mouse model of AD, reducing the accumulation of LC3‑II and p62 that is observed in this model106. Finally, two genes in which mutations have been identified in autosomal recessive HSP are implicated in the regeneration of lysosomes. The genes encode spastizin (ZFYVE26) and spatacsin (SPG11), which act together in a complex that is required for the initiation of lysosomal reformation. Loss or mutation of either gene results in a decrease in lysosomal number and an accumulation of autolyososomes107. Autophagy receptors for aggregate-prone proteins. p62 and other autophagy receptors are components of neuronal protein aggregates in various neurodegenerative diseases33,108–110. p62 and LC3 were shown to colocalize within a shell formed around mHTT aggregates in vitro, and knockdown of p62 enhanced mHTT toxicity, providing the first evidence for its role in the selective clearance of mHTT aggregates33. Indeed, in Drosophila melanogaster models of neurodegeneration, loss of p62 exacerbates the degeneration of the eye caused by overexpression of mHTT, a truncated form of ataxin 3 containing an expanded polyQ repeat (MJDtr‑Q78; associated with another polyQ disease, spinocerebellar ataxia type 3) or TAR DNA-binding protein 43 (TDP43; associated with ALS)111. However, these observations have not translated simply into vertebrate models. In a mouse model



REVIEWS of SBMA in which a polyQ-expanded androgen receptor was overexpressed, knockout of p62 enhanced the disease phenotype, whereas p62 overexpression was protective112. However, p62 overexpression, rather than enhancing clearance of the mutant protein, converted it from a soluble to an insoluble form, thereby increasing the number of nuclear inclusions observed in the brain112. Furthermore, depletion of p62 in mouse models of HD has been shown to ameliorate rather than enhance disease signs113. Although the number of nuclear mHTT aggregates was decreased in p62-knockout mice expressing mHTT, there was an increase in the levels of cytoplasmic aggregates and no apparent change in total HTT levels in the brain113. Another autophagy receptor implicated in the clearance of mHTT aggregates is ALFY, which co‑immuno precipitates with p62, ATG5 and LC3 within mHTT aggregates in mammalian cells expressing exon 1 of HTT or in fibroblasts from patients with HD38. Furthermore, in mammalian cells expressing mHTT, knockdown and overexpression of ALFY led to an accumulation of aggregates and a reduction in the number of aggregates, respectively, and similar effects were observed in vivo in a D. melanogaster model of polyQ toxicity 38. Together, these findings provide good functional evidence that ALFY plays a part in facilitating autophagic clearance of polyQ protein aggregates. Recently, a new class of receptors, the Cue5–TOLLIP proteins (CUET proteins; also known as CUE-domain targeting adaptors), were identified in yeast from a screen for novel ubiquitin–Atg8 (LC3) interactors114. Of these, Cue5 has a putative mammalian homologue, namely Toll-interacting protein (TOLLIP). TOLLIP was observed to colocalize with LC3 in mammalian cells, and its overexpression enhanced the clearance of mHTT, suggesting that this protein and maybe other yet‑to‑be‑identified CUET homologues act as mammalian autophagy receptors. Although both p62 and CUET proteins contain ubiquitin-binding domains, ubiquitylation is not the only signal for selective recognition. p62 can also bind to non-ubiquitylated protein aggregates and target these for autophagic degradation115. In addition, NDP52, already known as a selective autophagy receptor from work in bacteria116, has recently been shown to co‑immunoprecipitate with human tau through its SKICH domain. Although it contains a Lim‑L domain that is capable of binding ubiquitylated targets, this domain was not required for its interaction with phosphorylated tau. NDP52 expression is upregulated in response to nuclear factor erythroid 2‑related factor 2 (NRF2) activation, which leads to enhanced clearance of phosphorylated tau both in vitro and in vivo, providing good evidence that NDP52 can mediate selective autophagic clearance of phosphorylated tau in response to NRF2 activation117. Another protein that has been shown to have a role as an autophagy receptor is HTT itself. In the presence of mHTT, autophagosomes are able to form but seem to lack content, suggesting a failure in the sequestration of substrates118. Further investigation of this phenomenon led to the discovery that wild-type HTT is able to

act as a scaffold for selective autophagy by interacting with p62 and facilitating its association with LC3 and ubiquitylated substrates119. Wild-type HTT was also shown to be involved in autophagy initiation: domains of the protein distinct from those shown to interact with p62 are able to interact with ULK1, which reduces the inhibitory interaction of ULK1 with mTOR and leads to autophagy activation119. Mutations in the genes encoding p62 and another receptor protein, optineurin, have been associated with ALS110,120, along with other diseases such as Paget disease of bone and open-angle glaucoma. The proposed functions for p62 and optineurin in the clearance of aggregate-prone proteins suggest a clear pathogenic mechanism for forms of ALS; however, mutations in the genes encoding p62 and optineurin are located throughout the genes, affecting multiple domains121. Although mutations affecting the ubiquitin-binding or LC3‑interacting domain of optineurin have been demonstrated to inhibit its function as a receptor 122, it has yet to be established whether mutations affecting other regions of the proteins have the same functional consequences.

Autophagy as a therapeutic target In recent years, numerous studies have demonstrated that the aggregate-prone proteins at the heart of neurodegenerative disease toxicity are autophagy substrates and that pharmacological upregulators of autophagy can be beneficial in both cell and animal models of these diseases, in which they are able to reduce both intracytoplasmic aggregates and associated cell death (reviewed in REF. 123). However, understanding the exact mechanisms by which autophagy may be compromised in neurodegeneration is vital in the quest for therapeutics that act on the pathway. In the presence of a defect, pharmacological upregulators acting upstream would not enhance aggregate-prone protein clearance, and where autophagosome clearance is disrupted, enhancing formation may exacerbate rather than ameliorate pathology. Increasing our knowledge of the function and dysfunction of autophagy in neurodegenerative disease will facilitate the design of therapeutic interventions that can bypass a block in the autophagic pathway, should it exist. Achieving this goal is feasible given the diverse mechanisms of action of the autophagy-modulating agents that have so far been identified and the existence of others for which the mechanisms are not yet completely understood. The mTOR inhibitor rapamycin was the first drug to be identified as an autophagy inducer and, as such, has been at the forefront of studies establishing the role of autophagy upregulation as a therapeutic strategy for neurodegeneration, along with more soluble analogues (rapalogues) such as temsirolimus (previously known as CCI‑779)124–128. However, mTOR has diverse autophagy-independent functions, and its inhibition has adverse effects in patients, including immuno suppression and impaired wound healing. Efforts have therefore been made to identify mTOR-independent upregulators of autophagy.

352 | JUNE 2015 | VOLUME 16


REVIEWS Several compounds have been identified that regulate autophagy via the cAMP–EPAC (exchange factor directly activated by cAMP 1)–phospholipase Cε (PLCε)–inositol‑1,4,5‑trisphosphate (Ins(1,4,5)P3) pathway and the Ca2+–calpain–Gsα pathway129. One of these compounds, rilmenidine, which acts via Gi‑coupled imidazoline receptors, was subsequently shown to decrease aggregate load and disease signs in a transgenic mouse model of HD130 and is now in safety trials in patients with this disease (European Clinical Trials Database number: 2009‑018119‑14). Several other drugs that act on these pathways have since been shown to be effective in mammalian models of neurodegeneration. Lithium acts to induce autophagy by reducing Ins(1,4,5)P3 levels through the inhibition of inositol monophosphatase131 and is protective in mouse models of tauopathy 132, whereas carbamazepine, which inhibits inositol synthesis, has been shown to be protective in mouse models of AD133. Additionally, inhibitors of calpain activation, such as calpastatin, are able to promote the clearance of aggregate-prone proteins by autophagy and are protective in mouse and D. melanogaster models of HD and tauopathy 134. Another mTOR-independent autophagy regulator, the disaccharide trehalose, acts through unknown mechanisms to promote autophagy 135. Trehalose is neuroprotective in mouse models of tauopathy 136,137 and prolongs lifespan in mouse models of ALS138,139. Box 4 | New concepts and areas for investigation Glial autophagy Although much of the focus of the field so far has been on neuronal autophagy, recent studies suggest that autophagy in glia may be similarly important in maintaining neuronal health and in the pathogenesis of neurodegenerative diseases. Many neurological diseases are associated with inflammatory responses, and autophagy is required for the maintenance of mitochondrial architecture during inflammation in astrocytes179. Studies in the Long–Evans shaker rat, a model of demyelinating disease180, show that autophagy can protect vulnerable oligodendrocytes. In humans, duplications of or point mutations in the gene encoding peripheral myelin protein 22 (PMP22) in Schwann cells cause the progressive demyelinating disease Charcot–Marie–Tooth disease type 1A (CMT1A). The levels of the aberrant PMP22 and Schwann cell health appear to be regulated by autophagy in mouse models of CMT1A, and autophagy upregulation may be a suitable strategy to consider for this class of diseases181. In lysosomal storage disorders, astrocytic autophagy may be important for neuronal health. Astrocyte-specific loss of the enzyme impaired in multiple sulfatase deficiency causes a defect in autophagosome maturation. As the ability of these astrocytes to support normal neurons is compromised, it is interesting to speculate whether this noncell-autonomous mechanism can be partially attributed to the autophagy defect182. Autophagy and secretion of proteins Recent studies have highlighted the ways in which autophagy may regulate various secretory processes183. This concept may be of relevance to neurodegenerative diseases, particularly those in which cell‑to‑cell spreading of toxic proteins such as tau and α‑synuclein have been implicated184. However, the current status of this field is complex. In Parkinson disease, it seems that lysosomal impairment or defective autophagosome– lysosome fusion causes increased exosomal α‑synuclein release, which can be reconciled with the resulting increased pool of autophagosomes185–187. However, in the context of Alzhiemer disease, the situation appears to be different. In mouse studies, the secretion of amyloid‑β and the extracellular plaque load were reduced when autophagy was blocked by deletion of Atg7 in relevant neurons. However, the authors suggest that this was associated with increased intracellular amyloid‑β levels and enhanced pathology188. This study reiterates the possibility that intracellular amyloid‑β may be a major toxic entity. However, it is important to note that in all these cases the secretory– excretory route has not been precisely defined189.

The combined evidence from different compounds and different models of neurodegeneration strongly support the idea that upregulation of autophagy is a valid therapeutic strategy for neurodegenerative disease. However, upregulating autophagosome biogenesis is unlikely to be suitable or safe for all neurodegenerative diseases. For example, in diseases in which autophagosome degradation is slowed owing to lysosomal lesions, increasing autophagosome formation will result in largely unproductive formation of autophagosomes that are not properly cleared and therefore a marked increase in autophagosomes, which may be deleterious. It has been suggested that one approach to ameliorate lysosomal function in some of these diseases is to enhance the activity of TFEB. Viral transduction of TFEB has demonstrated promising results in mouse models of LSDs140, PD105 and tauopathies141. Another strategy for the treatment of LSDs may be to reduce the activity of cystatins, which are endogenous inhibitors of lysosomal cathepsin proteases. This strategy has also yielded promising results in mouse models of AD142.

Future challenges As our understanding of the autophagic process increases, we are beginning to identify new areas that will be important to investigate in the quest to understand the role of autophagy in neurodegenerative disease (BOX 4). One of the major difficulties in studying autophagic processes in the brain in vivo is to define flux on the basis of ‘snapshot’ data. For example, if a study finds that there are increased numbers of autophagosomes in the brain in a mouse model or an individual with a particular disease, this does not mean that autophagy is increased. An increase in autophagosome numbers could be due to increased formation (which can lead to increased delivery to lysosomes of substrates) or could be due to impaired lysosomal degradation or trafficking of completed autophagosomes (as occurs in LSDs). In mice, at least, good progress has been made in adapting methods to differentiate between these scenarios. Hetz and colleagues143 performed intraventricular injections of adenoviruses encoding monomeric tandem mCherry–GFP–LC3, which were then expressed in central and peripheral neurons. Before lysosome fusion, the LC3 vesicles were autophagosomes and exhibited both red and green fluorescence, but after fusion, the vesicles were autolysosomes and exhibited red fluorescence only (as GFP was more rapidly quenched in the acidic environment). Assessing the numbers of the two types (colours) of vesicles enabled a more reliable assessment of autophagic flux. Another important issue relates to the interpretation of whether autophagy is causing cell death in certain diseases in which increased numbers of autophagosomes are observed. First, one needs to know whether there is increased formation of autophagosomes or decreased degradation. If there is increased formation, then one needs to use a number of genetic manipulations of autophagy genes to perturb autophagy in order to test whether the increased autophagy is causally involved in the toxic process. This is preferable to



REVIEWS chemical strategies because all the drugs currently used to inhibit (or induce) autophagy have off-target effects. One also needs to consider that manipulations of some autophagy genes may have consequences unrelated to autophagy (for example, beclin 1 indirectly regulates p53 levels144). Recent data suggest additional non-autophagy roles of beclin 1 in retromer trafficking and phagocytosis of amyloid‑β in microglia, and this appears to be diseaserelevant because reduced beclin 1 and retromer levels have been observed in AD brains145. These challenges are difficult enough to negotiate in cell culture and mouse experiments. However, ultimately, we need to understand human diseases, many of which are complex and cannot be faithfully modelled in animals. Thus, a major challenge for the future will be to develop markers that enable inferences regarding autophagic flux in vivo and in post-mortem samples. Such new methods

1. Imarisio, S. et al. Huntington’s disease: from pathology and genetics to potential therapies. Biochem. J. 412, 191–209 (2008). 2. Rubinsztein, D. C. The roles of intracellular proteindegradation pathways in neurodegeneration. Nature 443, 780–786 (2006). 3. Tsvetkov, A. S. et al. Proteostasis of polyglutamine varies among neurons and predicts neurodegeneration. Nature Chem. Biol. 9, 586–592 (2013). 4. Hara, T. et al. Suppression of basal autophagy in neural cells causes neurodegenerative disease in mice. Nature 441, 885–889 (2006). 5. Komatsu, M. et al. Loss of autophagy in the central nervous system causes neurodegeneration in mice. Nature 441, 880–884 (2006). References 4 and 5 were the first to demonstrate the importance of autophagy for neuronal health (knockout of Atg5 or Atg7 in neurons resulted in a neurodegenerative phenotype in mice). 6. Kim, I. & Lemasters, J. J. Mitophagy selectively degrades individual damaged mitochondria after photoirradiation. Antioxid. Redox Signal. 14, 1919–1928 (2011). 7. Boya, P. et al. Inhibition of macroautophagy triggers apoptosis. Mol. Cell. Biol. 25, 1025–1040 (2005). 8. Ravikumar, B., Berger, Z., Vacher, C., O’Kane, C. J. & Rubinsztein, D. C. Rapamycin pre-treatment protects against apoptosis. Hum. Mol. Genet. 15, 1209–1216 (2006). 9. Korolchuk, V. I., Mansilla, A., Menzies, F. M. & Rubinsztein, D. C. Autophagy inhibition compromises degradation of ubiquitin–proteasome pathway substrates. Mol. Cell 33, 517–527 (2009). 10. Tsukada, M. & Ohsumi, Y. Isolation and characterization of autophagy-defective mutants of Saccharomyces cerevisiae. FEBS Lett. 333, 169–174 (1993). 11. Mizushima, N., Levine, B., Cuervo, A. M. & Klionsky, D. J. Autophagy fights disease through cellular self-digestion. Nature 451, 1069–1075 (2008). 12. Ravikumar, B., Moreau, K., Jahreiss, L., Puri, C. & Rubinsztein, D. C. Plasma membrane contributes to the formation of pre-autophagosomal structures. Nature Cell Biol. 12, 747–757 (2010). 13. Hayashi-Nishino, M. et al. A subdomain of the endoplasmic reticulum forms a cradle for autophagosome formation. Nature Cell Biol. 11, 1433–1437 (2009). 14. Yla-Anttila, P., Vihinen, H., Jokitalo, E. & Eskelinen, E. L. 3D tomography reveals connections between the phagophore and endoplasmic reticulum. Autophagy 5, 1180–1185 (2009). 15. van der Vaart, A., Griffith, J. & Reggiori, F. Exit from the Golgi is required for the expansion of the autophagosomal phagophore in yeast Saccharomyces cerevisiae. Mol. Biol. Cell 21, 2270–2284 (2010). 16. Yen, W. L. et al. The conserved oligomeric Golgi complex is involved in double-membrane vesicle formation during autophagy. J. Cell Biol. 188, 101–114 (2010).

will not only be helpful for understanding disease pathogenesis but also serve as a powerful tool for monitoring target engagement in therapeutic trials.

Conclusions There are now substantial data showing how autophagy may be compromised in various neurodegenerative diseases and extensive preclinical support for the use of autophagy inducers in certain neurodegenerative disease models. Further understanding of the basic biology of autophagy will provide additional insights into how disease-causing protein variants may affect this critical degradation pathway. Similarly, further drug discovery efforts coupled with the development of relevant biomarkers will facilitate the testing of whether autophagy induction can affect the genesis or progression of some neurodegenerative diseases.

17. Hailey, D. W. et al. Mitochondria supply membranes for autophagosome biogenesis during starvation. Cell 141, 656–667 (2010). 18. Puri, C., Renna, M., Bento, C. F., Moreau, K. & Rubinsztein, D. C. Diverse autophagosome membrane sources coalesce in recycling endosomes. Cell 154, 1285–1299 (2013). 19. Kabeya, Y. et al. LC3, a mammalian homologue of yeast Apg8p, is localized in autophagosome membranes after processing. EMBO J. 19, 5720–5728 (2000). 20. Ravikumar, B. et al. Dynein mutations impair autophagic clearance of aggregate-prone proteins. Nature Genet. 37, 771–776 (2005). 21. Moreau, K., Ravikumar, B., Renna, M., Puri, C. & Rubinsztein, D. C. Autophagosome precursor maturation requires homotypic fusion. Cell 146, 303–317 (2011). 22. Fader, C. M., Sanchez, D. G., Mestre, M. B. & Colombo, M. I. TI‑VAMP/VAMP7 and VAMP3/ cellubrevin: two v‑SNARE proteins involved in specific steps of the autophagy/multivesicular body pathways. Biochim. Biophys. Acta 1793, 1901–1916 (2009). 23. Furuta, N. & Amano, A. Cellular machinery to fuse antimicrobial autophagosome with lysosome. Commun. Integr. Biol. 3, 385–387 (2010). 24. Alers, S., Loffler, A. S., Wesselborg, S. & Stork, B. Role of AMPK–mTOR–Ulk1/2 in the regulation of autophagy: cross talk, shortcuts, and feedbacks. Mol. Cell. Biol. 32, 2–11 (2012). 25. Fullgrabe, J., Klionsky, D. J. & Joseph, B. The return of the nucleus: transcriptional and epigenetic control of autophagy. Nature Rev. Mol. Cell Biol. 15, 65–74 (2014). 26. Morselli, E. et al. p53 inhibits autophagy by interacting with the human ortholog of yeast Atg17, RB1CC1/FIP200. Cell Cycle 10, 2763–2769 (2011). 27. Russell, R. C. et al. ULK1 induces autophagy by phosphorylating Beclin‑1 and activating VPS34 lipid kinase. Nature Cell Biol. 15, 741–750 (2013). 28. Dooley, H. C. et al. WIPI2 links LC3 conjugation with PI3P, autophagosome formation, and pathogen clearance by recruiting Atg12‑5‑16L1. Mol. Cell 55, 238–252 (2014). 29. Mack, H. I., Zheng, B., Asara, J. M. & Thomas, S. M. AMPK-dependent phosphorylation of ULK1 regulates ATG9 localization. Autophagy 8, 1197–1214 (2012). 30. Papinski, D. et al. Early steps in autophagy depend on direct phosphorylation of Atg9 by the Atg1 kinase. Mol. Cell 53, 471–483 (2014). 31. Vicinanza, M. et al. PI(5)P regulates autophagosome biogenesis. Mol. Cell 57, 219–234 (2015). 32. Stolz, A., Ernst, A. & Dikic, I. Cargo recognition and trafficking in selective autophagy. Nature Cell Biol. 16, 495–501 (2014). 33. Bjorkoy, G. et al. p62/SQSTM1 forms protein aggregates degraded by autophagy and has a protective effect on huntingtin-induced cell death. J. Cell Biol. 171, 603–614 (2005).

354 | JUNE 2015 | VOLUME 16

34. Kirkin, V. et al. A role for NBR1 in autophagosomal degradation of ubiquitinated substrates. Mol. Cell 33, 505–516 (2009). 35. Pankiv, S. et al. p62/SQSTM1 binds directly to Atg8/ LC3 to facilitate degradation of ubiquitinated protein aggregates by autophagy. J. Biol. Chem. 282, 24131–24145 (2007). 36. Rogov, V., Dotsch, V., Johansen, T. & Kirkin, V. Interactions between autophagy receptors and ubiquitin-like proteins form the molecular basis for selective autophagy. Mol. Cell 53, 167–178 (2014). 37. Simonsen, A. et al. Alfy, a novel FYVE-domaincontaining protein associated with protein granules and autophagic membranes. J. Cell Sci. 117, 4239–4251 (2004). 38. Filimonenko, M. et al. The selective macroautophagic degradation of aggregated proteins requires the PI3P‑binding protein Alfy. Mol. Cell 38, 265–279 (2010). 39. Clausen, T. H. et al. p62/SQSTM1 and ALFY interact to facilitate the formation of p62 bodies/ALIS and their degradation by autophagy. Autophagy 6, 330–344 (2010). 40. Novak, I. et al. Nix is a selective autophagy receptor for mitochondrial clearance. EMBO Rep. 11, 45–51 (2010). 41. Schwarten, M. et al. Nix directly binds to GABARAP: a possible crosstalk between apoptosis and autophagy. Autophagy 5, 690–698 (2009). 42. Shen, H. M. & Mizushima, N. At the end of the autophagic road: an emerging understanding of lysosomal functions in autophagy. Trends Biochem. Sci. 39, 61–71 (2014). 43. Wong, Y. C. & Holzbaur, E. L. The regulation of autophagosome dynamics by huntingtin and HAP1 is disrupted by expression of mutant huntingtin, leading to defective cargo degradation. J. Neurosci. 34, 1293–1305 (2014). 44. Moreau, K., Renna, M. & Rubinsztein, D. C. Connections between SNAREs and autophagy. Trends Biochem. Sci. 38, 57–63 (2013). 45. Yu, L. et al. Termination of autophagy and reformation of lysosomes regulated by mTOR. Nature 465, 942–946 (2010). 46. Sardiello, M. et al. A gene network regulating lysosomal biogenesis and function. Science 325, 473–477 (2009). 47. Settembre, C. et al. TFEB links autophagy to lysosomal biogenesis. Science 332, 1429–1433 (2011). 48. Frake, R. A., Ricketts, T., Menzies, F. M. & Rubinsztein, D. C. Autophagy and neurodegeneration. J. Clin. Invest. 125, 65–75 (2015). 49. Haack, T. B., Hogarth, P., Gregory, A., Prokisch, H. & Hayflick, S. J. BPAN: the only X‑linked dominant NBIA disorder. Int. Rev. Neurobiol. 110, 85–90 (2013). 50. Saitsu, H. et al. De novo mutations in the autophagy gene WDR45 cause static encephalopathy of childhood with neurodegeneration in adulthood. Nature Genet. 45, 445–449 (2013). This paper shows that static encephalopathy of childhood with neurodegeneration in adulthood (SENDA; also known as BPAN) results from mutations in WDR45.


REVIEWS 51. Orsi, A. et al. Dynamic and transient interactions of Atg9 with autophagosomes, but not membrane integration, are required for autophagy. Mol. Biol. Cell 23, 1860–1873 (2012). 52. Metzger, S. et al. Age at onset in Huntington’s disease is modified by the autophagy pathway: implication of the V471A polymorphism in Atg7. Hum. Genet. 128, 453–459 (2010). 53. Metzger, S. et al. The V471A polymorphism in autophagy-related gene ATG7 modifies age at onset specifically in Italian Huntington disease patients. PLoS ONE 8, e68951 (2013). 54. Yan, J. Q. et al. Overexpression of human E46K mutant α-synuclein impairs macroautophagy via inactivation of JNK1–Bcl‑2 pathway. Mol. Neurobiol. 50, 685–701 (2014). 55. Subramaniam, S., Sixt, K. M., Barrow, R. & Snyder, S. H. Rhes, a striatal specific protein, mediates mutant-huntingtin cytotoxicity. Science 324, 1327–1330 (2009). 56. Mealer, R. G., Murray, A. J., Shahani, N., Subramaniam, S. & Snyder, S. H. Rhes, a striatalselective protein implicated in Huntington disease, binds beclin‑1 and activates autophagy. J. Biol. Chem. 289, 3547–3554 (2014). The results from this paper suggest a possible mechanism for the selective vulnerability of striatal cells in HD. 57. Sagona, A. P. et al. A tumor-associated mutation of FYVE-CENT prevents its interaction with Beclin 1 and interferes with cytokinesis. PLoS ONE 6, e17086 (2011). 58. Vantaggiato, C. et al. Defective autophagy in spastizin mutated patients with hereditary spastic paraparesis type 15. Brain 136, 3119–3139 (2013). 59. Harold, D. et al. Genome-wide association study identifies variants at CLU and PICALM associated with Alzheimer’s disease. Nature Genet. 41, 1088–1093 (2009). 60. Jun, G. et al. Meta-analysis confirms CR1, CLU, and PICALM as Alzheimer disease risk loci and reveals interactions with APOE genotypes. Arch. Neurol. 67, 1473–1484 (2010). 61. Ando, K. et al. Clathrin adaptor CALM/PICALM is associated with neurofibrillary tangles and is cleaved in Alzheimer’s brains. Acta Neuropathol. 125, 861–878 (2013). 62. Moreau, K. et al. PICALM modulates autophagy activity and tau accumulation. Nature Commun. 5, 4998 (2014). 63. Szatmari, Z. & Sass, M. The autophagic roles of Rab small GTPases and their upstream regulators: a review. Autophagy 10, 1154–1166 (2014). 64. Verhoeven, K. et al. Mutations in the small GTP-ase late endosomal protein RAB7 cause Charcot–Marie– Tooth type 2B neuropathy. Am. J. Hum. Genet. 72, 722–727 (2003). 65. Yang, Y. et al. The gene encoding alsin, a protein with three guanine-nucleotide exchange factor domains, is mutated in a form of recessive amyotrophic lateral sclerosis. Nature Genet. 29, 160–165 (2001). 66. Topp, J. D., Gray, N. W., Gerard, R. D. & Horazdovsky, B. F. Alsin is a Rab5 and Rac1 guanine nucleotide exchange factor. J. Biol. Chem. 279, 24612–24623 (2004). 67. Ravikumar, B., Imarisio, S., Sarkar, S., O’Kane, C. J. & Rubinsztein, D. C. Rab5 modulates aggregation and toxicity of mutant huntingtin through macroautophagy in cell and fly models of Huntington disease. J. Cell Sci. 121, 1649–1660 (2008). 68. Otomo, A., Kunita, R., Suzuki-Utsunomiya, K., Ikeda, J. E. & Hadano, S. Defective relocalization of ALS2/alsin missense mutants to Rac1‑induced macropinosomes accounts for loss of their cellular function and leads to disturbed amphisome formation. FEBS Lett. 585, 730–736 (2011). 69. Levine, T. P., Daniels, R. D., Gatta, A. T., Wong, L. H. & Hayes, M. J. The product of C9orf72, a gene strongly implicated in neurodegeneration, is structurally related to DENN Rab-GEFs. Bioinformatics 29, 499–503 (2013). 70. Farg, M. A. et al. C9ORF72, implicated in amytrophic lateral sclerosis and frontotemporal dementia, regulates endosomal trafficking. Hum. Mol. Genet. 23, 3579–3595 (2014). 71. Mizielinska, S. et al. C9orf72 repeat expansions cause neurodegeneration in Drosophila through arginine-rich proteins. Science 345,1192–1194 (2014). 72. Skibinski, G. et al. Mutations in the endosomal ESCRTIII-complex subunit CHMP2B in frontotemporal dementia. Nature Genet. 37, 806–808 (2005).

73. Lee, J. A., Beigneux, A., Ahmad, S. T., Young, S. G. & Gao, F. B. ESCRT-III dysfunction causes autophagosome accumulation and neurodegeneration. Curr. Biol. 17, 1561–1567 (2007). 74. Seaman, M. N. The retromer complex — endosomal protein recycling and beyond. J. Cell Sci. 125, 4693–4702 (2012). 75. Zavodszky, E. et al. Mutation in VPS35 associated with Parkinson’s disease impairs WASH complex association and inhibits autophagy. Nature Commun. 5, 3828 (2014). 76. Zimprich, A. et al. A mutation in VPS35, encoding a subunit of the retromer complex, causes late-onset Parkinson disease. Am. J. Hum. Genet. 89, 168–175 (2011). 77. Winslow, A. R. et al. α-synuclein impairs macroautophagy: implications for Parkinson’s disease. J. Cell Biol. 190, 1023–1037 (2010). 78. Singleton, A. B. et al. α-synuclein locus triplication causes Parkinson’s disease. Science 302, 841 (2003). 79. Cooper, A. A. et al. α-synuclein blocks ER–Golgi traffic and Rab1 rescues neuron loss in Parkinson’s models. Science 313, 324–328 (2006). 80. Freeman, C., Seaman, M. N. & Reid, E. The hereditary spastic paraplegia protein strumpellin: characterisation in neurons and of the effect of disease mutations on WASH complex assembly and function. Biochim. Biophys. Acta 1832, 160–173 (2013). 81. Ropers, F. et al. Identification of a novel candidate gene for non-syndromic autosomal recessive intellectual disability: the WASH complex member SWIP. Hum. Mol. Genet. 20, 2585–2590 (2011). 82. Vilarino-Guell, C. et al. DNAJC13 mutations in Parkinson disease. Hum. Mol. Genet. 23, 1794–1801 (2014). 83. Al‑Saif, A., Al‑Mohanna, F. & Bohlega, S. A mutation in sigma‑1 receptor causes juvenile amyotrophic lateral sclerosis. Ann. Neurol. 70, 913–919 (2011). 84. Luty, A. A. et al. Sigma nonopioid intracellular receptor 1 mutations cause frontotemporal lobar degeneration–motor neuron disease. Ann. Neurol. 68, 639–649 (2010). 85. Prause, J. et al. Altered localization, abnormal modification and loss of function of Sigma receptor‑1 in amyotrophic lateral sclerosis. Hum. Mol. Genet. 22, 1581–1600 (2013). 86. Mavlyutov, T. A., Epstein, M. L., Andersen, K. A., Ziskind-Conhaim, L. & Ruoho, A. E. The sigma‑1 receptor is enriched in postsynaptic sites of C‑terminals in mouse motoneurons. An anatomical and behavioral study. Neuroscience 167, 247–255 (2010). 87. Mancuso, R. et al. Sigma‑1R agonist improves motor function and motoneuron survival in ALS mice. Neurotherapeutics 9, 814–826 (2012). 88. Vollrath, J. T. et al. Loss of function of the ALS protein SigR1 leads to ER pathology associated with defective autophagy and lipid raft disturbances. Cell Death Dis. 5, e1290 (2014). 89. Lieberman, A. P. et al. Autophagy in lysosomal storage disorders. Autophagy 8, 719–730 (2012). 90. Swan, M. & Saunders-Pullman, R. The association between β‑glucocerebrosidase mutations and parkinsonism. Curr. Neurol. Neurosci. Rep. 13, 368 (2013). 91. Osellame, L. D. et al. Mitochondria and quality control defects in a mouse model of Gaucher disease — links to Parkinson’s disease. Cell. Metabolism 17, 941–953 (2013). 92. Mazzulli, J. R. et al. Gaucher disease glucocerebrosidase and α-synuclein form a bidirectional pathogenic loop in synucleinopathies. Cell 146, 37–52 (2011). This paper identifies the molecular link between PD and heterozygous mutations in GBA. Glucosylceramide, which accumulates in Gaucher disease caused by homozygous mutations in GBA, compromises lysosomal function and leads to an accumulation of α‑synuclein; conversely, α‑synuclein inhibits GBA, resulting in an increase in glucosylceramide. 93. Murphy, K. E. et al. Reduced glucocerebrosidase is associated with increased α-synuclein in sporadic Parkinson’s disease. Brain 137, 834–848 (2014). 94. Boya, P. & Kroemer, G. Lysosomal membrane permeabilization in cell death. Oncogene 27, 6434–6451 (2008). 95. Gabande-Rodriguez, E., Boya, P., Labrador, V., Dotti, C. G. & Ledesma, M. D. High sphingomyelin levels induce lysosomal damage and autophagy dysfunction in Niemann Pick disease type A. Cell Death Differ. 21, 864–875 (2014).


96. Ramirez, A. et al. Hereditary parkinsonism with dementia is caused by mutations in ATP13A2, encoding a lysosomal type 5 P‑type ATPase. Nature Genet. 38, 1184–1191 (2006). 97. Dehay, B. et al. Loss of P‑type ATPase ATP13A2/ PARK9 function induces general lysosomal deficiency and leads to Parkinson disease neurodegeneration. Proc. Natl Acad. Sci. USA 109, 9611–9616 (2012). 98. Lee, J. H. et al. Lysosomal proteolysis and autophagy require presenilin 1 and are disrupted by Alzheimerrelated PS1 mutations. Cell 141, 1146–1158 (2010). 99. Avrahami, L. et al. Inhibition of glycogen synthase kinase‑3 ameliorates β-amyloid pathology and restores lysosomal acidification and mammalian target of rapamycin activity in the Alzheimer disease mouse model: in vivo and in vitro studies. J. Biol. Chem. 288, 1295–1306 (2013). 100. Zhang, X. et al. A role for presenilins in autophagy revisited: normal acidification of lysosomes in cells lacking PSEN1 and PSEN2. J. Neurosci. 32, 8633–8648 (2012). 101. Coen, K. et al. Lysosomal calcium homeostasis defects, not proton pump defects, cause endo-lysosomal dysfunction in PSEN-deficient cells. J. Cell Biol. 198, 23–35 (2012). 102. Wilson, C. A. et al. Degradative organelles containing mislocalized α- and β-synuclein proliferate in presenilin‑1 null neurons. J. Cell Biol. 165, 335–346 (2004). 103. Neely, K. M., Green, K. N. & LaFerla, F. M. Presenilin is necessary for efficient proteolysis through the autophagy-lysosome system in a γ-secretaseindependent manner. J. Neurosci. 31, 2781–2791 (2011). 104. Cortes, C. J. et al. Polyglutamine-expanded androgen receptor interferes with TFEB to elicit autophagy defects in SBMA. Nature Neurosci. 17, 1180–1189 (2014). 105. Decressac, M. et al. TFEB-mediated autophagy rescues midbrain dopamine neurons from α-synuclein toxicity. Proc. Natl Acad. Sci. USA 110, E1817–E1826 (2013). This study shows that overexpression of α‑synuclein in rat midbrain results in a mislocalization of TFEB and the degeneration of dopaminergic neurons. This degeneration could be rescued by inducing autophagy via overexpression of TFEB or treatment with the rapamycin analogue CCI‑779. 106. Lee, J. K. et al. Acid sphingomyelinase modulates the autophagic process by controlling lysosomal biogenesis in Alzheimer’s disease. J. Exp. Med. 211, 1551–1570 (2014). 107. Chang, J., Lee, S. & Blackstone, C. Spastic paraplegia proteins spastizin and spatacsin mediate autophagic lysosome reformation. J. Clin. Invest. 124, 5249–5262 (2014). This paper presents a newly identified process in autophagy, in which lysosomes are replenished, involving two proteins in which mutations have been identified in HSP. 108. Zatloukal, K. et al. p62 is a common component of cytoplasmic inclusions in protein aggregation diseases. Am. J. Pathol. 160, 255–263 (2002). 109. Mori, F. et al. Optineurin immunoreactivity in neuronal nuclear inclusions of polyglutamine diseases (Huntington’s, DRPLA, SCA2, SCA3) and intranuclear inclusion body disease. Acta Neuropathol. 123, 747–749 (2012). 110. Maruyama, H. et al. Mutations of optineurin in amyotrophic lateral sclerosis. Nature 465, 223–226 (2010). 111. Saitoh, Y. et al. p62 plays a protective role in the autophagic degradation of polyglutamine protein oligomers in polyglutamine disease model flies. J. Biol. Chem. 290, 1442–1453 (2015). 112. Doi, H. et al. p62/SQSTM1 differentially removes the toxic mutant androgen receptor via autophagy and inclusion formation in a spinal and bulbar muscular atrophy mouse model. J. Neurosci. 33, 7710–7727 (2013). 113. Kurosawa, M. et al. Depletion of p62 reduces nuclear inclusions and paradoxically ameliorates disease phenotypes in Huntington’s model mice. Hum. Mol. Genet. 24, 1092–1105 (2014). 114. Lu, K., Psakhye, I. & Jentsch, S. Autophagic clearance of polyQ proteins mediated by ubiquitin‑Atg8 adaptors of the conserved CUET protein family. Cell 158, 549–563 (2014). This paper describes an entirely new class of autophagy receptor proteins that have a specific role in the clearance of aggregate-prone proteins.


REVIEWS 115. Watanabe, Y. & Tanaka, M. p62/SQSTM1 in autophagic clearance of a non-ubiquitylated substrate. J. Cell Sci. 124, 2692–2701 (2011). 116. Thurston, T. L., Ryzhakov, G., Bloor, S., von Muhlinen, N. & Randow, F. The TBK1 adaptor and autophagy receptor NDP52 restricts the proliferation of ubiquitin-coated bacteria. Nature Immunol. 10, 1215–1221 (2009). 117. Jo, C. et al. Nrf2 reduces levels of phosphorylated tau protein by inducing autophagy adaptor protein NDP52. Nature Commun. 5, 3496 (2014). 118. Martinez-Vicente, M. et al. Cargo recognition failure is responsible for inefficient autophagy in Huntington’s disease. Nature Neurosci. 13, 567–576 (2010). 119. Rui, Y. N. et al. Huntingtin functions as a scaffold for selective macroautophagy. Nature Cell Biol. 17, 262–275 (2015). 120. Fecto, F. et al. SQSTM1 mutations in familial and sporadic amyotrophic lateral sclerosis. Arch. Neurol. 68, 1440–1446 (2011). 121. Teyssou, E. et al. Mutations in SQSTM1 encoding p62 in amyotrophic lateral sclerosis: genetics and neuropathology. Acta Neuropathol. 125, 511–522 (2013). 122. Wong, Y. C. & Holzbaur, E. L. Optineurin is an autophagy receptor for damaged mitochondria in parkin-mediated mitophagy that is disrupted by an ALS-linked mutation. Proc. Natl Acad. Sci. USA 111, E4439–E4448 (2014). 123. Harris, H. & Rubinsztein, D. C. Control of autophagy as a therapy for neurodegenerative disease. Nature Rev. Neurol. 8, 108–117 (2012). 124. Rodriguez-Muela, N., Germain, F., Marino, G., Fitze, P. S. & Boya, P. Autophagy promotes survival of retinal ganglion cells after optic nerve axotomy in mice. Cell Death Differ. 19, 162–169 (2012). 125. Ravikumar, B. et al. Inhibition of mTOR induces autophagy and reduces toxicity of polyglutamine expansions in fly and mouse models of Huntington disease. Nature Genet. 36, 585–595 (2004). Using rapamycin to treat animal models of HD, this study was the first to demonstrate that upregulation of autophagy can have a beneficial effect in in vivo models of the disease. 126. Menzies, F. M. et al. Autophagy induction reduces mutant ataxin‑3 levels and toxicity in a mouse model of spinocerebellar ataxia type 3. Brain 133, 93–104 (2010). 127. Spilman, P. et al. Inhibition of mTOR by rapamycin abolishes cognitive deficits and reduces amyloid-β levels in a mouse model of Alzheimer’s disease. PLoS ONE 5, e9979 (2010). 128. Jiang, T. et al. Temsirolimus promotes autophagic clearance of amyloid-β and provides protective effects in cellular and animal models of Alzheimer’s disease. Pharmacol. Res. 81, 54–63 (2014). 129. Williams, A. et al. Novel targets for Huntington’s disease in an mTOR-independent autophagy pathway. Nature Chem. Biol. 4, 295–305 (2008). 130. Rose, C. et al. Rilmenidine attenuates toxicity of polyglutamine expansions in a mouse model of Huntington’s disease. Hum. Mol. Genet. 19, 2144–2153 (2010). 131. Sarkar, S. et al. Lithium induces autophagy by inhibiting inositol monophosphatase. J. Cell Biol. 170, 1101–1111 (2005). 132. Shimada, K. et al. Long-term oral lithium treatment attenuates motor disturbance in tauopathy model mice: implications of autophagy promotion. Neurobiol. Dis. 46, 101–108 (2012). 133. Li, L. et al. Autophagy enhancer carbamazepine alleviates memory deficits and cerebral amyloid-β pathology in a mouse model of Alzheimer’s disease. Curr. Alzheimer Res. 10, 433–441 (2013). 134. Menzies, F. M. et al. Calpain inhibition mediates autophagy-dependent protection against polyglutamine toxicity. Cell Death Differ. 22, 433–444 (2014). 135. Sarkar, S., Davies, J. E., Huang, Z., Tunnacliffe, A. & Rubinsztein, D. C. Trehalose, a novel mTORindependent autophagy enhancer, accelerates the clearance of mutant huntingtin and α-synuclein. J. Biol. Chem. 282, 5641–5652 (2007). 136. Rodriguez-Navarro, J. A. et al. Trehalose ameliorates dopaminergic and tau pathology in parkin deleted/tau overexpressing mice through autophagy activation. Neurobiol. Dis. 39, 423–438 (2010).

137. Schaeffer, V. et al. Stimulation of autophagy reduces neurodegeneration in a mouse model of human tauopathy. Brain 135, 2169–2177 (2012). 138. Castillo, K. et al. Trehalose delays the progression of amyotrophic lateral sclerosis by enhancing autophagy in motoneurons. Autophagy 9, 1308–1320 (2013). 139. Zhang, X. et al. MTOR-independent, autophagic enhancer trehalose prolongs motor neuron survival and ameliorates the autophagic flux defect in a mouse model of amyotrophic lateral sclerosis. Autophagy 10, 588–602 (2014). 140. Medina, D. L. et al. Transcriptional activation of lysosomal exocytosis promotes cellular clearance. Dev. Cell 21, 421–430 (2011). This study demonstrates that TFEB can promote fusion of lysosomes with the plasma membrane and the release of their contents. This process is protective in mouse models of the lysosomal storage disorder multiple sulfatase deficiency. 141. Polito, V. A. et al. Selective clearance of aberrant tau proteins and rescue of neurotoxicity by transcription factor EB. EMBO Mol. Med. 6, 1142–1160 (2014). 142. Yang, D. S. et al. Reversal of autophagy dysfunction in the TgCRND8 mouse model of Alzheimer’s disease ameliorates amyloid pathologies and memory deficits. Brain 134, 258–277 (2011). 143. Castillo, K. et al. Measurement of autophagy flux in the nervous system in vivo. Cell Death Dis. 4, e917 (2013). 144. Liu, J. et al. Beclin1 controls the levels of p53 by regulating the deubiquitination activity of USP10 and USP13. Cell 147, 223–234 (2011). 145. Lucin, K. M. et al. Microglial beclin 1 regulates retromer trafficking and phagocytosis and is impaired in Alzheimer’s disease. Neuron 79, 873–886 (2013). 146. Clark, I. E. et al. Drosophila pink1 is required for mitochondrial function and interacts genetically with parkin. Nature 441, 1162–1166 (2006). 147. Park, J. et al. Mitochondrial dysfunction in Drosophila PINK1 mutants is complemented by parkin. Nature 441, 1157–1161 (2006). 148. Narendra, D., Tanaka, A., Suen, D. F. & Youle, R. J. Parkin is recruited selectively to impaired mitochondria and promotes their autophagy. J. Cell Biol. 183, 795–803 (2008). This pioneering paper linked Parkin to mitophagy for the first time and led to the subsequent surge in understanding of the potential roles of certain genes that are mutated in PD. 149. Becker, D., Richter, J., Tocilescu, M. A., Przedborski, S. & Voos, W. Pink1 kinase and its membrane potential (Deltapsi)-dependent cleavage product both localize to outer mitochondrial membrane by unique targeting mode. J. Biol. Chem. 287, 22969–22987 (2012). 150. Greene, A. W. et al. Mitochondrial processing peptidase regulates PINK1 processing, import and Parkin recruitment. EMBO Rep. 13, 378–385 (2012). 151. Jin, S. M. et al. Mitochondrial membrane potential regulates PINK1 import and proteolytic destabilization by PARL. J. Cell Biol. 191, 933–942 (2010). 152. Geisler, S. et al. PINK1/Parkin-mediated mitophagy is dependent on VDAC1 and p62/SQSTM1. Nature Cell Biol. 12, 119–131 (2010). 153. Matsuda, N. et al. PINK1 stabilized by mitochondrial depolarization recruits Parkin to damaged mitochondria and activates latent Parkin for mitophagy. J. Cell Biol. 189, 211–221 (2010). 154. Narendra, D. P. et al. PINK1 is selectively stabilized on impaired mitochondria to activate Parkin. PLoS Biol. 8, e1000298 (2010). 155. Vives-Bauza, C. et al. PINK1‑dependent recruitment of Parkin to mitochondria in mitophagy. Proc. Natl Acad. Sci. USA 107, 378–383 (2010). 156. de Vries, R. L. & Przedborski, S. Mitophagy and Parkinson’s disease: be eaten to stay healthy. Mol. Cell. Neurosci. 55, 37–43 (2013). 157. Bingol, B. et al. The mitochondrial deubiquitinase USP30 opposes parkin-mediated mitophagy. Nature 510, 370–375 (2014). 158. Narendra, D., Kane, L. A., Hauser, D. N., Fearnley, I. M. & Youle, R. J. p62/SQSTM1 is required for Parkin-induced mitochondrial clustering but not mitophagy; VDAC1 is dispensable for both. Autophagy 6, 1090–1106 (2010). 159. Scarffe, L. A., Stevens, D. A., Dawson, V. L. & Dawson, T. M. Parkin and PINK1: much more than mitophagy. Trends Neurosci. 37, 315–324 (2014).

356 | JUNE 2015 | VOLUME 16

160. Ashrafi, G., Schlehe, J. S., LaVoie, M. J. & Schwarz, T. L. Mitophagy of damaged mitochondria occurs locally in distal neuronal axons and requires PINK1 and Parkin. J. Cell Biol. 206, 655–670 (2014). 161. Chu, C. T. et al. Cardiolipin externalization to the outer mitochondrial membrane acts as an elimination signal for mitophagy in neuronal cells. Nature Cell Biol. 15, 1197–1205 (2013). This paper describes an alternative mitophagy activation signal differing from that involving Parkin and PINK1, in which cardiolipin, a lipid normally localized to the mitochondrial inner membrane, is externalized. 162. Do, C. B. et al. Web-based genome-wide association study identifies two novel loci and a substantial genetic component for Parkinson’s disease. PLoS Genet. 7, e1002141 (2011). 163. Ivatt, R. M. et al. Genome-wide RNAi screen identifies the Parkinson disease GWAS risk locus SREBF1 as a regulator of mitophagy. Proc. Natl Acad. Sci. USA 111, 8494–8499 (2014). 164. Fang, E. F. et al. Defective mitophagy in XPA via PARP‑1 hyperactivation and NAD+/SIRT1 reduction. Cell 157, 882–896 (2014). 165. Cuervo, A. M. & Wong, E. Chaperone-mediated autophagy: roles in disease and aging. Cell Res. 24, 92–104 (2014). 166. Wang, G. & Mao, Z. Chaperone-mediated autophagy: roles in neurodegeneration. Transl. Neurodegener. 3, 20 (2014). 167. Xilouri, M. & Stefanis, L. Chaperone mediated autophagy to the rescue: a new-fangled target for the treatment of neurodegenerative diseases. Mol. Cell. Neurosci. http://dx.doi.org/10.1016/j. mcn.2015.01.003 (2015). 168. Cuervo, A. M., Stefanis, L., Fredenburg, R., Lansbury, P. T. & Sulzer, D. Impaired degradation of mutant α-synuclein by chaperone-mediated autophagy. Science 305, 1292–1295 (2004). This landmark paper is the first to suggest the possibility of dysfunction in CMA as a factor in the pathogenesis of PD. 169. Orenstein, S. J. et al. Interplay of LRRK2 with chaperone-mediated autophagy. Nature Neurosci. 16, 394–406 (2013). 170. Wang, Y. et al. Tau fragmentation, aggregation and clearance: the dual role of lysosomal processing. Hum. Mol. Genet. 18, 4153–4170 (2009). 171. Pacheco, C. D. & Lieberman, A. P. The pathogenesis of Niemann–Pick type C disease: a role for autophagy? Expert Rev. Mol. Med. 10, e26 (2008). 172. Ko, D. C. et al. Cell-autonomous death of cerebellar Purkinje neurons with autophagy in Niemann–Pick type C disease. PLoS Genet. 1, 81–95 (2005). 173. Kimura, S., Noda, T. & Yoshimori, T. Dissection of the autophagosome maturation process by a novel reporter protein, tandem fluorescent-tagged LC3. Autophagy 3, 452–460 (2007). 174. Elrick, M. J., Yu, T., Chung, C. & Lieberman, A. P. Impaired proteolysis underlies autophagic dysfunction in Niemann–Pick type C disease. Hum. Mol. Genet. 21, 4876–4887 (2012). 175. Ordonez, M. P. et al. Disruption and therapeutic rescue of autophagy in a human neuronal model of Niemann Pick type C1. Hum. Mol. Genet. 21, 2651–2662 (2012). 176. Maetzel, D. et al. Genetic and chemical correction of cholesterol accumulation and impaired autophagy in hepatic and neural cells derived from Niemann-Pick type C patient-specific iPS cells. Stem Cell Rep. 2, 866–880 (2014). 177. Sarkar, S. et al. Impaired autophagy in the lipidstorage disorder Niemann–Pick type C1 disease. Cell Rep. 5, 1302–1315 (2013). This paper attempts to clarify previous contentions in the role of autophagy in NPC1 and demonstrates a decrease in autophagic function in the disease. 178. Meske, V., Erz, J., Priesnitz, T. & Ohm, T. G. The autophagic defect in Niemann-Pick disease type C neurons differs from somatic cells and reduces neuronal viability. Neurobiol. Dis. 64, 88–97 (2014). 179. Motori, E. et al. Inflammation-induced alteration of astrocyte mitochondrial dynamics requires autophagy for mitochondrial network maintenance. Cell. Metab. 18, 844–859 (2013). 180. Smith, C. M., Mayer, J. A. & Duncan, I. D. Autophagy promotes oligodendrocyte survival and function following dysmyelination in a long-lived myelin mutant. J. Neurosci. 33, 8088–8100 (2013).


REVIEWS 181. Rangaraju, S. et al. Rapamycin activates autophagy and improves myelination in explant cultures from neuropathic mice. J. Neurosci. 30, 11388–11397 (2010). 182. Di Malta, C., Fryer, J. D., Settembre, C. & Ballabio, A. Astrocyte dysfunction triggers neurodegeneration in a lysosomal storage disorder. Proc. Natl Acad. Sci. USA 109, E2334–E2342 (2012). This paper demonstrates that astrocyte-specific deletion of Sumf1 in mice, a gene associated with multiple sulfatase deficiency, results in the loss of autophagy function in astrocytes and degeneration of cortical neurons. 183. Deretic, V., Jiang, S. & Dupont, N. Autophagy intersections with conventional and unconventional secretion in tissue development, remodeling and inflammation. Trends Cell Biol. 22, 397–406 (2012). 184. Kaufman, S. K. & Diamond, M. I. Prion-like propagation of protein aggregation and related

therapeutic strategies. Neurotherapeutics 10, 371–382 (2013). 185. Danzer, K. M. et al. Exosomal cell‑to‑cell transmission of alpha synuclein oligomers. Mol. Neurodegener. 7, 42 (2012). 186. Ejlerskov, P. et al. Tubulin polymerization-promoting protein (TPPP/p25α) promotes unconventional secretion of α-synuclein through exophagy by impairing autophagosome–lysosome fusion. J. Biol. Chem. 288, 17313–17335 (2013). 187. Lee, H. J. et al. Autophagic failure promotes the exocytosis and intercellular transfer of α-synuclein. Exp. Mol. Med. 45, e22 (2013). 188. Nilsson, P. et al. Aβ secretion and plaque formation depend on autophagy. Cell Rep. 5, 61–69 (2013). 189. Nixon, R. A. Alzheimer neurodegeneration, autophagy, and Abeta secretion: the ins and outs (comment on DOI 10.1002/bies.201400002). Bioessays 36, 547 (2014).


Acknowledgements

The authors are grateful for funding from a Wellcome Trust Principal Research Fellowship (D.C.R.), a Wellcome Trust and Medical Research Council (MRC) Strategic Grant on neuro degeneration, a Wellcome Trust Strategic Award to the Cambridge Institute for Medical Research, the Alzheimer’s disease National Institute for Health Research (NIHR) Biomedical Research Unit at Addenbrooke’s Hospital, the Tau Consortium and Alzheimer’s Research UK.

Competing interests statement

The authors declare competing interests: see Web version for details.

DATABASES European Clinical Trials Database: https://eudract.ema. europa.eu/index.html ALL LINKS ARE ACTIVE IN THE ONLINE PDF


The Function of Autophagy in Neurodegenerative Diseases.

Therapeutic implication of autophagy in neurodegenerative diseases.

Impaired autophagy: a link between neurodegenerative and neuropsychiatric diseases.

Autophagy and Human Neurodegenerative Diseases-A Fly's Perspective.

Emerging role of autophagy in pediatric neurodegenerative and neurometabolic diseases.

Autophagy in neurodegenerative diseases: from mechanism to therapeutic approach.

Autophagy in aging and neurodegenerative diseases: implications for pathogenesis and therapy.

Autophagy and Its Impact on Neurodegenerative Diseases: New Roles for TDP-43 and C9orf72.

Mitochondria, autophagy and age-associated neurodegenerative diseases: New insights into a complex interplay.

Selenium, selenoproteins and neurodegenerative diseases.

Ubiquitin, lysosomes, and neurodegenerative diseases.

Protein misfolding and neurodegenerative diseases.

Scrapie and human neurodegenerative diseases.

Ubiquitin, lysosomes and neurodegenerative diseases.

Autophagy and apoptosis dysfunction in neurodegenerative disorders.

Genetics of neurodegenerative diseases.

Optogenetics for neurodegenerative diseases.

Spermidine on neurodegenerative diseases.

Induced pluripotent stem cell-based modeling of neurodegenerative diseases: a focus on autophagy.

Human transmissible neurodegenerative diseases (prion diseases).

Emotional dysfunctions in neurodegenerative diseases.

Gene therapy for neurodegenerative diseases.

DNA damage in neurodegenerative diseases.

Pain Assessment in Neurodegenerative Diseases.